Practical RevOps Analytics Series: Lead Source Analysis – Don’t Get Tricked by Dirty Data

By Rob Kall, CEO & Co-Founder, Cien.ai

“If you have one lead source that converts at 5% and another one that converts at 15%, the difference in success rate is 300%.”

– Rob Kall, CEO & Co-Founder, Cien.ai

Lead Source Analysis – Propensity to Convert?

Prospecting for a New Logo business is hard for SDRs & AEs. One of the most common complaints I hear from sales teams is that “the leads are bad….”. When generating leads for your sales team to prospect into, it makes sense to understand which lead sources are working well.

The formula is super simple:

Propensity to Convert per Lead Source (%) = Successful Outcomes / # of Leads

Should be easy, but there are a lot of pitfalls…

Data Quality Considerations – Your Grouping Field

First, you need a set of standardized lead sources. Most companies let marketing define lead sources, and, over time, leads are added from a myriad of sources with very specific names. It is common to see over 200 unique lead sources in a single CRM instance. You may see things like “DreamForce 2019” and “Dreamforce 2020”. Those are both trade show events and should be classified as such. To address the issue, use a tool like Excel and:

1. Create a unique master list of your current lead sources (Data -> Data Tools -> Remove Duplicates).
2. Map them into new standardized groups. Here is a list of 20 suggested names to use: 

3. Use a function like VLOOKUP to update all leads/accounts with the cleaned-up lead source. Now you will get statistically significant results from a shorter master list.

You may find quite a lot of “non-determinable” records, which indicates that you have issues with data completeness. By looking at the “added by” field, you can determine if they came from marketing sources or were added by the reps themselves (if they were rep-entered, update the lead source accordingly).

Data Quality Considerations – Your Grouping Field

Now let’s turn to the actual list of leads. If you analyze data over a few years, many times there may be duplicates in leads and accounts. If you do B2B Sales, you need to clean up that list so you have one entry per potential customer (not including child accounts). Many times, duplicates have similar but not identical names (E.g., “Microsoft Corp” vs. “Microsoft Corporation.”). For analysis purposes, these need to be consolidated into a single record. A tool like the Fuzzy Lookup Add-on can assist. 

Equally important is the prevalence of more than one lead per account. Double counting will significantly skew the results. Instead, let’s say you have 5 leads from a single account; pick the most appropriate lead source (usually the first) and treat it as a single lead.

Data Quality Consideration – Your Outcome Field

Any analysis is dependent on your definition of “success”. In analytics, we refer to it as “outcome”. As discussed, lead duplications can “dilute” the perceived effectiveness of the lead source due to having too many items in the formula’s denominator. Another problem is not having an indication of a successful outcome for some converted leads. Most CRM databases have a converted date field or similar, but that field is only populated if the sales rep followed the correct process. We all know that is not always the case. So, looking through all your opps, make sure there is a corresponding lead (or account if you don’t use leads) tied to each opportunity – if not, use the same approach as above to match the company name and add the missing conversion date based on when the first opportunity was created.

What Does Success Look Like?

Success is when you perform this analysis on a continuous basis (Cien.ai automates this process…). This lets you understand the true propensity of conversion, with a clean list of lead sources, there are almost always at least one or two lead sources that generate much lower conversion. Remember, if you have one lead source that converts at 5% and another one that converts at 15%, the difference in success rate is 300%! Reallocate spending & effort from the worst ones to higher-yielding lead sources to create a much stronger new logo pipeline (and tell your reps that lead quality is a corporate priority to keep them motivated).

About the Cien.ai Practical RevOps Analytics Series

This article is part of our Practical RevOps Analytics Series, inspired by our work with B2B business leaders, growth consultants, and PE operating partners. These articles focus on the technical aspects of improving GTM performance. If you want to dig in on the business details of how to improve the concepts we use here, please refer to our “Growth Essentials Analytics Series”.