The Harmony Algorithm: Solving for the Next Best Lead

Unlocking New Levels of Scoring Optimization to Directly Impact the Bottom Line

Prioritization.  By definition, it’s the act of organizing or dealing with something according to its level of importance. If we could aggregate the most commonly used words in the workplace and translate that into a word cloud, I would venture a strong and educated guess that it would be one the top results amongst the populous.  What’s not to love? It alludes to logical and objective reasoning to deliver proper direction.  A dream come true.  However, when you boil it down, the cause of why we employ that word comes into question.  

The need for prioritization is a direct consequence of an overwhelming environment.  An environment where the lines of what, where, when, and how to execute are blurred and indiscernible.  In the context of CRM utilization, this is an epidemic amongst most organizations.

A CRM being enacted upon, like Salesforce, is meant to give a rep a systematic approach to take action against a lead, contact, or account at opportune periods of time. The utopia for any Sales or Marketing leader is to have the best leads in a reps hands at the right time.  A simple ask, but a complicated solution.  Antiquated systems, processes, unmanageable data structures, and ideologies plagued our ability to truly achieve this.  No matter our respective industry, total addressable market, or business dynamics the issue at hand has proven to result in disadvantages in productivity which has lead to suboptimal unit economics.

How we solve for this? In an effort to crack the code, I went through an exercise to define the combination of variables that cause a lead to be ‘perfect’.  I concluded that there are three key variables in determining this:

  1. Brand Engagement
  2. Account Fit
  3. Contactability

Let’s break down each of these variables and depict the underlying technologies needed to support them.

Brand Engagement: To what degree of significance has a lead engaged with our brand?  Answering that question has typically been determined by a lead scoring mechanism that lives and breathes in a marketing automation tool like Marketo, Pardot, Silverpop, or Eloqua.  

The levels of scoring are determined by specific types of behaviors and data points that can be attributed to an individual’s journey e.g. site visits, form submissions, email opens, click paths, campaign engagement, persona + firmographic data, frequency, and so on.  Historically this score has been the sole way to prioritize leads in CRM for sales reps.  While this is the best place to start, and an industry standard to have in a digital world, this metric is not strong enough to stand on it’s own.

Account Fit: Does this lead belong to a company whose DNA closely resembles our critical mass of successful customers?  It’s one thing to have a lead that is really engaged with the brand, creating hand raising events and demonstrating interest, but it’s useless if they do not fit your understood TAM.  

That said, the first order of operation is to define what’s your ideal customer.  Arguably that answer could be hacked by just appending firmographic data upon conversion of a lead and baking that into the lead scoring model.  However, that’s a static analysis that does not lend itself to dynamically evolve over time as new types of deals and companies begin to show demand for your offering.  

Tools like Everstring enable you to integrate a machine learning mechanism that is constantly looking for the positive and negative events that occur in CRM and then analyzes that cohort to parse out the commonalities.  Those commonalities can be based on their firmographic make-up, technological and departmental utilization, volatility of growth, and a whole host of other data points to determine what is the DNA of your perfect customer.  Once known, the tool will cross-reference that lead with that perfect customer archetype and render a score.  Powerful stuff.

Contactability: What’s the probability that we can have a meaningful conversation via proactive human outreach in a given period of time?  If we follow a path to optimal conversion, we note that a conversation with the lead has happened almost immediately after brand engagement. 

At the core of that ideal scenario is the belief that the more conversations we have, the greater the likelihood is that we can progress a lead to the next step of the buyer’s journey. That said, how can we manufacture more conversations? As part of their platform, insidesales.com provides two features that enable you to increase the probability of having conversations:

  • Local Presence – Specifically, it’s the ability to appear to a lead that you are calling from a number that matches their area code aka a local number.  Studies show that a person is significantly more likely to pick up the phone coming from a local number as opposed to a number they do not recognize.
  • Neurolytics – Weather, traffic, news and infinite data points are cross-referenced with the leads location to determine the likelihood of that person picking up.  For example, let’s say the lead is located in Dallas, Texas and there is a Tornado Warning in effect.  The chances of that person picking up in the middle of a storm are way less likely than a lead who is not. This scoring model is dynamic and adjusts in real time to account for these types of changes and ranks leads based on that probability.

Now, the fun part.

With the means to place a numerical scale/score for each of the three variables previously mentioned, the harmony algorithm combines the disparate values into one weighted average score to determine the best leads as seen below:

Screen Shot 2017-04-12 at 11.38.28 PMThis algorithm weighs out the importance of each variable enabling you to determine which should have a bigger seat at the table.  Should the account fit score outweigh the behavioral score? That determination can be resolved by performing a regression analysis on historical data or by creating a hypothesis and testing from there.  The moral of the story is that every organization is going to be different and you should allow the data to drive your decision-making.

After replacing the previous method of prioritization with this metric, the results we experienced were nearly instantaneous. An average 293% lift in almost every meaningful KPI and unit economic we had as a marketing and sales org.  So only one question remains, why not you?

Want to learn more? Check out the podcast discussing the topic here!

Leave a Reply

%d bloggers like this: