Sales Excellence: The Top Funnel – Universe To MQL

By Stu Schmidt, president of Zend, a Sapphire Ventures portfolio company

This is the second blog in a series authored by Stu Schmidt, in which he looks at how business leaders can scale their sales organizations to new heights via increased control, visibility and emerging new tools that can fundamentally transform the selling process. Read the first blog in this series here, and look for subsequent blogs coming soon.

In my previous post, How to Diagnose a Broken Revenue Process, I introduced the concept of defining three unique “funnels” and the need to define processes for each that are tightly integrated and fanatically measured. Let’s now take a closer look at the very familiar “top funnel”, the one focused on the marketing qualified lead (MQL). Here, we are trying to identify, in a massive universe of potential suspects, the people that are actually interested in what we have to offer.

First of all, the universe is big…very big. And very noisy. The proliferation of top-of-funnel tools, led by marketing automation and various online advertising tools, fill our inboxes and overwhelm our eyeballs daily. Combine this with the undeniable fact that every decision maker has more potential things in which to invest than they have time or money to invest, and it’s surprising that we get any hand-raisers at all.

So, what to do?

Define your lead flow model
Even though the marketing automation market is fairly mature, it is surprising how many companies have not done the basics in their implementation of these tools. I’ve spoken to customer success leaders from most of the major vendors in the marketing automation space, and they all concur that the “adoption” of their products is generally weak resulting in far less return than the customer originally envisioned.

The place to start addressing this issue is at the definition of your lead flow model. Here is an example of a typical lead flow model.

Lead Flow Model

Consider the following ways you can refine and get much more out of each step in the model:

  • Each stage or “lead state” must be clearly defined, and all the parties involved must agree to that definition. For example, your sales development reps need to know exactly what constitutes an MQL and what the inside reps will expect of them in producing a sales-accepted lead.
  • What is happening before “Raw”? Are your demand generation methods working effectively? What are you measuring there? How often? Who gets involved in the measurement meetings? The top performers in this discipline involve every stakeholder group in weekly metric review meetings. Sometimes your most important insights will come from very unexpected players; for example, finance, direct sales, product management and the sales development reps can provide insights from a non-marketing point of view.
  • Define the optimal path – the one shown along the green band across the middle of the diagram above.
  • Ensure that every possibility is accounted for. What else can happen other than the optimal path? Think about anything you want to be able to measure and ensure you have a clear stage defined for that possible outcome.
  • Define and constantly iterate your scoring methodology. Is it really working? At my current company, we realized that certain key attributes of a lead were not being considered in the scoring model. That realization came from the sales team, not marketing. Communicate openly and look at it more like a game than blame.

Go beyond behavioral scoring
Most scoring methodologies today are based on the actions taken by prospects on websites or responses to emails. A click, a download, a free trial –these are all actions taken and most commonly used in determining MQL status. Big data has changed the way we can evaluate a lead. There are literally thousands of indicators or signals that can be ascertained about someone (or their company) regardless of what action or behavior they take. Unfortunately, there are just too many of these indicators for a human being to evaluate, especially in a big, noisy universe.

Amid this massive proliferation of data, two very important trends are emerging in the “find me an MQL” space, which can help you augment the behavioral scoring you are already doing:

Trend One: Predictive analytics powered by big data and combined with your customer data. Category specialists such as Infer, SalesPredict, Mintigo and others use a sophisticated approach to create a “model” of this myriad of indicators based on your past successes. The model then evaluates anything you feed it – for example, an inbound inquiry – and predicts the likelihood of success. It “scores” your leads in a new way. My experience has shown that by combining behavioral scoring with predictive analytics you can dramatically increase the conversion ratios of your complete pipeline…all three funnels.

Trend Two: Trigger events. Good sales people have known for a long time that a specific external event or signal can indicate a likely propensity to buy. This approach is making its way into the marketing tool space and holds significant promise to proactively identify a good prospect before they even come to your website.

An example of a trigger event implementation is the “Watch Lists” in InsideView, which allow users to identify companies and get instantaneously notified when there’s news related to that company. In another example, at a previous company we built web-crawling technology to identify companies that were hiring insides sales reps or sales development reps. We knew that hiring for these positions was an indicator of a specific motive that directly linked to the value that we offered. We would call and more often than not get an audience because there was an active motive with urgency to solve a specific problem with which we could help.

Today, more advanced methods that include social media activity tracking, technology signals on websites, job board postings and literally hundreds of other signals are being used by many predictive marketing solutions. Thanks to big data and advanced machine learning we can now be alerted of potential targets that we would have never found with traditional methods.

What external signals or trigger events exist for your solution?

Next time we’ll tackle that nasty, overwhelming problem of taking your bright shiny new MQL and turning it into an actual (I won’t say qualified…you’ll see why later) opportunity in the sales funnel.