DLA Piper’s marketing team revealed the inner workings of its massive client retention data analysis program at ILTACON 2017.
Even the biggest firms can have trouble retaining and growing revenue from clients. That was the issue facing DLA Piper in 2014. Out of 20 clients that the firm had placed in its key partner program, 18 had revenue to the firm stay flat or decrease over the previous year. And in the firm’s marketing department, many were at a loss why this decrease was happening, let alone how to reverse it.
The answer they found was in analytics—but not just in looking at the numbers, but rather forming a data action plan focusing on specific (and perhaps unexpected) metrics that the firm found influenced retention rates more than others.
The marketing team, alongside analytics partner Axiom, revealed the results of this endeavor to a standing-room-only crowd at the “Harnessing Predictive Analytics to Drive Client Growth and Retention” session on Monday at ILTACON.
The idea for the project came from an article Kim Rennick, director of sector marketing at DLA Piper, read in Time magazine. The article explained new and interesting ways Big Data was being used on college campuses, such as utilizing influential metrics to identify students most at risk of dropping out, then giving them extra care and tutoring. Rennick saw numerous applications to the law: “You could say that’s client service for the student,” she explained to the conference crowd.
However, firm leadership was not yet on board. Rennick explained that they were “not as excited as we might have wanted” at the proposal from DLA Piper chief marketing officer Barbara Taylor. But the marketing team was allowed to take on the project as skunkworks.
Building the Model
To actually launch the project, DLA Piper undertook a four-step process. The first key step, said David Kuhlman, partner at Axiom Consulting Partners who helped spearhead the initiative for the firm, was to establish the foundation of what the data was looking to solve.
“It’s not important to set up a question at the beginning. You just have to know the direction you’re going,” Kuhlman said.
For DLA Piper, that meant finding out more about client retention. Heather Reid, director of practice marketing at DLA Piper, explained that this meant the firm would “single out what we did differently with clients that grew vs. clients that shrunk.”
Second, the team actually had to build a data model, which Kuhlman described as “70 percent of the work.” In this case, that meant combing through four years of data and about seven million records to find the insights the firm wanted.
Third, those analyzing the data needed to examine the relationships between variables and see what made sense. This is because some variables that seem statistically significant can actually have nothing to do with what you want to find out (such as geographic location of clients) or can draw distracting and incorrect conclusions.
Finally, the team would build the predictive model with the variables it deemed worthwhile. Here, Kuhlman had two pieces of advice. The first was to find “the biggest impact on the least number of things in the least number of places”—in other words, reducing the number of variables so you actually have something actionable to take to firm leadership.
His other piece of advice was to pick a model that works well, but perhaps not best—it’s more important to have something that’s easy to explain and adoptable. “Unless people take action on the result, it’s all useless.”
The DLA Piper Results
Ultimately, the firm settled on four key variables that its data model found directly affected client retention:
• Reducing the size of matter teams to five or less and increasing time per team member proportionally where possible;
• Introducing one new professional to the relationship;
• Adding one more industry expert to the team (which could coincide with point two); and
• Running a focused, relevant marketing initiative for each client.
Using the model, Axiom helped DLA Piper identify 1,200 at-risk clients that could be a target for the strategy, which the firm then pruned by considering the partners involved with the client, client size, and industry.
Now, DLA Piper had its list, but it still needed attorney buy-in. The marketing team learned through analyzing past data that focusing on these variables could see a 75 to 80 percent retention increase, but actually convincing partners required a lot of legwork.
“This was the signature project of the three of us here [from DLA Piper on the panel],” Taylor explained. We understood how cool it was … and were available at a moment’s notice to give advice when needed.”
The team also set up a teleconference every other week with partner attorneys to monitor progress and talk about how the variables were being carried out. As Taylor said, “The team got the sense that this wasn’t one and done. We kept at it, hammering away.”
That hammering occurred through successes and failures. The DLA Piper members of the panel all said they were forced to be more flexible than they originally had anticipated, even kicking members of the project team off if they weren’t being accountable. Rennick also adjusted her sales strategy to partners partway through, going from talking about “at-risk” clients to clients “primed to grow.”
In DLA Piper’s case, the project turned out to be a massive success. When comparing a control group to a group focused on improving those key variables, DLA Piper was able to prevent 85 percent of fee loss on a year-over-year basis. The actions started to take effect within about six months, and the team found that the more variables that were acted upon, the better retention would be.
That translated into hard dollars for DLA Piper: The firm estimates an increase in revenue of $37.6 million for the firm, given the high-leverage clients involved in the process. That means that not only was this project a massive success, but it’s likely to be expanded in the future—Taylor explained that the firm is now updating its client and internal partner lists and securing increased capacity for the project.