Lies and statistics

You may recognize the title to this post as part of a famous quote popularized by Mark Twain. There are two key words missing from the quote, but this is a family friendly blog, so….

Let’s establish up front that I’m a big fan of marketing. Especially today, where marketing is truly data-driven and often plays in the sales team’s sandbox. Branding is also fun, but that’s the jumping off point to the fluffier marketing stuff that isn’t so determinable. Data-driven marketing creates visibility into the straight lines between cause and effect, which is exceptionally valuable in business. It also, fortunately and unfortunately, prompts the use of statistics.

Marketing and statistics are like those friends of yours who seem like a perfect match but do nothing but bicker when they’re together. The problem is that statistics can be manipulated to tell nearly any story, regardless of whether or not it is a true story. A common example of statistical manipulation is the comparison of changing averages in a time series without reference to the underlying changes in population. Even more common is choosing analysis start and end dates to support an argument when different dates paint a contradictory picture. Read financial news with some skepticism and you’ll see these ALL THE TIME.

In indirect consumer finance, one of the “we’re the best!” claims that emerge from the union of marketing and statistics is the approval rate claim. High approval rates are good for contractors, of course. But when is an approval really an approval? And when 4 of 5 lenders claim to have the highest approval rate in the industry, has it ceased to mean anything?

Here’s a typical way approval rate loses meaning. Imagine a finance company that approves 50% of the applications they review. One day a risk person realizes they can get a small boost in loan volume if they counter-offer deals that are close to what they’d normally approve. A request for $26k counter-offered to $25k is something most contractors would happily accept. The change increases the finance company’s approval rate to 55%, the positive press marginally increases business, and there’s no effect on portfolio risk.

When marketing validates the impact, they start thinking of how to really take advantage of high approval rates. They eventually think…what if an unqualified applicant requests $26k and we only offer $5k? Most contractors won’t take the offer and, if a contractor takes it, the loss given default is tiny. In discussions, the risk people agree that it doesn’t impact the portfolio meaningfully, but why bother? Because counter-offering all those big requests for small amounts boosts the approval rate to 75%. The change results in an notably high approval rate but provides little to no benefit to the applicants or the contractors.

It is tempting to make marketing claims that are true but not true. I think people are generally good and, even in this case, the end goal is usually a genuine interest in acquiring customers and serving them well. But in our personal interactions, we would typically call claims like these “white lies”. The basic claim is true but the intention is to deceive or manipulate to some degree. And, because flimsy marketing claims can be found everywhere, sticking to the simple facts does little to attract customers. So that leads to two questions: is such marketing a bad business practice or an innocent means of making a product sound attractive? Is it even possible to completely avoid this kind of marketing?

[/vc_column_text][/vc_column][/vc_row]