How I Created A B2B Customer Lead Prioritization Model

How I created a Customer Lead Prioritization Model to ensure that we focus on those customers who are critical to growing the business.

Sourcing and prioritizing leads is important to ensure resources are used efficiently and are directed towards a common goal. Focusing on a common goal also helps to track the progress and stops us from taking a spray-gun approach.

Recently, I created a lead prioritization model as a common ground for our product and sales teams to expand our business. Below are the steps I followed:

Understanding The Organization’s Strategy

Even before I started, I took some time to understand what’s the priority for the executive team. Where they think the organization is headed, what’s the five-year plan, and which markets to focus on.

Asking the right questions to the right people is important here. It helped me understand who “we think” our competitors are, and in which markets our focus should be. The five-year plan included aggressive international expansion. These discussions helped me understand a few things:

• International expansion is a priority (so international leads are valued more).
• Certain industries are more important because we have been less successful in those, and those industries are expected to grow tremendously in the next five years.
• Since our company deals with security and communication products, it also made me understand that we need to round-up the use cases keeping the target industries in mind so that we can pitch a complete product to the customers. This might include a “build or buy” decision.
• In case we are unsuccessful in selling our products to the industries in question, I need to find an alternate route to reach them.

Sourcing The Leads

With the base understanding, I started sourcing the leads. There were mainly three ways we capture the leads:

• Inbound leads that visit our website, proactively check out our products and either contacts Sales or drops off. They are mostly product managers, engineering leads, and managers of other departments.
• Inbound Leads that visit our website and sign up for a free trial of our products. They are mostly individual developers or developers at smaller companies/startups who are looking for a cheaper way to validate their product-market fit. This is by far the easiest of all and the most difficult to convert (the website needs to be near-perfect to score high SEO points and source organic leads. I have found that our website traffic nearly dropped by a factor of three when paid advertising was called off).
• Outbound Leads that are generated through in-house market research and Sales reach outs.

I wanted to aggregate the list of those who visited our website and churned away, and those who we thought qualifies as a lead through in-house market research.

Understanding Market Forces And Refining The List

Once I had the list, it was important that I prioritize among the existing leads. In a raw, unfiltered list, there will be anywhere between 3000 and 5000 leads acquired from multiple channels. Targeting all of them is a waste of time and money. We needed to focus on a subset of the list, let’s say 50 or 100.

I did this through external market research. The list had companies grouped by industries and sub-industries (such as Gaming — MMO). The list also had the location of the companies, their monthly web traffic, app downloads (if they had one), and others that made up to around 100 data points. Definitely, there was a lot of noise that needed to be cut through.

This is where understanding the overall company strategy helped. I knew we had to expand internationally. There was also a potential profit target. So I tried to figure out the industries that were riding the growth wave, the countries/regions they were expanding to, the overall growth curve of the industry, etc.

For example, I figured that after the US, e-commerce is expected to be a major phenomenon in Latin America (LATAM) and Asia. The predicted annual growth curve in LATAM and Asia were the highest among any other regions.

If I needed to sell communications product, I need to know where we fit into the overall value chain.

We can approach the Amazons and Walmarts of the world directly, but would it help? What’s the competition we’re going to face? Would that shrink the profit margin per transaction? Instead, I mapped out an e-commerce buyer’s journey and found that there were multiple entry points without going to the giants directly.

Take Shopify’s example. A lot of companies are running on Shopify, and Shopify’s business model is very modular. They aggregate every part of the shopping experience and provide it as a bundle to its customers. So, in an e-commerce buyer’s journey, a customer would expect some sort of communication (mostly SMS) when they purchase a product, sellers might also want to remind customers about an abandoned cart, etc.

So one entry point would be to go to those companies who aggregate communications for those medium and small scale businesses selling on Shopify and similar platforms.

The other option could be to partner with electronic point-of-sale software manufacturers (such as Square) who have modularized in-store purchasing.

While ranking companies, I tried to understand the global companies that are either based on any of these three regions (US, LATAM, Asia) and has operations in at least the other two. In fact, it’s best if they’re just starting in the other regions. High chances that they’ll be not big enough yet to have already received product pitches from other competitors. If we can nail the customer service, we can increase the customer lifetime value.

Understanding The Value Of The Data Source

Another major problem with having too many data points is that very few of those data points come from authoritative sources. Data obtained from credible sources (such as research databases or websites like Alexa) are trustworthy but we don’t always find everything that we’re looking for in those data sources.

This is when we start turning to not-so-credible data sources. However, there are high chances of creating a very wrong model if we factor in too much non-authoritative data and assign them weights comparable to authoritative data.

I wanted to strike the right balance of having enough data to build a comprehensive model, but also not relying too much on non-credible data sources (such as # of app downloads, web traffic on websites of companies not indexed by Alexa, etc). Hence, initially, I wanted the model MVP to be mostly built on credible data sources with limited non-credible but necessary sources.

For example, the # of app downloads were not from a credible source but to prove the hypothesis that higher the app download, more customers use the product and higher the traffic on our products if we partner with them we need to weight in that data.

In the future, once I have enough evidence that the model works, I can add the non-credible data one by one and test the incremental effectiveness of the model.

Using Experience And Intuition To Create The Ranking

This is mostly an extension of understanding the market forces and leveraging business judgment.

For example, even though I see a certain lead doesn’t have too many app downloads but I see the company is backed by famous VCs like Sequoia or Softbank, I predict that the lead (often a Series B or C startup) has strong potential. With other factors constant, I try to prioritize them.

In the end, I assigned weights to each data point based on our mission. Since the international expansion is the primary objective, leads with operations beyond US score high. Leads with operations in LATAM and Asia is prioritized, startups with high potential get bonus points.

All of these taken together, I had the minimum viable prioritization model that needs to be tested for success.

It’s important to note that models like these are never final. They should be re-evaluated once in a quarter or once in six months to re-adjust the weights on data points, add more data points, and even looking through the critical lenses to understand if it’s still serving the purpose. However, having one such model for every major task helps reduce the amount of time spent on redundant work.

These models need not be automated through machine learning (although if that’s the case, there’s nothing better). However, very few companies have such bandwidth. In such scenarios, having one on an Excel sheet works just fine.

Image Credits: Google Images