To Build a Retail Change Engine, Stop Fixing Squeaky Wheels

MAY 9, 2018

Retail customer experience is changing profoundly. Every year, companies roll out major new strategies and technologies to stay relevant to customers, but then struggle with their implementation and measurement at the store level. Their solution, understandably, is to set new standards for stores and use metrics to hold them accountable. But metrics can be tricky and sometimes make the problems worse. Are you looking at too many? Not enough? How do we get to the “just right” metrics?  

As a COO, I have launched metrics-driven initiatives that failedbecause they were too complicated or gameable; because it was unclear how individuals could affect them; or because there were just too many of them. I’ve learned from my mistakes, and I’m fortunate to work for a software company that can also help solve these problems for our retail customers. If retail is a well-oiled machine, then a brand’s metrics are the individual parts that make the machine move. Retailers have to be sure they’re focusing on improving the engine instead of fixing one squeaky wheel at a time.


To motivate and manage the shifting retail scene, companies tend to rely on measurement. Analysts in HQ crunch reams of supporting data to set targets for every store; the field organization shares these targets, sometimes along with training materials, execution plans, and accountability checks.

From the corporate perspective, these tools and standards are a no-brainerthe market demands them, and it’s hard to imagine how any major change could happen without them. But they often don’t work.

A typical store manager has access to hundreds of store metrics and is working on over 10 “top” priorities at a time. No one person can get all of them completely right. In addition, every minute a store manager spends on reading, data analysis, measurement, and reporting is a lost opportunity for impacting the customer experience firsthand on the floor. With more than one corporate change initiative running at a time, each one fragments and detracts from focus. Some people are better at interpreting the numbers than others, but in the end, store managers should be people people, not number people.


Imagine you’re a store manager and you’ve just learned that your store is currently below its NPS target, you’ve missed some points on a recent audit, your payroll is above budget, your foot traffic is a little lower than last month, profitability is down, sales of a few top sellers have slowed, your stockout rate is too high, you’re missing an opportunity to earn an incentive, and your shrinkage has just gone up. Overwhelming, right? But wait: you also have access to a database with 3,000 other store metrics, many of which show you’re deficient. What do you do?

Good store managers will pick at least one of these problems and target it with an improvement plan. But which one? In spite of all the meticulous measurement, there’s usually no clear answer.


When numbers fail us, we rely on experience and emotions. Often, they point us to the problem that screams at us the loudest: the squeakiest wheel of our retail machine.

For a store manager, this could mean spending lots of time satisfying the customer that complains the most, focusing on the lowest-performing metric, or reacting to the latest or loudest communication from corporate about what needs to change.

While these approaches will certainly help a little, as soon as the squeakiest wheel gets the grease, something else screeches into view. Suddenly store managers are trapped and moving frantically from one thing to the next, oiling each problem just enough to keep it from squeaking for the moment.

What if we let some of the wheels squeak and instead focused on upgrading the engine?


Numbers can be the solution as well as the problem. Store performance measurements must be put in context:

  •      Is this problem isolated or a part of a pattern?
  •      What will happen if I do nothing?
  •      What will I gain if I am able to improve?
  •      How am I doing compared to others?
  •      What can I do to improve?

If we had the answers to these questions for each of our 3,000+ metrics for a particular store, we could ignore the thousands of isolated, inconsequential issues and spend time fixing the few things that matter to it the most. Unfortunately, no human being has time for that.

The flaw isn’t always that we have too many metrics. It’s that we’re using only humans to interpret them.


The flaw isn’t always that we have too many metrics. It’s that we’re using only humans to interpret them. Machines with the right AI in place can crunch through thousands of metrics and find important ranks, trends, and correlationsspecific to a storein seconds.

In our store manager scenario, the #1 root cause of some problems may be high associate turnover. Perhaps high turnover led to hasty hiring and training, which in turn led to inefficiency, employee theft, stocking errors, and an overall decline in customer experience. In this case, the right solution could be not to address any of the issues directly, but to improve associate NPS through training and development efforts.

Any good store manager or district manager will see this and address it once it’s a big problem.  However, in the age of AI, we can uncover these things earlier, and upgrade the engine’s acceleration…all while it continues humming.

If we let the machines do the analysis instead of just the measurement, we can free people to be the change agents they need to be.


We believe that to drive real change at the store level, retailers need to escape the Squeaky Wheel Problem. By using software and data science to deliver a short list of store-specific insights and recommendations, endless metrics and reports become more meaningful and, most importantly, actionable. This is exactly what Square Root’s solution, CoEFFICIENT®, sets out to do.

If we let the machines do the analysis instead of just the measurement, we can free people to be the change agents they need to be.


Elizabeth keeps Square Root operations running like a well-oiled machine. With over a decade of experience moving corporations to data-driven decision making, she’s our resident business analytics expert. She has consulted extensively within the automotive, manufacturing, software, and energy sectors on business processes and enterprise software solutions. Elizabeth holds a degree in Computer Science from Harvard University.