Traditional Lead Scoring Methods are Expensive and Flawed.
You Can Do Better — Here’s How
A few days ago our friends at Periscope Data posted an excellent description of how they score their leads using Clearbit-provided enrichment data. We do something similar on the enrichment side, but we also score leads based on engagement. The Periscope piece inspired me to share how we score engagement and, more importantly, how we’ve used this scoring to drive marketing and sales alignment.
Aligning marketing and sales is f*cking hard. It’s something you hear about all the time, and think you’ll never struggle with, until you do.
There’s so much software out there that offers some sort of solution. Maybe I haven’t given it enough chance to work. But it always feels like I’m throwing money at a solution that isn’t really tailored to me, or my prospects.
So, as Director of Growth for RJMetrics, achieving this alignment has become my number one priority. Why? Because highly aligned organizations grow faster.
As any good analytical growth-focused marketing leader at a data analytics company would… I solve this problem, with data. Conversion rates, leaky funnels, misaligned nets, prioritization — all of these can be found, easily. You just need your funnel data — all of it, from all systems — in a central analytical data warehouse.
Setting lead scores based on data, not intuition
What are each of my marketing touches worth? And knowing this which leads should be prioritized by sales?
At RJMetrics, the answer to these questions is just a SQL query away. We’ve aggregated our sales, marketing and advertising data into a data warehouse. In this post I’ll show you the exact queries we use to answer these two questions.
What you need to do this yourself
To run these queries you’re going to need Salesforce Opportunity, Account and Contact data as well as your marketing automation data in a single analytical data warehouse. This is not a small feat if you’re going it alone, but using RJMetrics Pipeline makes this process trivially easy. Manual exports or querying the Salesforce and Pardot API to get this data would take days or even weeks, and, honestly, the whole process is an unsustainable mess. Trust me–before we built Pipeline I tried to do this myself and gave up.
On the other hand, Pipeline draws the data together into a central location so it is ready for you to query in minutes, instead of weeks or months of data wrangling. And then it keeps your data up to date with little to no latency. It requires next to no configuration.
I’ve posted the entire SQL query here, so you get a sense of the end result, but in this post I’ll walk you through each of the sections. The outputs are:
- a table of marketing activities and their relative values
- a table of prospect/lead IDs ordered by their aggregate lead score
Find All Activities
This is the simple part. We use our marketing automation data to get a list of every form submission and view. We count them to know how many times prospects have participated in each activity.
Find All Activities for Won Opportunities
Finding data specifically for Won Opportunities is much more complicated. We join four objects together just to get marketing attribution data linked to opportunities: Salesforce Opportunities, Salesforce Accounts, Salesforce Contacts, and Pardot Prospects. Once we have Prospects, we can join into Visitor Activity. This gives us activities for prospects related to Won Opportunities. I’ve limited touches to those occurring within the 3 month window before the creation of the opportunity, but this may or may not appropriate for your purposes. In addition, you may want to use Opportunity Contact Roles or something similar, but I wanted to see results using all Contacts, not just the ones our reps say are directly involved in a deal.
Find Total Revenue Generated by Activity
We know how much revenue is generated for each Opportunity, and we know the activities that have influenced that Opportunity. Taken together, we can split attribution of revenue to each activity, and sum across Opportunities. More or less, we can figure out the amount of revenue each activity is responsible for. In this case, we are using equal weighting, but you could choose to use other attribution methods.
Find Revenue “Generated” For Each Activity
So since we know the total revenue generated by each activity, and we know the total number of times all prospects, regardless of whether there is an associated Won Opportunity, have engaged in the activity, we can find the value per activity.
Getting the list then is as simple as running the query SELECT * FROM mrr_per_activity.
Find Aggregate Lead Scores (or: which leads should be prioritized by sales)
Finally, now that we know what each activity is worth, we can sum the value of all the activities listed for each prospect and find the aggregate lead score. This will be the final output used by sales to guide focus.
Besides surfacing the best leads for our sales team, we now have a shared understanding of what marketing success looks like. No longer is it “get us more leads.” Instead, it is “drive meaningful activity from our targets.”
Notably, the results here are very similar to what you get when you pay thousands of dollars a month for marketing attribution systems. Personally, I’d rather be in control of the algorithm we use to value our leads, and I’d also rather save a few dollars in the process!
Most importantly, we’ve found that looking at the results in this way allows all of us to make sense of our funnel. It helps to speak a common language — data — and that reinforces our shared goal, growing RJMetrics.
Special thanks to our data scientist, Yevgeniy Meyer, for helping me to make data-driven lead scoring at RJMetrics a reality (and for helping me to write this post).
