Don't Trash Your Leads

Lead generation is one of the most draining costs in the marketing and sales funnel, where expenses can reach hundreds of USD per lead. According to many sales leaders, the emotional connection between the prospect and the salesperson plays one of the most pivotal roles in the buying decision. Research has shown that people have different communication styles, and those with similar communication styles will establish better emotional connections, which are critical determinant of the success of the sales interaction.

However, saying that the conventional wisdom of assigning sales representatives (SDRs) and account executives (AEs) to leads, i.e. lead routing, is inefficient, is a gross understatement. More precisely, in over 95% of cases, it is random, round-robin, or based on gut feeling. The current methods do not take the prospect’s and salesperson’s characteristics into account.

Handle Your Leads With a Data-Driven Approach

In order to maximize revenue and make sure each lead is handled by the person with the highest probability of success, we must first understand the factors (AKA features) that influence the emotional compatibility of the sales team and prospects.

There is no right or wrong here; the features may differ from one business to another and can depend on the product or service, market and vertical, culture, and business processes of each business. This is why a data-driven approach is a must in order to eliminate all biases.

The Process

Step 1 — Know Your Data

DB Schema
Typical CRM Database Schema

Just like in the old saying, “you don’t know what you don’t know”, the first step is to encompass as much data as possible. We never know in advance which data point is more valuable. Most of your sales data already sits within the CRM, but some of it is unrelated to specific prospects or salespeople, while other data may be in other systems, such as marketing channels (e.g., Google and Facebook Ads). This is why gathering ALL available data and linking it to the prospect and salesperson is key. The more data you have available, the more accurate your decision is going to be.

Step 2 — Top 5 Features of Emotional Compatibility in Your Business

In order to extract the most unbiased features on the compatibility of prospects with salespeople, we must not presume or assume anything. Deep learning is the best tool for this kind of job. We need to run our deep learning algorithm to seek the contributing factors for a prospect to convert into a sale, as well as the contributing factors in the opposite case when the sales effort fails. Once this is done, we should receive the contributing factors that led to a positive or negative result for each lead. For instance, in the case of a successful sale, we should see something like this (names of features and values have been redacted):

A schema which shows the features that contributed a ML-based prediction
Positive contributing features are in red and negative contributing features are in blue

From this graph, we can see that “path” being “NaN” (this is a made up value for privacy reasons) is the most important feature, followed by spoken languages, referrer link, and so on. In addition, the number and values of positive contributing factors outweigh the number and values of negative contributing features.

In the case of an unsuccessful sale, we should see something like this (names of features and values have been redacted):

Positive contributing features are in red and negative contributing features are in blue

From this graph, we can see that the lack of spoken language data and marital status being “single” are very important negative features, outweighing the positive features.

These features are unique to each prospect-salesperson pair. The next steps are to extract these positive and negative features for each possible prospect-salesperson pair (factorial analysis), aggregate them, and weigh them properly in order to understand the overall positive and negative feature set and values.

Step 3 — What If?

In the previous step, we extracted both the positive and negative features and values that affect each salesperson’s compatibility with a prospect. The question remains, however, did I make the right decision handing that prospect to this salesperson? Perhaps another salesperson could have done better. Alternatively, taking it one step further, perhaps I “wasted” my best “closer” on a prospect my juniors could have converted.

Probability for each Account Executive to convert each prospect (brighter color means higher chances to convert)

In order to tackle these “what if?” questions, we need to utilize a field in mathematics called “mixed integer programming”. First, we cluster the prospects using a multi-dimensional clustering algorithm based on the positive and negative features we extracted. Then, we calculate the probability of each salesperson to convert each prospect according to their individual feature values versus the prospect’s, which were calculated in the previous step. From here, we can generate a heatmap that shows us which salesperson has the highest compatibility with each prospect. This heatmap represents the optimal lead distribution among your sales team.

Step 4 — Actionable Insights and Dynamic Quota Management

So far, we’ve generated a heatmap showing the compatibility of each salesperson with each prospect, enabling us to make data-based decisions accordingly. For example, we can spot “easy” leads, characterized by a high probability of conversion: They appear as bright horizontal lines and can be handed off to our junior team members. We can also spot the leads that are only compatible with a few salespeople. However, if we start to route leads according to the map, couldn’t it create a situation where the leads only flow to the “closers”, starving the others? The answer to that is a very definite NO, simply because of time. Not all leads arrive on your doorstep simultaneously; they flow continuously, and routing decisions are made according to the current prospect map. This means that when a new prospect arrives, the best decision, which will maximize the overall revenue of the business, can be to reassign an already assigned (but not handled) lead to another salesperson who is more likely to convert it. This is dynamic quota management. The “difficulty” of the conversion, juxtaposed with the overall likelihood of conversion across the entire sales team, can be taken into account when assigning quotas and managing salesperson compensation.

Conclusion

Lead generation is expensive, and emotional compatibility with the client is the most determining factor in sales. Using a data-driven approach and data-science tools, a sales leader can assign each prospect to the salesperson most likely to convert the prospect into a sale. Using emotional compatibility metrics also enables sales leaders to continuously optimize lead assignments and create better quotes and compensation. This approach maximizes the overall performance of the entire sales team for three reasons:

  1. Salespeople are assigned more compatible prospects, which will be more easily converted into sales.
  2. Understanding the emotional compatibility matrix of the entire team provides the marketing department with highly actionable information, enabling them to focus their campaigns on generating prospects that are more compatible to begin with.
  3. The two reasons above create a positive feedback loop, which improves conversion performance even more, and your entire team is “in the zone”.

We at Buzzer have developed the tools and techniques required to perform these four steps, and we’ve successfully improved the conversion rates of sales teams by 10%-20%. We’d be happy to show you how you can implement our tools in your environment to boost your sales team’s performance and generate more revenue with zero additional resources.

For a Proof of Concept, contact us today!

How does outperforming quotas sound?

See for yourself how Buzzer AI is the unfair sales advantage you’ve been looking for.