How to Pick the Right Accounts for Your ABM Program

Account-based marketing (ABM) is fast becoming the preferred approach to marketing for many B2B companies. Last summer, in a survey of B2B business and marketing leaders by Demand Metric, 45% of respondents said they were testing or already using ABM, and another 26% said they were interested in adopting it.

The defining characteristic of account-based marketing is that it focuses marketing efforts on a specified group of target accounts. Therefore, choosing which accounts to target is an essential step in implementing ABM, and most ABM thought leaders and practitioners agree that account selection is the most critical component of any ABM program. Choosing the right target accounts is not the only thing you need for success, but it will be impossible to build a successful ABM program if you target the wrong accounts.

ABM can be used for acquiring new customers and for marketing to existing customers. In this post, I'm focusing on account selection for new customer acquisition.

Most companies choose target accounts based on how closely those organizations resemble their existing customers. This approach is often called look-alike modeling, and the process is fairly straight forward:

• Companies identify the attributes and behaviors that their best existing customers have in common. For example, firmographic attributes might include company size, industry vertical, number of employees, and location.
• Then, they use these shared attributes and behaviors to create a profile of their "ideal customer."
• Lastly, they will choose their target accounts based on how closely they match the ideal customer profile.

Most companies select target accounts manually, but mature ABM practitioners are increasingly using predictive analytics to support the account selection process. Virtually all predictive analytics solutions use a sophisticated version of look-alike modeling to identify target accounts. They extract data regarding existing customers from your CRM and marketing automation solution and combine that information with external data about those customers to construct a customer data model. The solution provider then runs your customer data model against its database of businesses and/or applies the model to prospects already in your marketing database to identify the accounts that resemble your existing customers.

The advantage of predictive analytics is that it can incorporate and process far more data points than humans can realistically use. Therefore, predictive analytics solutions can enable companies to build and use more comprehensive customer data models and thus do a better job of identifying accounts that most closely resemble their existing customers.

In many ways, look-alike modeling is the essential foundation for effective account-based marketing. But like any business tool or methodology - no matter how powerful it may be - look-alike modeling is not without limitations. Therefore, it's important for marketers to understand what look-alike modeling is really trying to achieve, what assumptions or inferences underlie it, what the limitations of look-alike modeling are, and how to tweak the look-alike modeling process to reduce or eliminate some of those limitations. I'll discuss these issues in my next post.

Illustration courtesy of Emillo Kuffer via Flickr CC.