Identifying new clients is one of the key tasks in achieving successful organic growth for companies selling B2B. Due to the usually long sales cycle, each sale costs more which puts pressure on a successful sales organization as well as giving good service to get a good customer retention rate. Hence an important part for B2B companies is to decide which customers to focus on, to avoid wasting resources on developing relations that ultimately will not create opportunities for sales. This is often carried out in a non-systematic way through human estimations in the sales teams. But what if there is another way, a bit more data-driven? Wouldn’t it be nice to get some indicators from publicly or 3rd party available data in order to prioritize your efforts? I see you are nodding; well, predictive analytics might be just right for you then!

What is Predictive Analytics?

Predictive Analytics is all about exploiting patterns found in historical, transactional or any other 3rd party data to identify risks and opportunities. Models capture relationships among many factors to allow assessment of risk or potential associated with a particular set of conditions, guiding decision-making for candidate transactions. Of course, in order these insights to be  accurate you need a lot of data and preferably data that you can trust. So if you don’t have enough data or the data is inaccurate with duplicates and anomalies, then you are likely to get inaccuracies in your predictions (garbage-in, garbage out rule)

How to get started

• First things first; start by measuring the effect of doing nothing. In other words, what would happen if you didn’t advertise at all at a specific channel or if you didn’t set a data-driven prospect prioritization process? Thus, measuring the effect of doing nothing allows you to create a baseline to measure against & predict the return-on-investment of any use case that you will think of.Although this might sound trivial, most of the companies omit this step since they think that it is not necessary or it takes too much of their time.

• Once you have a full overview of where you are standing, defining control groups is the next step. A control group is a vital part of your testing process since it helps validate testing results and proving the return-on-investment of a specific use case.Put simply,  a control group is a test cell of customers or prospects who receive no special treatment. In that way, you will be able to understand if you really moved the needle in the right direction or it was just a coincidence, that would have happened anyway.

• It is now time to define your predictive analytics use cases.  A use case let’s you show people what to expect and the outcomes that you will achieve. A use case is a very powerful and reusable piece of content, since people tend to digest the information and see your point much easier than reading paragraphs of text. No idea where to start from? A few use cases that I will describe in detail in my future posts (both from a business and technical perspective) that it is worth considering of, are the following:

  • New Customer Acquisition

  • Churn Prediction

  • Customer Activation

  • Cross/up selling

• Last but not least, choosing your KPIs (reflecting strategic goals) and metrics (reflecting tactical goals) will ensure that you get your company’s credibility and show how your specific use case has contributed to the overall growth.Without measuring the right KPIs/metrics against specific goals, you cannot build a rapport with the decision makers in your organization since different parts of the organization care about different issues.

As you probably have noticed so far, I have only concentrated on the business side of this project, omitting intentionally any software or product recommendations. Those four steps described above are crucial to be in place (defined and agreed), no matter how your current Marketing Stack looks like today.

Bringing it all together, B2B Marketing & Sales teams should start dipping their toes into predictive analytics to derive intelligence out of customer interactions, adopting a customer-centric approach. In order to succeed on that journey, a mind-switch should be made by both teams, realizing that they are both responsible for bringing in new leads, as well as retaining and up-selling existing customers. In my next (a-bit-more-technical) articles, I will describe the MarTech architecture needed in order to put your first use cases in production. Stay tuned!