Like a good portion of families on Thanksgiving, the Lions and Cowboys took their places at our table during the holiday festivities. The Lions are on because we’re Michiganders, and we love the usual “Black and Blue Division” grudge matches with our turkey and dressing (the other games are OK, too). We watch the Cowboy games because football. It was during the Dallas vs. Washington game that my wife’s uncle said something that made my nerves tingle.
Out of nowhere, he commented that the NFL needed new ways to measure quarterbacks. He thought that pass completion was a poor measurement of a QB’s performance. In his words “some QBs can hit a guy’s hands all day long, but if he has crappy receivers, it’s somehow his fault.” We have those same issues inside of car dealerships, and we’re part of the problem.
I spent a great deal of time this summer contemplating the metrics we use to measure the success of Internet departments. It’s something that Joe and I have written about extensively. Hell, data analysis has been a giant part of my career. It’s what I did while working in venture capital, it built my success in retail, was the very basis of developing lead scoring, and is the cornerstone of our consulting efforts. I concluded that the measurements that have been thrust upon us simply ignore all of the variables that make up a successful dealership. It’s not that the current metrics should be ignored, but they need to be considered as part of a wider group of more significant measurements.
Romance and measurement.
The science of sports analytics isn’t much of a secret after Michael Lewis’ book Money Ball became a hit movie. Far too much has been written about it in our space, so I won’t wear down that path. What isn’t made evident is how the data is processed, and what resulting metrics are meaningful. The Signal and the Noise by Nate Silver rattled something loose in my brain. What Silver uncovered for me with was the sheer volume of measurement that can take place. Any and all variables can be analyzed from the most obvious, to the most minute, all corrected for the passage of time. Moreover, as new technologies become available, things that were previously hidden from measurement, can now be analyzed with precision, thus altering the formulas that were previously used to determine success.
However, even in the face of new data, traditions can color our judgment. After all, people blindly love sports. Just look at any fan who roots for Philadelphia. The algorithms used in many sports analytics programs contain biases built-in by the programmers and users. These “rules of thumb” can easily skew results and exacerbate failures that were not anticipated during the creation of the model (Lewis recently wrote another book, the Undoing Project, where he addresses those oversights in Moneyball). The metrics that we so romantically hold dear are from the easy-to-measure days, and often cloud our judgment, sometimes forcing us to overlook areas where we’re actually successful.
Moreover, these biases alter our perception of what’s factual. Because these biases are often built into our tools, the data that’s presented is colored by the provider of said tool. Simply put, it’s up to the team that created the tool or module to decide how data is presented, and sometimes it’s not even consistent.
Take “appointments set,” for example. While writing this, I ran three separate reports inside of a widely used CRM, using the same date parameters, only to find three wildly differing numbers for the exact same metric. The same yardstick should measure thirty-six inches weather it’s in Michigan, Montana, Mongolia, or Monaco. If a company can’t clearly define the same metric for its teams, what hope do we have?
Let’s put sports completely aside for a second, and go to science. The advent of new measurement tools and technologies fundamentally alter the way we see everything around us. Gas was the smallest composition of reality until it wasn’t. Molecules gave way to atoms. Atomic particles became the building blocks until they weren’t. Subatomic particles are the smallest known quantities until we can crack into the quarks, leptons, and bosons that comprise them. That’ll be sure to keep the Big Bang Theory on TV for the years to come. We’ve harnessed new technical breakthroughs to measure our predictions, and when they’re found to be true, we alter our measurements to reflect that new reality.
Unfortunately, retail automotive doesn’t have that same commitment to measurement. Our CRM companies rely on programmers to develop the reports. Our DMS companies hold our data hostage. Our dealer academies rely on a curriculum developed decades ago. Our dealer associations hold onto tradition. Our dealers have too many other things to worry about. Despite the potential for understanding our success down to the subatomic level, those in the retail community are somehow stuck on a shiny metal’s finish, buried in the same biases that have been perpetuated for decades.
While others have been content to stick with antiquated metrics, we’re tired of it. Quietly, we’ve been gathering data at a clip that would make most vendors squirm. How? Since the big guys aren’t doing it, we developed a proprietary technology that can provide a far deeper analysis than what a CRM can provide. In the last six months, we’ve analyzed in excess of 100,000 leads, and over five million data points to search for metrics that can better measure the performance of the agents who are working inside the CRM. Using the previous generation of our technology, it took five years to analyze 20,000 leads. Now we can do that in a weekend.
We recently collaborated with VinSolutions to explore new variables that can help dealerships monitor CRM utilization and benchmark themselves against a broader population. Not just those who are in their DMA, or the best in class, but every VinSolutions dealership in the country. In all, we analyzed data of over 300,000 leads for the study, and as our technical capabilities grow, we’ll be certain to gain even more insights as time goes on.
We’re just getting started, and we’re pushing all of the chips in.
Simply measuring leads, appointments set, appointment shown, and units sold gives us an abstract view, at best, into how our Internet operations are performing. Metrics such as quality of response, response methodology, lead attribution, behavioral correction, among dozens of undiscovered variables, will help dealerships to fine tune their operations, more precisely measure their employees, and, of course, sell more cars.
Just like scientists classifying states of matter, the metrics we look at should be broken down to the subatomic level. We need to be constantly exploring the correlation of these variables as it pertains to the whole retail ecosystem, completely free of the biases that we’ve created for ourselves. We do not live in a two-dimensional world of sales volume and revenue. We live in a three-dimensional world that’s governed by time, in an electronic age where data can be measured to the infinitesimal. DealerKnows will no longer be content with the measurements that were.
As 2017 fast approaches, we’re exploring new partnerships and solidifying opportunities that will help us with our agenda. And, we’re open to aligning with others who will join our cause. There are a lot of great quarterbacks out there, throwing to less than spectacular receivers. Like the analysts that are rapidly making up the front offices of professional sports, we’ll be there leading the charge to minimize the influence of tradition and maximize the results of every dollar spent. It’s finally time that we give our dealer partners the information they need to make smarter decisions about the entire value chain that makes up their operations. As the dealer community faces even stronger external threats, we need to adapt like our livelihoods depend on it. If we somehow fail to win the war against the dealer community, at least we will know precisely why, and we won’t be forced to guess.