Since launching, over fifty thousand batches have been run and 20 million records enriched! Batch has quickly become one of the best ways to try out Enrichment.

Today, we've launched Batch Reports, an upgrade to the batch process that gives you a free breakdown of exactly what your data looks like before you purchase anything.

As soon as you upload a list of company domains, we'll start processing and generating a series of beautiful visualizations breaking down the data by location, category, company size, and more.

If you upload a list of emails, we'll also break down the results by the role and seniority of the people behind those emails.

Using reports

A great way of using this data is by analyzing your customer base or target accounts and analyzing the results.

For example, let's take a look at Clearbit's customers. Interestingly, over 80% of them use Google Apps. Looking at technologies shared between your customers can be a useful component of your lead scoring and qualification process.

If we look at the employment role of Clearbit buyers, sales and marketing positions stand out. As our data is quite useful for sales and marketing, this makes some sense.

We can see that over 80% of our customers are in the B2B space - another useful signal we can use when qualifying leads.

Lastly, if we breakdown Clearbit's customer base by company headcount, you can see our sweet spot is companies ranging from 50 to 250 employees.

Understanding your target customer is crucial, and what better way to do so than analyzing historical data. Batch Reports will give you answers in minutes, and best of all it's free. Give it a whirl and let us know how you get on!

In this example we're building a prospect list from companies with more than 200 employees, who have visited our website within the last day!

These target accounts can then be de-duplicated against your existing accounts, before you start searching for new sales prospects.

Our recent chapter on Automated Outbound Sales details how Segment’s growth team leverages Reveal data to drive highly targeted and personalized email campaigns. By focusing outbound on companies that have visited their website, the campaign has seen an open rate of 55%, and increased outbound email open rates by 200%.

With the Reveal Loop, we’ve empowered any organization using Salesforce to build lead lists based on website visitors.

If you are interested in learning more or seeing a demo, please contact success@clearbit.com

Live chat is a fantastic way to engage your site's visitors and can have a huge impact on increasing your conversion rates. However, live chat also has its downsides.

Unqualified sales conversations and increased support requests, all that need to be responded to in real time, can be a huge drain on any team.

The new Clearbit Reveal and Drift Messaging integration now allows you to proactively target your site visitors and only show live chat to qualified prospects.

Using this approach the marketing teams from our beta have been able to double the number of qualified opportunities they are generating from chat!

Segment’s VP of Growth Guillaume Cabane set up personalized campaigns for anonymous visitors and saw a 5x increase in engagement and 2x increase in conversion, and as a result, messaging has become one of Segment’s top sources of qualified opportunities.

How it works

Here are four ways your marketing team can start using the new integration.

1. Target based on industry and company size.

Clearbit company tags let you target companies based on high level industry categorization like Software or Banking. There are also tags for company business model, SAAS, Enterprise, Marketplace, etc.

To give you one example: at Drift our sweet spot is B2B SaaS companies, so we have a personalized message that only shows up when someone from a B2B SaaS company hits our website:

You can also target messages by employee size -- so here we’re targeting B2B SaaS companies with more than 50 employees:

Then a personalized message can be displayed to any website visitors that match that criteria, like this:

Not a B2B SaaS company with more than 50 employees? You’ll see a different message -- or nothing at all.

2. Target based on technology

In addition to targeting based on company tags, you can also send messages based on the technology that their company uses.

So let’s say you want to target B2B SaaS companies with more than 50 employees that are also using Salesforce, the rules look like this:

3. Send personalized messages to each website visitor

Here’s where this starts to get really cool.

If I go and visit the Segment website while I’m in the office at Drift, I see this personalized message “We have advice for Drift”:

How the heck did they know!?

Well, Segment is using Drift and Clearbit Reveal to serve up personalized messages for every website visitor in their target industries. Just like you’d personalize an email in other marketing automation products, with Drift you can personalize messages on your website.

But to take it a step further, with Clearbit Reveal you can deliver personalized messages to anonymous website visitors – it’s like magic. All you need to do is include the Clearbit company name tag when you’re creating a campaign in Drift:

4. Route the best conversations to your sales team

Lastly you can also use the data to determine which visitors your sales team is asked to engage with, so that their time is only spent with the best possible prospects. To do so, you can build routing on top of the firmographic data within Drift, such as company size:

Or industry (in addition to routing conversations based on technologies used and the amount of funding raised):

And here's the best part: all of this can happen before a website visitor has to fill out a form.

Imagine if your sales team is only asked to chat with the high-value accounts or specific target accounts you've hand-picked as part of an ABM strategy. This is exactly the reason why the Segment sales team loves this new approach to B2B marketing.

Instead of forcing people to fill out long forms and then waiting to hear back, they are having highly targeted conversations in real-time—and reps now know ahead of time what companies people work at, since all of the information from Clearbit gets displayed right inside of Drift as well:

With Clearbit and Drift, not only will you know when your best leads are live on your website - but you'll have everything you need to be able to reach out in real-time to give them the VIP treatment and take the fast lane right to your sales team.

If you'd like to learn more, and are near Boston, Drift will be hosting Guillaume from Segment to chat growth and tactics (including proactive chat).

Therefore it is critical that this channel to your customers and potential customers remains open. A low domain sender score will ruin your email reputation and quickly close that door. Your domain sender score is your trustworthiness score for mailbox providers. It's what decides whether they let you deliver your emails to their users, or whether they think you are spam and send your message into the ether.

Not understanding your sender score and what goes into it will reduce your ability to reach leads and customers, and more importantly directly hurt revenue. Understanding your sender score and how you can increase your score and deliverability of your emails, helps you make sure you can always reach the right customers without fear of being blacklisted.

What is your sender score?

Your sender score is a proxy for your domain's reputation as an email sender. The most natural analogy is with your credit score (credit history). The higher it is, the more mailbox providers—Gmail, Microsoft, Yahoo, et al.—are going to trust you with access to their customers. When it's low, you can find your emails undelivered or your domain blacklisted and unable to deliver to that mailbox at all.

There are a number of different companies that assess your domain sender score:

• Return Path's Sender Score and WatchGuard's Reputation Authority are numerical scores ranging from 0 - 100.
• Cisco's Senderbase and Barracuda Networks Reputation System are categorical classification systems that give Good, Neutral, or Poor ratings.

For some of these scores, you can checkout your domain or IP right now to see your current score. Whichever you choose to follow, they will take into consideration the following information to compute your score:

• Complaints: How often your emails are “marked as spam,” or forwarded to a postmaster or blacklisting service to make sure they are removed from the recipient's inbox?
• Blacklists: Is your IP address or domain already on a blacklist for previous email abuses?
• Sending to spam traps & unknown users: Are you consistently sending to email addresses that are out of use or set up to catch spam lists?
• Engagement: Does your content have low open rates, reply rates, or other metrics of engagement?
• Consistent volume: Are you sending to the same-size list constantly? You can be penalized for sending to fluctuating size lists as it shows that you don't have a consistent user base.
• Consistent frequency: Are you sending on a regular basis to that list. Again, changing the frequency constantly can hurt your score.

How each of these are weighted is different for each proprietary algorithm. But the most important offenses are sending to people who complain and sending to non-existent addresses. These are red-flags for spam.

The importance of sender scores for sales & marketing

It is vital to continually check and improve your sender score for two reasons.

1. If you have a low score, then your email will not be reaching the people you expect. They'll be undelivered or spammed. You'll have low engagement. You'll make less $$.
2. Your sender score is a symptom, not a cause of bad email etiquette. If your score is low, that means you aren't sending high-quality content to high-quality addresses.

Effectively, your sender score is not just a proxy for your reputation with mailbox providers. It is also a proxy for your reputation with leads, customers, and prospects. A low score means that you are not only not reaching the right people, but that you are not respecting the people you are reaching. It also gives you a measure of your efficiency.

This might seem that you are beholden to your recipients and random companies for whether your marketing emails or prospecting emails will be a success. But just like your credit score, your sender score is under your control.

The good news is that your sender score is dynamic. Return Path's Sender Score, for instance, is calculated on a 30-day rolling average. This means that if your score is currently low it can be improved and you can go from 0 to 100, or from poor, through neutral, to a good rating.

If you have a good sender score (for instance, above 80 on Sender Score) then you should look for continual ways to optimize your marketing and keep your score high. If you have a sender score less than 80, it can be rescued. Each of the data points used by companies can be improved through good list practices and good email practices.

How to avoid ruining your domain score: quality and hygiene

This entire process is in your control. Just like with credit scores, following a few basic guidelines can get you to a top score.

The quality of your email list is perhaps the single most important determinant of your sender score.

If you email a list that contains a high number of non-existent or non-working addresses, this is a strong signal to mailbox providers that you aren't taking time and consideration in your email marketing. They won't want to risk you sending ill-targeted emails to their customers, so they will blacklist you for this.

Therefore, having a list with a low bounce probability sets you up for success. Having a high-quality list comes down to two things:

1. List Quality: Getting correct email addresses in the first place
2. List Hygiene: Constantly assessing and pruning your list to only email desired recipients

List quality

The first hurdle is high-quality email addresses. If you are using a lead generation form, it's possible people will use a fake email address just to get access to gated content. You should verify these addresses either through a double opt-in verification email or through enrichment (http://clearbit.com/enrichment).

If you are building an outbound list, then you have to be extremely careful about using the emails on the list. Suppliers that emphasize quantity over quality will include unknown user and spam trap addresses that will kill your domain reputation:

• Unknown users: An unknown user could be an email that never existed in the first place, or has been abandoned or terminated. These will return a 550 error code. If just 1 in 10 of your emails returns a 550 error, this will cause email deliverability issues for the rest of your list. If you then resend to those addresses and it bounces again, it signals to mailbox providers that a) you don't care about list quality, and b) you don't care about list hygiene.
• Spam traps: These are email addresses set up exclusively to catch spam marketers and list providers that indiscriminately scrape websites for email addresses. The email address is usually hidden from human view but will be harvestable for a bot. If you send an email to a spam trap, it shows you aren't actively trying to reach real people and your domain reputation will take a severe hit.
• Email guessing: Some email list providers will include calculated (or guessed) email addresses that are typically some combination of first name + last name @companydomain.com. Lists of generated emails typically have an average bounce rate of 30% and will quickly destroy your sender score.

List hygiene

If you are unsure of the legitimacy of the emails on your list, then enacting list hygiene techniques can ameliorate this to a certain extent:

• Remove obvious email addresses that have no specific end-user (support@company.com, test@company.com), as well as common errors (matt@emial.com).
• Quarantine new information and check for these errors before adding to your main email list.
• Email regularly and consistently so that bad addresses are removed over time.
• Use bounce management to determine the source of bad emails.

This last technique, bounce management, is something that you should take seriously. It allows you to make sure your list gets more accurate over time, increasing your domain score. As each bad email is sent back, don't just delete it from your list. Use the data gathered to find out where you got this specific email from and why it bounced. If it was from a user, look at double opt-in and other verification options. If from a list, use a different provider that emphasizes quality over quantity.

Continual improvement

Though list quality is the most fundamental component to a high sender score, your score is a combination of other factors that you should also take into account:

• Email content: Anti-spam filters will scan your emails looking for red-flag content and issues. While some words that will get you blacklisted are well-known, other elements that evoke filtering are high use of images, and incorrect use of HTML. Check that you have a good balance of text to images in your email, and that everything is rendered correctly.
• Engagement: Marketers will use engagement metrics to assess the success of their campaigns—opens, clicks, conversions—but mailbox providers will also look at these metrics to determine the quality of your emails. Low engagement is a signal that the content is unwanted and therefore spam. They also have access to extra quantitative information about your campaigns such as spam complaints, forwarded email, and email archive rates, that they can use to effectively monitor your comprehensive reputation.
• Consistency: You want to send to a regular list on a regular basis, showing that you are just flooding the mailbox providers' customers with email spam. You want your list to grow naturally over time instead of adding masses of new emails all at once.

Each of these points are important whether your list is inbound or outbound. Ultimately what separates legitimate email marketing and sales from spam is targeted content to a targeted list. Curate a great list, send some great emails, and you'll always have a solid sender reputation and stay out of your prospects spam folders.

TL;DR

Integrating Auth0 and Clearbit can make the signup process as smooth as possible. In this article, we will see how to combine these amazing solutions to automate signup forms, which can lead to improved conversion rates.

Introduction

Signup forms play a major role on conversion rates, for better or worse. We don't want to ask for too much information upfront, otherwise we will add friction and reduce the total number of signups. But, on the other hand, we can get a lot of value from knowing more than a user's email address. For example, if we have enough data about our customers, we can send highly personalized onboarding emails to increase adoption and ultimately conversion.

This tricky situation is what we are going to tackle throughout this article. Here, we will see that we can solve this issue by making the signup process smooth, with Auth0, and by automatically enriching users' profiles, with Clearbit

What is Auth0

Auth0 is an enterprise-grade platform for modern identity. The company provides tools that eliminate the friction of authentication for applications and APIs. With it we can easily and quickly connect our apps, choose identity providers to rely on, set up sign in and sign up rules, and customize our login page. It really is identity made easy for developers.

What is Clearbit

Clearbit is a company that provides intelligence about customers and their workforce to help our business grow. By using Clearbit, we can get access to valuable information about our users without even asking for it. We can check their role on the company that they work for, get the company size, and use this data to help our marketers and salespeople to outperform the competition.

Signing Up for Auth0

The signup process to Auth0 is quite simple. Let’s head to Auth0’s website and click on the Sign Up button at the top. This link will bring up a beautiful component that will allow us to sign up, or log in if we already own an Auth0 account.

On the demo application that we are going to develop, we will use exactly the same component as Auth0 in the sign up/sign in process. This component is called Lock and is an open source project that Auth0 made available on GitHub. Lock is very easy to configure and to extend, as we will see soon.

For now, let’s just choose one of the available signup options to complete the account creation. After choosing it, Auth0 will ask some questions about how we intend to use its solutions and then we will be good to go.

Signing Up for Clearbit

Now that we have our Auth0 account, let’s go to Clearbit and create an account there as well.

Creating Our Project

Having both the Auth0 and the Clearbit accounts created, we can now focus on the demo application that we are going to build. We will go step by step, creating and configuring everything that we need. But, if you prefer, you can download the fully configured demo app from GitHub.

Node and npm are expected to be installed on the development computer so, if you don't have them, please follow the installation instructions on npm's website.

To start, let's go through the following commands:

mkdir auth0-clearbit  
cd auth0-clearbit  
npm init -y  
npm install --save angular-jwt angular-lock auth0-js clearbit  
npm install -g http-server  

The first two commands issued are responsible for creating a directory for our application and for changing the current working directory to it. The third command, npm init -y, is responsible for making this directory the root folder of our demo app, by creating a package.json file that will hold the details about our project.

The last two commands are responsible for installing all the dependencies that our application have. The first one of them install runtime dependencies, and the later installs a npm module that runs a lightweight HTTP server that is perfect to test our application.

The next step that we will perform is to create the index.html file in the root directory of our application. This file will have the following code:

<!DOCTYPE html>  
<html lang="en">  
  <head>
    <meta charset="utf-8">
    <meta http-equiv="X-UA-Compatible" content="IE=edge">
    <meta name="viewport" content="width=device-width, initial-scale=1">
    <!-- The above 3 meta tags *must* come first in the head; any other head content must come *after* these tags -->
    <title>Next Big Thing</title>

    <!-- Bootstrap -->
    <link href="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/css/bootstrap.min.css" rel="stylesheet">
    <style>
      [ng\:cloak], [ng-cloak], [data-ng-cloak], [x-ng-cloak], .ng-cloak, .x-ng-cloak {
        display: none !important;
      }
    </style>
  </head>
  <body>
    <div class="container" ng-app="myApp">
      <div class="row">
        <div class="col-xs-12" ng-controller="homeController as homeCtrl">
          <h1>Next Big Thing</h1>
          <div class="jumbotron" ng-show="!homeCtrl.isLoggedIn()">
            <p>Welcome to my great, actually amazing, application.</p>
            <p><a href="" ng-click="homeCtrl.login()">Sign up now</a>, it is ultra fast.</p>
          </div>
          <div class="jumbotron" ng-show="homeCtrl.isLoggedIn()" ng-cloak>
            <h2>Completing Your Data</h2>
            <p ng-show="homeCtrl.user == null">
              Sit back and relax while we fetch your data.
              <img style="width: 18px;" src="http://cdnjs.cloudflare.com/ajax/libs/semantic-ui/0.16.1/images/loader-large.gif" />
            </p>
            <p ng-show="homeCtrl.user != null">
              Please, confirm your data below.
            </p>

            <div class="form-group">
              <label>Full Name</label>
              <input class="form-control" ng-model="homeCtrl.user.enrichment.name.fullName">
            </div>
            <div class="form-group">
              <label>Full Name</label>
              <input class="form-control" ng-model="homeCtrl.user.enrichment.location">
            </div>
            <div class="form-group">
              <label>Company</label>
              <input class="form-control" ng-model="homeCtrl.user.enrichment.employment.name">
            </div>
            <div class="form-group">
              <label>Position</label>
              <input class="form-control" ng-model="homeCtrl.user.enrichment.employment.title">
            </div>
            <div class="form-group">
              <label>Biography</label>
              <textarea class="form-control" ng-model="homeCtrl.user.enrichment.bio"></textarea>
            </div>
            <button type="submit" class="btn btn-primary pull-right disabled" ng-show="homeCtrl.user == null">
              Loading your data...
            </button>
            <button type="submit" class="btn btn-primary pull-right" ng-show="homeCtrl.user != null">
              Go Ahead
            </button>
          </div>
        </div>
      </div>
    </div>
    <script type="text/javascript" src="https://cdn.auth0.com/js/lock/10.13/lock.min.js"></script>
    <script type="text/javascript" src="https://cdnjs.cloudflare.com/ajax/libs/angular.js/1.6.1/angular.js"></script>
    <script src="node_modules/auth0-js/build/auth0.js"></script>
    <script src="node_modules/angular-lock/dist/angular-lock.js"></script>
    <script src="node_modules/angular-jwt/dist/angular-jwt.js"></script>
    <script src="app.js"></script>
  </body>
</html>  

This is the only HTML file that we will create throughout this article and its contents are quite simple. The body section is divided into two parts. The first one is shown when the user is not logged in, and provides a link to sign up. The second one is shown after a successful sign up.

The magic for the user happens in the second part. Here they will face a form to complete the signup process. Usually the user would be responsible for filling this data but, with the help of Clearbit, the user does not have to fulfill these fields manually. The whole form is filled automatically by the data fetched by Clearbit.

To complete the demo application we still need to create the JavaScript file that will give life to our app. Let's now create this file, naming it app.js, and fill it with the following code:

(function () {
  'use strict';

  var myApp = angular.module('myApp', [
      'auth0.lock',
      'angular-jwt'
  ]);

  myApp.config(function config(lockProvider) {
    lockProvider.init({
      clientID: 'ADD_YOUR_AUTH0_CLIENT_ID_HERE',
      domain: 'ADD_YOUR_AUTH0_DOMAIN_HERE'
    });
  });

  myApp.controller('homeController', function(lock) {
    var ctrl = this;

    ctrl.login = lock.show;
    ctrl.loggedIn = false;

    ctrl.isLoggedIn = function () {
      return ctrl.loggedIn;
    }

    lock.on('authenticated', function (authResult) {
      localStorage.setItem('id_token', authResult.idToken);
      ctrl.loggedIn = true;

      lock.getProfile(authResult.idToken, function (error, profile) {
        if (error) {
          return console.log(error);
        }
        profile.enrichment = profile.enrichment ? profile.enrichment : {};

        ctrl.user = profile;
        localStorage.setItem('profile', JSON.stringify(profile));
      });
    });
  });
})();

This JavaScript file has three responsibilities:

1. it creates an AngularJS module, called myApp, that is used by our index.html file
2. it configures the Lock component with our Auth0 Client ID and our Auth0 Domain
3. it creates a homeController that controls our HTML page, defining what to show and when, and providing integration with Lock

Whenever a user signs up to our demo application, the Lock component gets an authenticated event with the authResult (authentication result) object. We then use the idToken property of this object to get the profile of the user that just signed up. As we can see in the code above, the Auth0 answer to the getProfile call will return to us a profile object that contains an enrichment property. This is the property that will hold all the employment data about our new user.

Attention! If you want to use Auth0 authentication to authorize requests to APIs, you'll need to use a different flow depending on your use case. Auth0 idToken should only be used on the client-side. Access tokens should be used to authorize APIs. You can read more about making API calls with Auth0 here.

It is important to note that this property is not part of the getProfile response by default. The data contained in this property will be fetched from Clearbit by an Auth0 rule that we are going to create soon.

Note that the source code above exposes two properties that must be changed: ADD_YOUR_AUTH0_CLIENT_ID_HERE and ADD_YOUR_AUTH0_DOMAIN_HERE. The values that we must add in the place of these two properties can be found in the Clients Page of our Auth0's Dashboard. In this page we will find all the clients created for our account–which will be just one if we haven't manually created any other–and choosing the Settings option of one of these clients will lead us to a page where we can find the properties.

From this page, we can copy the contents of the Domain field to replace ADD_YOUR_AUTH0_DOMAIN_HERE property in the JavaScript file, and then copy the contents of the Client ID field to replace the ADD_YOUR_AUTH0_CLIENT_ID_HERE property.

As the last step to properly configure our integration with Auth0, we will add http://localhost:8080 as the value of the Allowed Callback URLs field of this page. This configuration is important to allow our demo application to communicate with Auth0 when running it from our development machine. If we were to publish this demo application on a real server, responding on behalf of a real domain, we would need to add this domain to this field as well. Otherwise, Auth0 would deny, as a security measure, the integration.

Our demo application is now ready to run, and it is already working with Auth0 to allow users to sign up. But it still lacks the integration with Clearbit to automatically fill the personal data of the user signing up. Let's tackle this last task now.

Integrating Auth0 with Clearbit

Clearbit provides a RESTful API that is well structured, properly documented and easy to use. They also provide libraries for common development platforms–like Ruby, NodeJS, and Python–that make integration with their APIs even easier.

To complete our demo application we are going to use the NodeJS SDK, but we won't add it on our source code. We will perform this integration with Auth0 Rules.

Auth0 Rules are functions, written in JavaScript, that are executed in Auth0 as part of the sign in and sign up transactions every time a user authenticates to our application. These rules allow us to easily customize and extend Auth0's capabilities and, in our case, we will use them to make requests to Clearbit's APIs to fetch data from the user that is signing up.

To start this integration, let's head to the API Keys page on the Clearbit's Dashboard and copy the first key shown there, which Clearbit refers to as Secret API Key. This secret key will be used by our Auth0 rule to communicate with Clearbit's API, as we will see below.

Now that we have this key, let's head to the Rules page on Auth0's dashboard and click on the Create Rule button. After clicking on it we will see a page that allows us to start creating our rule based on a template. For now we will choose the first button, called Empty Rule, which will allow us to create a rule from scratch.

When we choose this option, we are redirected to the page where we will create the rule. On it, the first thing that we will do is to define the rule name to something like Clearbit Enrichment Rule. Below the rule name, there is a dark text area where we can add the code that will perform this enrichment. Let's add the following source code to the field:

function (user, context, callback) {  
  var clearbit = require('clearbit@1.2.3');
  require('bluebird@3.4.6');

  var clearbitClient = clearbit('YOUR_CLEARBIT_SECRET_KEY');

  var Person = clearbitClient.Person;

  // It is important to use `stream: true` because we need to
  // wait for a response from Clearbit.
  // `stream: false` can lead to Clearbit responding that the definitive response
  // is queued and will be available in the future.
  Person.find({email: user.email, stream: true}).nodeify(function (err, person) {
    if (!err) {
      user.enrichment = person;
    }

    callback(null, user, context);
  });
}

Auth0 rules are expected to hold a single function, as shown above, that have access to three arguments:

• user – that holds all the user's data, like email, name and etc
• context – an object containing contextual information of the current authentication transaction, such as user's IP address, application, location and so on
• callback – a function to send back the potentially modified user and context objects back to Auth0 (or an error)

In our rule, we change the user object by adding data that we fetch from Clearbit. This is done with the help of the clearbit-node npm module that we require on the second line of the code above, and by calling the Person.find function on this module. Note that before running our application, we must change the YOUR_CLEARBIT_SECRET_KEY property in the source code above with the Secret API Key that we copied from Clearbit's dashboard.

This was the last step that we needed to perform before being able to run our demo application. Therefore we can head to our app's directory and run http-server on it. After running this command we can open http://localhost:8080 on a web browser and test the signup process with automatic enrichment.

Conclusion

As you can see, integrating Auth0 to Clearbit to enrich our users' profiles is a piece of cake. We just need to create a single rule, on Auth0, that communicates with Cleabit's API and we are good to go.

Note that the demo application that we've just built is just a proof of concept and its workflow could be improved even more. We could, for example, avoid showing the form that asks for the employment data of our new users by relying on the data that Clearbit sends us.

Having a signup process that relies on Auth0 and Clearbit can give our users the best signup experience ever. With Auth0, users can signup by pressing a single button and with Clearbit they don't even need to fill tedious forms.

Unfortunately, it's also one of our hardest products to trial and is not entirely self service. Installing a Salesforce package is no mean feat, and can require some tricky upfront configuration.

That all changes today with the release of our Salesforce Lite Chrome Extension, which installs with one click then injects our Salesforce integration directly into Lead, Contact and Account pages in Salesforce CRM. No configuration or installed packages required!

The best news is that we're giving this away for free! Give it a whirl. We support both Salesforce Classic and Lightning versions.

Contextual demographic & firmagraphic data

Salesforce Lite displays fresh and real time person & company information directly next to your Salesforce Leads, Accounts and Contacts.

We realize screen real-estate is super important in Salesforce, so we've provided a minimized view of this information too.

Fetch contact emails

We've also baked in our Prospector product, so you can find up to 20 emails every month for free. You can create a new Lead or Contact inline, and we'll pre-fill in email, name, and title information.

Upgrade available

Now we want to emphasize that this extension, while useful standalone, is a taster of all the functionality you get with our paid product, Salesforce Pro.

With our Salesforce Pro enrichment solution you can embed the data in custom objects, standard Salesforce fields, and reports. This gives you the ability to use the data anywhere you could want it in Salesforce. This helps you keep your CRM data quality high, and your reps happy.

Additionally we've got a souped-up version of our Prospector which will let you search for companies by 80 different criteria, find leads and de-dupe against your existing database.

Marketing would be so easy if my site visitors just knew exactly what to read, where to click, how to signup/purchase….

Easier said than done. Prospects have found your website and you want to point them towards the best information or solution designed for their needs. But you have so many products, case studies, resources, pricing options…. it can be overwhelming and challenging for your buyer to know where to even look, let alone engage.

However – there's hope!

Reveal enables real-time personalization of your site based on the company information of the visitor - even if they haven’t self-identified or signed up!

With dynamic web personalization in place, you can adjust the site experience based on a visitor's specific company, industry, size, tech stack, and and more. And the best part is that no-coding is required!

How it works

Drop in a small javascript snippet and Reveal will start listening to your site’s anonymous traffic. With that in place, you can define your various visitor segments and begin creating personalized experiences.

To create, manage, and measure your site variations, we suggest using Personalization software. Behind-the-scenes, we've created integrations with 3 powerful, modern platforms:

If you use Reveal with Google Optimize, Optimizely, or GetSmartContent, plug-and-play Personalization is at your fingertips! Unlike some of the old school companies in the space, you won’t need a 3 month consulting engagement or dev resources to start running personalization.

For example, below is a .gif that shows creation of an audience segment in Optimizely that targets any visitors at companies with 50+ employees.

Using a personalization platform also allows you to incorporate behavioral attributes in as well - so you can create the ultimate experience based on their actual site engagement and their company firmographics.

Some practical examples

• Running an account-based marketing strategy? Present your new case study about how enterprise-capable your product is.

• Have vertical-specific solutions? Display your healthcare customer logos when a Pharma company is checking out your site instead of a generic set.

• Someone is using your competitor's tech? Highlight the “free trial” button and share your new competitive analysis white paper.

Today, companies like IBM, Invision, and Segment have successfully deployed Clearbit Reveal for dynamic site personalization. We’ll be publishing much more as far as best practices and success stories - stay tuned!

Pricing starts at $999/month and tiers based on your site's web traffic. To get started, ping success@clearbit.com.

B2B companies invest millions in demand gen and then instantly lose visibility into how buyers actually engage with their site, content, and products. This leads to missed conversions, poor marketing attribution, and the same digital experience for your inherently unique visitors. But what if you had more insight into who was behind each site visit?

While personalization is becoming the norm for post-signup engagement, it can be game-changing to have if you had this data ahead of time – even if only at the firmographic level.

By identifying anonymous traffic, you can tailor your site (or app) experience, leading to higher conversion rates, quicker buying cycles, and better insight into how the market engages with your business.

Marketing's survivorship bias

Website conversion rates in SaaS sit at about 7%:

Think about what this means. 93 out of every 100 people who come to your site leave. This could be for many reasons:

• They aren't finding your content or site appealing.
• They might not be a good fit for your products.
• You aren't interesting.

We can't help you with the third bracket and you probably don't want prospects from the second. The first bracket, however, is a problem that can be solved.

One of the reasons you may be losing them is survivorship bias.

In the second world war, the Navy wanted to know where to put armor on planes to help them survive. Initially, they plotted all the hit marks on returning planes and suggested adding additional armor where there was damage.

Then came along statistician Abraham Wald. He pointed out that they should do the opposite. They should armor the parts were there was no damage. The returning planes had been hit in these areas and survived. That means these weren't critical areas. The results were skewed by a bias towards surviving aircraft—survivorship bias.

Wald showed that they should be reinforcing the areas that were unscathed. This was where the non-returning bombers were hit to cause them critical damage.

(Source: wikipedia)

Your site suffers from survivorship bias. It is built around the visitors that make it through and convert to customers. This is to be expected as they are the only ones you really understand. You only have data on those people—who they are, what they like within your product, what their needs and motivations are—so everything is skewed towards them.

Conversely, you can't build your site better for your anonymous visitors, the ones that don't survive to conversion, because you have no data on them. To fathom these churned visitors, you basically only have two sparse choices:

• General demographics from Google Analytics (GA) about location, referrer, and consumer interests, or
• Demographics from lead generation forms offer some gated content in exchange for an email address.

But general GA traits don't really offer much in the way of business insights for B2B companies. Lead gen can work, but it still relies on your content being appealing in some way to them.

This is the paradox of pre-conversion visitors. You need to understand who they are to offer what they need, but you can't offer if you don't have more data about them.

Increasing your survival rate by identifying your anonymous visitors

You don't have any data on those 93 anonymous visitors. You can't know what works for them because you don't know them.

But you could know them. And with intelligence around your anonymous visitors, you can optimize your marketing funnel to convert more of the 93 visitors into qualified leads.

If you follow this guide to setting up Reveal in Google Analytics, you'll be able to pull Clearbit company data into your analytics for visiting companies. Every site visit has an IP address - Reveal matches an IP address with company profile information. So if someone from Salesforce HQ went onto your site, Reveal would tell you that you have site visits from Salesforce.com, neatly located within GA.

Let's run through two approaches that use Reveal to de-anonymize your visitors and increase your conversion rates for them.

Increase conversion from different industries

With Reveal you can easily see your conversion rates for different types of industries. In Google Analytics, you can create a widget that will constantly update with the number of visitors from an industry, along with their conversion rate:

This shows data on all visitors, not just those that have converted. You can see the four different categories of visitors here, segmented by high traffic (>200 sessions) and high conversions (<6%):

1. High traffic, high conversion: Software and hardware companies
2. High traffic, low conversion: Advertising
3. Low traffic, high conversion: Diversified capital markets, general internet, and consulting
4. Low traffic, low conversion: HR, corporate, research, and payments

The companies in group 1 are the ones already finding value in the marketing efforts. These are the survivors. But they survive possibly to the detriment of the other categories.

Advertising is the best example from this group of industries. There is substantial traffic from this industry, but they don't have quite the conversion numbers of others. The fact that so many come to the site means they are interested, but the fact that they don't convert shows dissatisfaction. It could be that the site is too geared towards software and tech companies and that advertising visitors can't find the right information to stay interested.

Through this quick insight you can have a quick optimization:

• Change the copy on the page to be slightly less tech-heavy and more laymen.
• Include advertising customer testimonials, case studies, or logos.

These are changes that can be implemented in a day. Then, with this data, you can test and check if the changes had a positive effect.

You can then optimize for the two low-traffic groups, but off-site. Diversified capital markets and consulting, for instance, have the highest conversion rates. By advertising towards these groups you can drive traffic to your site and increase conversions simply through increased numbers.

You can find out how to set up this segmentation here.

Increase conversion from ideal companies

Many B2B businesses will have strategic, or "named", accounts they want to focus on. Using Reveal, its simple to generate a widget showing if, when, and how these key accounts are engaging with you:

This shows you the de-anonymized page views on your site from different visitors. If these visits aren't leading to meaningful interactions, you can now use intelligence to better capture their hearts and minds. There are a few ways you could do this:

• On-site case studies: Pick similar companies to these that are already your customers to study. Show these visitors exactly how they could be using your product. For instance, if you already have CRM customers, profile these to show Salesforce what you are all about.
• Off-site marketing: If you are running account-based marketing and sales, then when people from your named accounts visit, use this as a trigger to contact the company. Say you saw that they visited and want to help them find the information they are looking for.
• On-site customization: If you are using the Reveal API, you can personalize your site in real-time while key accounts are browsing. It could be something as obvious as changing the homepage to “We ❤ Stripe” or something as subtle as putting a testimonial from a similar company on the front page.

You can find out how to set up this segmentation here.

You need to know more to convert more

You need to understand your entire funnel to optimize your entire funnel. If you are just using your successful, surviving customers as a guide, you could potentially be missing out on a whole swathe of other great opportunities.

By identifying your anonymous visitors, you have that chance to understand who is in every stage of your funnel, and why some don't make it all the way through. Then you can start to ask why, start to help those visitors and start to turn those visitors into customers.

Salesforce is an ideal place to do lead qualification and scoring as it is typically the system of record and where all lead, contact, and account information lives. Building a lead scoring process in Salesforce offers two key benefits:

1. Enriched data with demographic attributes can be easily added.

2. All recorded sales activity can be used within your scoring formulas.

It's important to note you don't have to include all of your data in your lead scoring. It is often best to start with a basic lead scoring model and then build over time as you define your ideal customer profile (ICP) and understand what activities predict conversions.

When starting, demographic and firmographic data can form the base of your model. Who your lead is (and what company they work for), is very often more important than what actions they are taking on your site or within your app. This seems to be especially true in the B2B SAAS market.

The first step to scoring leads is to get that data into your Salesforce instance. Using the Clearbit Salesforce Integration, or something like data.com, you can append all relevant personal, social, and company information to each and every record you have in Salesforce.

Clearbit data inside Salesforce: All enrichment fields are available for any Salesforce record

From there, you can start building out your scoring rules based on your new rich standardized data set.

Below, we've put together a quick tutorial to get you started using just a custom field and a bit of Apex to build a real-time custom lead score field on every lead in SFDC.

Lead qualification isn't particularly difficult, but it is extremely important to get right. When you aren't effectively (and automatically) qualifying leads as a part of your sales process, you are simply wasting time and money.

Simple lead qualification and scoring is easy to set up with Clearbit and Salesforce. Doing so will take you five minutes, but save you hours. It will also boost the quality of leads sent to your sales reps, and make sure you are only talking to the right leads for your business.

Building your own custom lead score in Salesforce.

Requirements:

1. Salesforce permissions to create custom formula fields
2. Clearbit for Salesforce installed

Step 1: Create a custom formula field on the lead object

Choose Formula under Data Type.

Name your new Custom Field and set Formula Return Type to Percent and Decimal Places to 2.

Step 2: Paste in this pre-built Apex code.

(
CASE(cbit__Clearbit__r.cbit__CompanyTags__c, "B2B", 5,0)  
+
CASE(cbit__Clearbit__r.cbit__CompanyType__c, "public", 5,0)  
+
CASE( cbit__Clearbit__r.cbit__CompanyGeoCountryCode__c , "US", 10, 0)  
+
CASE( cbit__Clearbit__r.cbit__CompanyMetricsEmployeesRange__c, "1000+", 7, "251-1000", 8, "51-250", 10, "11-50", 4, "1-10", 2, 0)  
+
IF(CONTAINS( cbit__Clearbit__r.cbit__CompanyTech__c , "salesforce"), 10,  
IF(CONTAINS( cbit__Clearbit__r.cbit__CompanyTech__c , "marketo"), 8,  
IF(CONTAINS( cbit__Clearbit__r.cbit__CompanyTech__c , "google_analytics"), 6,0)))  
)
/
40  

Step 3: Modify fields and scores to match your own requirements.

Don't forget to change the final number to reflect the total potential sum if each criteria returns the max value. For example, in the default code above, the maximum possible value is 40.

Pro Tip: Custom fields in Salesforce tend to have names that can get quite long and unwieldy, so don't forget to use the Insert Field button above the formula edit box to search through available fields and add them to the formula with a single click.

Step 4: See it in action!

That's it! You now have a simple and robust lead scoring system directly within Salesforce, which you can use for lead routing, prioritization, and reporting.

Next step: Try incorporating other attributes like lead source, or sales activity into your model.

At the time the company consisted of me working out of my bedroom. When we added a sales team they made calls from the laundry room. Fast forward to today, and we have a team of 17 working out of a beautiful converted yoga studio in San Francisco's Mission district. A lot has changed, but some things haven't.

We've always tried to do things a little differently than your typical San Francisco startup, largely eschewing the traditional funding route by building a sustainable business. We've been profitable almost since day one, and while we've made some big bets, we've always been careful to keep costs under control. When you're relied upon at a systems level, it's important that you're here to stay.

We set out with an ambitious vision to make the world's best data provider. Visions take time to build though, and it's all too easy to lose sight of the forest for the trees. One good yardstick I've found to measure your progress along the way is by looking at the impact you have on your most innovative customers. Therein lies your future.

Take Segment for example. Clearbit data courses through that company, informing everything from lead qualification, to customized drip emails, to shortening their signup form, to customizing their home page. This is a truly data driven company, and Clearbit is behind the scenes powering a lot of it. We believe this kind of approach represents the future of sales and marketing.

While our market research initially consisted of taking a stab in the dark, these days we do things differently. We work with our best customers, iterating on feedback and improving our products. And sometimes our customers come up with completely new product lines. Indeed, later this year we're releasing a product first conceived by the team at Zendesk.

A good rule of thumb is that the simpler an interface, the more complexity lies behind the scenes. That certainly holds true here. Powering Clearbit is some serious tech that deals with parsing billions of rows of data, looking at government records, ssl certificates, whois info, classifying site text, and technology detection. Our engineering team has built out something quite incredible. Underlying Clearbit are 45 internal APIs and 180 git repos running on a Kubernetes cluster of 100 servers.

At our heart, we're a technology company. And as an engineer by trade, it's important to me this never changes. We've invested a lot in tooling. In fact, we even have an internal company policy that it should never take more than a five minutes to boot up a new API and service. This ensures we scale horizontally and that engineers can feel free to experiment.

Personally the most exciting aspect of this for me, is that the best is yet to come. We've laid the groundwork for something remarkable, and most importantly we have the team to execute on it. I'm looking forward to sharing with you more of what we've been working on this year.

The process of filtering through those thousands of opportunities to find the best ones is called lead qualification. This is an integral part of the sales process. Without it, your sales and marketing teams won't be able to prioritize their activity and will waste time on leads who will never convert.

In this post, we will take you through the basics of lead qualification, show where some of the traditional approaches are failing, and show you some new techniques that are enabling modern companies to quickly and effectively score leads. We've even included a bonus lead qualification formula at the end for Salesforce users.

What is lead qualification?

Lead qualification is the process of determining whether a prospect fits your ideal customer profile (ICP), has a high chance of becoming a customer, and most importantly has a high chance of being a successful long-term customer.

Sales funnel: Your qualification process is part of your sales funnel and determines which leads turn into prospects.

In order to effectively qualify and score your leads, you need to know as much as possible about each one. You are essentially gathering data and insights that will help you pick the leads to focus on.

On the surface this seems like you might be leaving money on the table by not pursuing all leads. But in reality, your sales team should only pick up the phone for people who have a strong likelihood of becoming a successful customer.

Successful is an important qualifier in that phrase. There are costs to acquiring non-ideal customers that can easily be overlooked in the race to acquire more customers. Increased churn, support costs, and unhappy users are all side effects of selling to non-ideal customers.

Without proper qualification, your sales and marketing team will spend equal time on all leads, whether they are a good fit for the company or not. With qualification, your teams can be confident they are prioritizing the best potential customers for your product.

How lead qualification works

Traditionally, lead qualification has been a manual task performed by Sales Development Representatives (SDRs) in the early stages of the sales process. Reps would reach out to any and every lead coming in, asking them pre-determined questions to assess if they were a good fit for the product.

There are a number of different frameworks created to help reps qualify prospects.

• BANT (Budget, Authority, Need, Timing)—devised by IBM. This is the original lead qualification framework and is still used widely today. Potential BANT questions are:
  • Do you have a specific budget allocated for this acquisition?
  • Do you personally have the authority to make the buying decision?
  • What are the top challenges, needs, and pain points that your team is facing at the moment?
  • Is this a priority acquisition for your team at this time?
    
    
• GPCTBA/C&I (Goals, Plans, Challenges, Timeline, Budget, Authority, Negative Consequences and Positive Implications)—devised by HubSpot to put the sales rep in a more advisory role for informed buyers. Potential questions are:
  • Do you have specific company goals?
  • Are the resources in place right now for you to implement this plan?
  • How are you currently dealing with these challenges?
  • What's the timeline for implementation of this plan?
  • Is the budget already being used to solve the problem we’ve discussed?
  • What concerns do you think the company's authoritative decision makers will raise?
  • What are the personal consequences or implications for you hitting your goal or not?
    
    
• CHAMP (Challenges, Authority, Money, Prioritization)—devised by InsightSquared (https://www.insightsquared.com/2014/03/dont-bant-just-champ-sales-qualification-questions-for-champions/). CHAMP starts with the challenges, the needs and pains, that the prospect is experiencing. Potential CHAMP questions are:
  • What objectives are you looking to achieve by solving this pain? (Challenges)
  • How are purchasing decisions made for products like ours and who is involved in looking at this solution? (Authority)
  • What are your expectations for the investment necessary to purchase the solution? (Money)
  • When were you planning on starting the implementation? (Prioritization)

Each of these pros and cons: BANT is simple and quick, but misses some modern aspects of the buying process, such as the increasing number of stakeholders involved in each decision. GPCTBA/C&I takes into consideration the sophistication of the modern buyer, but is a complex qualification process not optimal for smaller teams. CHAMP puts the emphasis on the problems the buyer is facing, but can potentially slow down your sales cycle.

Though there is still some need for a manual component to lead qualification, modern lead qualification looks to automate as much of the process as possible. To do this you can compute a lead score based on data available about the lead. There are two ways to do this:

1. Data enrichment in sales and marketing CRMs allow you to build lead scoring models using the specific data points important to your company and customers.
   
2. Predictive lead scoring tools, such as Infer or MadKudu, use thousands of different indicators (including 3rd party enrichment), to assign a lead score to each new lead.

The one downside of these predictive tools is that they are “black box” software. You are unable to see exactly what activities and attributes are used to formulate your lead scores.

Often, building your own simple model within your CRM will give you better insight into your scoring, and more control over tweaking your model over time. Though this may seem difficult, it is actually quite easy to set up a simple model in your CRM.

The individual components of a lead score

A lead score is a numerical expression of a prospects potential. It quantifies behavior and information into a single number that can be used to automatically channel leads to the right team.

Though a lead score can potentially include dozens of data points, it effectively consists of three core components:

1. Behavioral attributes

These are explicit actions that indicate interest or lack of interest in your product. Consider two visitors to your marketing site:

• Visitor A lands on your site through a Google search for your company name. They visit “Product”, “Pricing”, and “Customers” pages, read three articles on your blog and enter their information into a lead form for an upcoming webinar.
• Visitor B comes to your site through a keyword search, visits your “Pricing” page, reads a single blog article, then leaves.

Lead scoring through behavior

From these descriptions it is clear that Visitor A has a higher interest in your product than Visitor B. This component of lead scoring quantifies this activity. If in the above example each page visit was worth +1, a direct referral +2, and form submission +3, visitor A would score +11 while visitor B would only score +2.

If you set your threshold for contact at +10, your sales team could then be flagged about interest of visitor A. They could then reach out immediately with the information from the lead form.

2. Demographic components

Not all visitors to your site are equal. Not only the actions but also the discoverable attributes of the visitor are important to factor into your lead score. Consider two visitors to your marketing site:

• Visitor A, a CEO of a 100+ employee fintech company, lands on your site through a Google search for your company name. They visit “Product”, “Pricing”, and “Customers” pages, read three articles on your blog and enter their information into a lead form for an upcoming webinar.
• Visitor B, a sales rep from a 5-person marketing startup, lands on your site through a Google search for your company name. They visit “Product”, “Pricing”, and “Customers” pages, read three articles on your blog and enter their information into a lead form for an upcoming webinar.

Lead scoring through behavior and demographics

The activity is the same, the person is not. Lead scoring has to over-weight the CEO in this scenario so that this prospect is quickly brought to the attention of the sales team.

This demographic score is the most important part of your lead score as you can assign large weights to your ICP criteria. If your ideal customer profile is a large fintech company, then it's important that no matter what their behavior, your sales team is alerted to interest of a qualified lead. In this scenario you might score large companies +10 and fintech companies + 10 so that these always push you over your threshold.

You can obtain this demographic information either through long and detailed lead forms or third-party enrichment solutions.

3. Sales components

Finally, once the lead has been contacted by your sales team, you should continue to add to and update their lead score.

You can include sales activities such as emails, conversations, and meetings in your scoring model. For instance, every successful contact through phone or email could elicit a score of +5. This makes sure that interest from the prospect continues to equal stronger scoring in your model.

Combining components

When you have your behavioral activity score, your demographic/firmographic fit score, and your continual sales touch score you can add them together into a single lead score for that prospect.

Lead scoring model: Each individual component can go into your lead score, which in turn will automatically tell your sales team whether the lead is worth reaching out to or not.

While this is the ideal way to score, there can be some significant hurdles to implementation without engineering support. Luckily, there are some simple alternatives that can be built within your CRM or Marketing automation system.

Alternative scoring models:

• If you have a well-designed ICP, you can score just on related demographic criteria. This allows you to establish a base model where only attributes that fit those criteria will be positively scored and each lead will need all attributes to reach threshold and go through to the sales team. This can be done directly in Salesforce, Marketo, or most other systems.
• The next option is as described above plus behavioral attributes. Within your marketing automation system, you can combine demographic, firmographic, and behavioral attributes to create a close to perfect scoring system.

Where to start

If you're new to lead scoring, we suggest going for the first (simpler) option and building a demographic / firmographic scoring system. It will produce a score that closely matches your successful customers. It also requires little set up, whereas the other version will require some instrumentation of your site / system to capture behaviors.

Lead qualification isn't particularly difficult, but it is extremely important. When you aren't qualifying leads, you are wasting time and money. But more importantly, if you aren't qualifying leads efficiently, using automated lead scoring or basing your process on your ideal customers, you are also wasting time and money.

If you're using Salesforce, checkout our tutorial on building your own lead scoring system right within SFDC.

Issue Summary

From 7:41 AM to 8:10 AM PT, requests to Clearbit APIs resulted in unauthorized error response messages. The root cause of this outage was the incorrect setup of one of our developer machines that gave direct access to the production database while running automated tests. This briefly removed access to Clearbit’s accounts as automated tests rely on an empty database.

Timeline (all times Pacific Time)

• 7:40 AM: Test run finished
• 7:41 AM: Pagers alerted team
• 7:45 AM: Complete outage after our caching system was invalidated
• 7:48 AM: Accounts database backup started from a snapshot
• 7:55 AM: Root cause discovered
• 8:08 AM: Account backup restored
• 8:10 PM: 100% of traffic back online.

Root Cause

At 7:40 AM PT, one of our developers started a test run with a setup that inadvertently had the development environment configured to use our production cluster. While preparing the test run, our code completely drops all rows from our authentication database.

Resolution and recovery

At 7:41 AM PT, the monitoring systems alerted our engineers who investigated and quickly escalated the issue.

By 7:48 AM, the team started a backup to recover accounts but the root cause was still not clear.

At 7:55 PM, the root cause was found by the developer that caused the issue after noticing how the timeline of events matched the test run.

These problems were addressed and we successfully recovered from a backup at 8:10 PM.

Corrective and Preventative Measures

The following are actions we are taking to address the underlying causes of the issue and to help prevent them from happening again:

1. Improve checks to the testing environment to prevent usage of production settings.
2. Change database recovery process to be more time efficient.
3. Increase backups to an hourly cadence.

We appreciate your patience and again apologize for any inconvenience. We thank you for your business and continued support.

Today those just got a whole lot better:

1. New tracking and detection system that lets us more accurately and widely detect tags. This includes detecting javascript variables, CNAME resolution, and SenderID records. We've already seen this increase our coverage on certain tags by over 30%!

2. 130 new technologies added to our trackers. These include Tealium, Visual Website Optimizer, Outbrain, Taboola, and many others. Full list here.

Great news if you're an API or Salesforce customer, these new tags are already being added to your records. No updates needed on your end.

This technographic profile can be super helpful for both sales targeting, and communication personalization. For example, we dog-food this data internally by sending super targeted onboarding emails to anyone signing up to Clearbit we detect using one of our native integrations, such as Segment:

As always, if you're interested in adding this data to your CRM or internal systems - just give us a shout. You can also request additional technologies here.

And a few have also died along the way. We wanted to understand these trends and see what makes a YC company. To do that, we have analyzed data on over 1,000 of these companies to see what they do, where they are, and how they work. Here's what we found.

The growth of YC & its companies

The program has slowly expanded from small beginnings in 2005 when it took in just 9 companies. The intake has grown by approximately 1/3 each year, taking in more and more companies each year. The only exception was 2013, when the accelerator intentionally took in fewer companies to deal with growth issues that had arisen with the summer 2012 batch.

After that hiccup, they continued to increase batch size all the way through to 2016 when there were almost two hundred companies combined in the winter and summer batches:

Growth of YC batches over time.

90% of these YC-funded companies are still active today, belying the Icarus trope associated with startups. Only 4% seemed to have died entirely, with another 6% acquired by other companies:

Proportion of active, exited, and dead YC companies.

If we look at the OG batch from summer 2005, most have exited. TextPayMe, an SMS-based transaction service, was acquired by Amazon only a year after its YC inclusion for an undisclosed amount. It now directs to Amazon Payments, though the TextPayMe service was deprecated in 2011.

Probably the biggest name from this genesis batch is Reddit. Started that year by Steve Huffman and Alexis Ohanian, it was also acquired a year later in 2006 by Conde Nast. Though it is now independent of Conde Nast, parent company Advance Publications is still the major shareholder.

In terms of funding raised, Loopt seems to have been the big winner in this initial batch, raising almost $40 million in total before it was acquired by Green Dot, a financial services company, in 2012.

This original batch does have a far greater proportion of exits than other batches. This could be that, as the elite of this new movement in 2005, they were hot property and quickly snapped up. Alternatively, this could be the fate of most of the companies, but these have had longer to play out.

The standard deal for YC companies is $120,000 funding for 7% equity. But money for these companies doesn't stop there. Combined, YC-funded companies have raised almost $10 billion in funding. The overwhelming amount of money raised, 97%, has been pumped into still active companies.

Dead companies contribute only about $50 million of that funding. Homejoy, the home-cleaning marketplace raised the majority of that funding—$39.7M, but ceased operations in 2015 as they couldn't find a path to profitability.

There is no obvious pattern when looking at funding through each year:

Funding raised per year.

This is because so many factors influence funding raised:

• Size of batch—there have been more companies available for funding in the latter batches. 2012 and 2013 both had multiple companies that raised over $100M in funding.
• Maturity—companies in the older batches have had more time to raise different rounds
• External factors—economic cycles have an effect on the financial ability to raise rounds
• Unicorns—some companies raise funding far exceeding their YC brethren. It is that last factor that is supremely evident in the graph above, particularly in 2009. The winter batch that year included Airbnb, who have raised almost $2.4B in funding.

Two companies, Airbnb and Dropbox (Summer 2007) have raised over $1B in funding. Another 13 have raised over $100 million:

Amount of funding raised by top YC companies.

There is no real doubt that YC is the premier seed accelerator. The fact that so many tech names have launched through the program shows that the team has a gift for identifying winners. This is backed up by the fact that so many seem to continue to be active—supposedly, 90% of startups fail, yet 90% of YC startups succeed.

This is no mean feat in an industry that is moving at such speed. YC has been around for more than a decade. 26 companies had already been through the program before the iPhone was invented. The fact that the accelerator continues to take on more companies, and more diverse companies, shows promise that it will still be relevant a decade from now.

The types of companies going through YC

Beyond the headline numbers, we wanted to look at what type of companies are being funded by Y Combinator. If you are at a startup that is thinking of applying, or just interested to know what is getting the top funding firms excited in tech, then knowing what type of companies are common within accelerators can give you a better understanding of the startup ecosystem.

To do that we broke down companies according to their Clearbits tags. These tags are automatically generated based on website content, and allow for high-level categorization of companies. Companies can have more than one tag. For instance, a company could be tagged as both SaaS and B2B, or B2C and E-Commerce.

Since 2005, the number of B2B companies and the number of B2C companies have been fairly similar. About 950 companies are categorized as either B2B or B2C, with approximately 55% of these companies being B2B:

Proportion of B2B and B2C type companies in YC.

However, it wasn't always like this. The early batches were skewed more heavily towards B2C companies. It is only since 2011 that B2B companies have been in the majority, though 2016 saw a shift back to B2C dominance:

Change in proportion of B2B/B2C companies at YC over time.

When we break out the proportions of other type of companies, we see that SaaS is unsurprisingly the predominant type of company at YC:

Proportion of types of companies in YC.

However, there are a substantial amount of marketplace companies coming through YC, over 200 in this dataset. E-commerce is the main category with the fewest companies coming through the accelerator, though the proportion per year seems to have remained about steady in each batch:

Change in proportion of company types at YC over time.

If anything, though SaaS has striking dominance, it is on the wane, with proportional numbers dropping each year. Conversely, marketplaces seem to be taking more and more of the share with each passing year.

Going deeper into the type of companies in YC, we also looked at the sub-industry tags for these companies. Perhaps unsurprisingly, tech hardware and software account for just over 50% of all YC companies.

Proportion of industries in YC batches.

The 'Other' in that graph contains 67 other tags including advertising, human resources, financial services, payments, and aerospace.

This won't always be the case. Whereas early batches had more generally categorized companies, as tech pushes out into other fields, Y Combinator is also stretching out to help these types of company. As an example, here are the numbers of fintech and biomedical companies in different YC batches:

Increase in fintech and biomedical companies in recent batches.

Though fintech has been around since the very beginning (that one 2005 company is TextPayMe) it has grown substantially in the last few years, accounting for 15 companies in the 2015 batches.

The rise of biomedical companies has been even more impressive. There were 0 biomedical YC companies before from 2005 to 2013. There have been 30 in the three years since, with a massive 19 in the 2015 intakes. The winter 2015 batch contained 13 biomedical companies ranging from disease testing - Diassess, to affordable medical devices - Shift Labs, to Laboratories-as-a-Service - Transcriptic, to medical cannabis - Meadow.

This shows that even if you aren't straight “internet software,” YC could be the place for you. As tech expands out into other arenas, those other fields become interesting for VC firms and accelerators. If you are in finance, medicine, automotive, or aerospace, this could be the right time to be looking into applying.

And if you are straight SaaS, it's always a great time to apply.

San Francisco is the YC hub

Y Combinator is a 3-month residential program. The accelerator invites the founding team of the company to their HQ in Mountain View, California, where they can meet fellow founders, talk to experts, and participate in talks from Silicon Valley luminaries.

The benefits of asking everyone to a single place are the networking and access to a huge network of early adopters/customers. However, it does mean that the accelerator is dominated by companies that are easily able to base themselves in the Bay Area for 3 months. YC is heavily dominated by US-based companies, with about 86% of YC companies based within the United States:

Proportion of YC companies from different countries.

The accelerator does take applications from any country though. If successful, they will put international founders in touch with people who can help them get the required visas. However, some international teams might still be put off by the overhead required to pack up and head to another country for three months.

Outside the US, the next main contributor of companies is Canada, with about 3% of YC teams. Next up are the UK and India.

Within the United States, California predominates:

Proportion of YC companies from different states.

75% of American YC companies are California-based. The other significant home states for YC companies are New York, Massachusetts, and Washington. Within these states, New York City, Boston, and Seattle are the respective hubs. Unfortunately some states, particularly it seems in the Midwest, don't seem to have much representation at Y Combinator.

We've previously shown that San Francisco is still, by far, the main startup hub in the world. It is no surprise then that, within California, SF dominates in YC companies as well:

Proportion of YC companies from different cities in California.

Don't be fooled by that hotspot in LA, all the action is around the Bay Area:

Proportion of YC companies in the bay area.

That archipelago accounts for about 90% of the California-based YC companies.

This isn't some location-based form of nepotism. Rather the sheer totality of tech-based companies in Silicon Valley (about one per one thousand people) means that there are just far more available, and far more applying. In fact, this can explain why over a quarter of all YC companies are based within the San Francisco city limits.

Another reason for the high number in this locale could be post-YC relocation. This data is from the current locations of all YC companies. It could be that they all started in Kansas, but stayed in San Francisco!

The YC ouroboros

Any good tech company dog foods their product. It seems that YC dogfooding is common as well. Perhaps the ultimate affirmation of YC companies is that other YC companies use them.

Using our tech tag, we looked at the stacks underpinning each of the YC companies in our dataset. 11 YC companies are commonly part of the tech stack of other YC companies, with Mixpanel (s2009) being the most common:

YC tech used by other YC companies.

43% of YC companies are using Mixpanel for their analytics. That is pretty impressive for the analytics provider. Remember, not all YC companies are even tech companies, so might not need web/mobile analytics. Mixpanel seems to be the YC SaaS that is most useful across board.

In 2nd place was A/B testing platform Optimizely (w2010), used by about 22% of their fellow YC companies. These other companies were all used by over 10% of YC companies:

• Hosting service Heroku (w2008)
• Data platform Segment (s2011)
• Email API service Mailgun (w2011)
• Live chat tool Olark (w2009)
• Payment system Stripe (s2009)

From this, it seems that 2009 and 2011 were great years for YC SaaS.

Overall though, Google Analytics was the tech most often seen in YC stacks:

Tech used by YC companies.

This is unsurprising as it turns up in the stack of almost every company. Google Apps also has over 50% usage within the dataset. But it seems to be Mixpanel that is the hit with their fellow Y Combinators, being the 3rd most used piece of software by the group overall.

Three YC companies—Mixpanel, Optimizely, and Heroku are in the top 10 for YC tech stacks, with Segment, Mailgun, and Olark within the top 20. Of course, this doesn't include companies that are using Zenefits for backend HR functions, Dropbox for storage needs, or DoorDash when they are getting hungry!

Understanding Y Combinator

With over 1,400 companies going through the YC program it's honestly quite difficult to draw conclusions about what makes a great YC company. The fact that so many continue to succeed is a testament to the program and founders.

Not everyone can raise the type of money Airbnb or Dropbox saw, but a massive majority of these companies are not only succeeding, but thriving. They might not be huge names, but they form the backbone of the tech ecosystem. Developers now couldn't imagine life without Docker, Heroku, or Segment. Marketers couldn't survive without Mailgun or Olark. And anyone founder that hates paperwork thanks the stars for Zenefits and Gusto!

Probably the only concrete conclusion could be that the program will continue to diversify, while most likely being software dominated. The YC companies of five years hence will likely still be primarily SaaS and tech, but they will be pushing those technologies into completely new industries and verticals. This makes for an exciting future for the program, and for startups overall.

As the definitions of ABM are ambiguous and quite varied, let’s agree to define Account Based Marketing as this: a simple approach that aligns sales and marketing teams on the same pre-defined target accounts.

In practice this typically means:

• Sales and marketing teams agree to prioritize a list of target accounts.
• Marketing’s Key Performance Indicators (KPIs) become more directly tied to down funnel sales goals.
• Marketing provides Sales Development Representatives (SDRs) with lists of contacts to cold call and manually email.
• Marketing organizes targeted events and invests in enterprise level SaaS tools to personalize messaging to this top handful of accounts.

But what about the accounts outside of your target list? Think about that huge government deal from last quarter that nobody was expecting.

Segment, Slack, and Atlassian are all landing impressive enterprise logos with Account Based Marketing. However, they take traditional ABM a step further by using data and automation to provide a targeted and personalized experience for all of their prospects, and not just those on their "target list." These fast-growing companies personalize website content, marketing outreach, and outbound sales emails when they’re selling to IBM, but also when they’re working with Joe’s Muffler Supply.

How?

The most innovative companies have realized that personalization and automation aren't mutually exclusive (see excerpt from a recent Segment Marketing Whitepaper). With the right tools, data, and organizational support, Marketing can support Sales across your company’s entire universe of potential accounts (without hiring 100s of new sales reps!).

Let’s call it ABM at Scale? Or Programmatic ABM? Or maybe just data-driven marketing.

Here are a few ways Clearbit’s customers are using this scaled-ABM approach to drive sales:

1. Automating lead generation: Identifying anonymous website visitors and automatically sending them personalized outbound emails without any human touch. Learn More →
2. Website chat only for qualified leads: Surfacing a proactive chat (Intercom / Drift) only when someone from one of their target verticals at a company with over 50 employees visits the website. Learn More →
3. Personalized email campaigns: Enriching Salesforce or their marketing automation system with highly detailed company firmographics to run analysis and create personalized email campaigns at scale. Learn More →
4. Website personalization: Dynamically altering website content based on an anonymous visitor’s industry, company size, or whether the company uses a technology like Marketo.

What would it mean to your business if you could differentiate a highly qualified website visitor from a lower value (SMB) visitor, but provide both with a satisfying and personalized buying experience?

Reach out to sales@clearbit.com or create a free trial if you’d like to learn more.