Proof Restaurant “Analytics” Companies Aren’t Actually Analytics Companies

Facebooktwitterlinkedin

The history of solution providers in brick and mortar is mostly the same, and that has made for an abysmal outcome for the industry. We will start with an often avoided, but nonetheless true allegory to show what we mean. This is a copy-paste from POSitouch’s own website.

Bill and Ted Fuller, owners of the Gregg’s restaurant chain, a high volume, low check average, family restaurants in Rhode Island, wanted to install a computerized point of sale system for their operations. In 1980 they began a search and evaluation of existing systems specific to the food industry.

After two years of searching, they found no system that fully addressed their needs. This prompted a decision to develop a system. It was clear that there was a need that the POS industry was not filling.

Development on POSitouch began in 1982 and a product was readied for testing in Gregg’s restaurants in 1985. Also in 1985, Peter Lipman joined the Fuller’s. With his technical guidance and the Fuller’s industry knowledge, the product continued to evolve.

The creation of POSitouch, a restaurant point of sale software, stemmed from two people who were not software engineers: they were restauranteurs. They were undoubtedly familiar with some industry problems, but did that make them suitable to build a software product, particularly one that would touch lots of data? An answer might be concluded with two following points:

  1. POSitouch is currently reliant on a third party solution (Omnivore) to offer cloud connectivity for its merchants. This increases costs for customers.
  2. Its largest reseller has built its own cloud POS as its customers couldn’t achieve suitable value from POSitouch configurations. This creates uncertainty for customers.

But POSitouch customers in the early days were satisfied, right?

Maybe. To truthfully answer this we must put things into relative terms. Think about the core process of problem solving. People purchase solutions because they solve problems. If a customer is unaware of their own business problems but they set out to build solutions that similar business should use, can the solution be that valuable?

Think about your answer in terms of an analogy. The year is 2017. Science and technology have proven that vaccines minimize material risk to human health. Yet there are a number of remote tribes that do not so much as keep basic hygiene. Would we seek their counsel in the development of new biomedical treatments? Should we trust people who have such a hostile response to new ideas (as we talked about previously as expedient escape) that they attack anyone showing them how to improve?

The problem with brick and mortar is that you (generally) have unqualified people building shoddy solutions because they think they can do it better. If we were talking about more sophisticated industries – like finance – we might be more inclined to believe in their solutions. Jeff Bezos, Amazon’s founder, was an engineer who worked on Wall Street, after all.

The perpetual cycle where brick and mortar merchants build new solutions because all existing solutions are “crappy” has this sort of self-fulfilling prophecy going for it: most solutions in brick and mortar are total crap.

So when merchants set out to build a “better” solution they’re likely driven by the right conclusion. It’s just that the merchant, unaware of what’s possible or what’s been successfully implemented elsewhere, has no objective basis on which to measure the quality of their own solution. It’s this level of sophistication that stops them short of executing something great.

Like the POSitouch allegory, most “analytics” solutions in brick and mortar were not started by data scientists. Nor were they started by engineers or computer scientists. In fact, they weren’t even started by professionals who had seen good solutions implemented in other industries.

Thus brick and mortar solutions leave much to be desired… but merchants are totally unaware of what to desire!

For proof, we’re going to share analysis on a 600-unit restaurant chain using a “best-of-breed analytics” tool. We took the POS data from this chain and determined annualized ROI for implementing operational changes with machine learning. Since the POS data is in itself limited – that is, one would need data from other systems like inventory or marketing to pair with it to analyze a business more completely – the ROI is superficial. By that we mean that ROI could be substantially higher if more data were available.

If you need more details on what these metrics spell out you can ping us directly, but in general Employee Performance is optimizing employee behavioral patterns to work the right people on the right shifts using machine learning, and Labor Forecasts is predicting demand with machine learning on externalities (school schedules, sports schedules, weather, etc.) and matching labor schemes to reduce over and understaffing.

So this is a lot of money, right? Of course it is! But if you were to tell merchants that there were better ways to solve their problems, or that they have problems that they don’t even know about, most merchants would disagree.

I buy solution X from Y. Y used to run a restaurant so they know best!

And that’s if they even respond at all. If you were a CEO of a company, and it was your job to make sure the company delivered maximal returns for shareholders, you would seek the best solutions, wouldn’t you? You wouldn’t care where those solutions came from so long as they were fair and reasonable, right?

Yet that’s not how merchants think. That’s why we’re shocked when low margin retailers don’t even bother to return phone calls or emails. When a decision maker ignores something like this for months or years they need to be held accountable. The length of time ignored * the ROI per unit of time needs to come right out of this person’s pocket. And since no merchant “executive” can pony up $40MM a year, they should be fired. Immediately.

Unfortunately too many merchant CEOs and board members are no different than the people they employ. So, because the industry sources products from its own ilk, it has not evolved.

The good news? This is changing. And since so many merchants are objectively run so poorly, there’s only one direction they can go: up.

POS companies, with access to merchant transaction data, can start using said data to produce “audit reports”. These reports can showcase how analytics will objectively improve a merchant’s specific business, and often times open the merchant’s eyes to what they’re leaving on the table. If the merchant is already using a solution, the audit can prove the deficiencies of their current provider. And if the analytics are undertaken by qualified data scientists (usually via a partnership with a third party like us), and white labeled by the POS company to remove merchant friction, it’s really hard for a merchant to refute any of it.

This now becomes an opportunity for the POS company to sell an analytics tool to their merchants that leverages advanced data science and fixes these problems. Selling on ROI: what a novel idea. It’s almost like we predicted this inevitability…

This, among many others, is a reason why POS companies should be collecting transaction data. With access to data and the right partnerships, these types of solutions can be made available to a POS company’s merchants. The POS company makes more money, merchants become more successful, and merchants are more likely to stick with their current POS provider, reducing churn.

Of course some merchants won’t believe the data, or will just refuse to change. Even showing them how much they’re screwing up will do little to convince them to uphold their fiduciary responsibility to shareholders. You can’t help everyone. But at least POS companies will be made aware of their at-risk merchants. As far as we’re concerned this group of customers can suffer with “crappy” solutions built by former merchants on their way into insolvency.

Facebooktwitterlinkedin