How to Take Control of Information
In our industry, the current availability of analytics and forecasting tools can provide a retailer with a significant competitive advantage. I wish those tools would have been available early in my career. I frequently was impatient and quickly reacted without the benefit of all of the facts. I didn’t understand that I might have actually been making the situation worse instead of better. I would often start my week reviewing the previous week’s sales data. Anxious to make a good judgment, I would compare the results against budget and prior year. If the results were better, I was reasonably comfortable that the trend was up and things were going well. If the results were worse,
I would begin to pull the levers and hope for a solution. Inadvertently, I was causing more variation in the system; the net effect was that I was tampering.
Fortunately, I began to learn more about Deming’s management method and came to realize that two points a year apart did not represent a trend. Our role as managers and leaders is to make good decisions that continuously improve the system vs. creating so much variation in a system that it is nearly impossible to detect a trend. In fact, one of the biggest lessons I learned was that not every dot on a run chart graph comparing sales to budget or prior year is important. The trajectory is important and tells the story.
What I learned was to use control charts to look at data trends over time and make meaningful comparisons. There are three basic rules for interpreting data trends in a control chart (see example below) to gauge if something “special” is happening:
- Trends: Six points in a row trending up or down represents something “special” is happening in the system. A trend of at least six in a row indicates that the system is shifting.
- Shifts: Eight or more points on either side of the center line also indicates that the system likely has shifted.
- Outside the Limits: Statistically, the limits are set roughly three standard deviations on either side of the mean. A point outside of that area is a special cause. Data graphing and statistical tools prevent weekly gyration and overreacting to a situation instead of letting the system establish a trend. Acting too quickly actually puts more variation in the system.
As I learned more about how to use control charts, I began to appreciate the importance of hypothesis testing. I relied on internal and external experts to assist with the analytics and the development of the statistics. If your desire is to grow and the trends have been established, it is important to carefully change inputs that drive better results. Be sure to monitor your change and look at the effect of the changes. This process may seem slow, but it can make a positive difference. How often have you tried several new ideas, only to ask yourself if any of them made a difference, or which one made a difference?
Once I started graphing information and looking at the rules, then it became more predictive. Granted, I wasn’t always happy with the performance or the time to make an improvement, but just setting a goal was not the method to change performance. It is like yelling at the oven as a way to get a better cake instead of changing the recipe. A specific recipe for improvement or action is required. Then it is important to have the patience to wait for a result.
While many times the change did not produce the desired outcome, at least I knew for the future what the data was telling me and the possible meaning behind the trends. I felt more in control and better able to know which levers to pull.
A word of caution: Be wary of common theories about “why” business is up or down. They are often untested. Using data to debunk a myth is important. Examples include:
- High retail gas prices are causing customer count declines.
- Unemployment is reducing sales.
- I’m losing my morning business to a quick-service restaurant.
If these examples sound like familiar statements, you may want to compare stores sales to other stores in your chain. You may discover that high gas prices may cause customers to stop more often. If a store has many customers paying cash, the store could have an increased transaction count. The customer paying cash still spends $20 for gas per stop. If gas retailed at $4, they get 5 gallons; if $3, they get 6.66 gallons. Very few drive less, but many comeback more often. Note: The average gallon fill on credit cards decreases as retail rises.
Break It Down
Later in my career, I also became more aware of special cause vs. common cause. For instance, if a spike in sales occurred more than three standard deviations above or below the mean (outside of control limits), I would look for a special cause. (Was it a blizzard, power outage or something that happened once or seldom?) Reacting to a special cause does not produce results. However, identifying the cause can create understanding.
Finally, the most difficult challenge is looking for input measures to affect your business. Personally, I found this activity the most challenging. But the breakdown of a transaction created some clarity. In sales you can get more customers, sell them more units and/or sell at a higher retail per basket. The trifecta is to have a higher customer count, an increase in unit count per customer and a higher average retail per item in the basket:
- More Units: Basic merchandising affects unit count in the basket. Bundling, multiple pricing, merchandising related items in proximity, good traffic pattern in the store and good displays, plus other methods (such as suggestive selling) also make a difference
- Higher Retail/Basket: Retail price management can improve average retail per item in the basket. The pricing component, along with merchandising the higher retail items in first position, product selection and price sensitivity studies, will improve basket retail. Don’t gravitate to price sensitivity without having the facts.
- More Customers: Customer transaction count gets to the basic principle of “why” your customers select the store. Naturally, you must have a convenient location. Yet once that is established, what really makes a difference? In my opinion, the differentiators are friendly people, speed of service and product availability(including offer, and not having out-of stocks).Your offer must meet the consumer need and most likely include food. Combine this with a clean store and clean restroom, which are must-have components. If you noticed price was not a major factor, it is because it is assumed the stories reasonably competitive, not necessarily low-priced.
What levers do you pull when faced with a declining sales trend or wanting to increase sales growth? Analyze the situation first. What makes up the sales loss? Is it units per transaction, average retail price or customer transactions? If it is transactions, then ask why customers are selecting another retailer. Or has your environment changed? If so, what is the root cause? Are the employees friendly? Do you measure by customer survey? Are employees satisfied? It is hard to be friendly when an employee is not happy. Is your organization creating barriers to fast service? Does the store have the offer your customers want, and is it offered in a clean environment? Do you have the facts you need? Great solutions start with great questions.
These things are harder to deliver and even harder to duplicate if you are succeeding in a friendly store with the products that the customer wants in a clean environment. I doubt a major promotion or a great price on an item will be anyone’s silver bulletin fact, it can be negative to average retail, and you can create a negative spiral, being judged for commodity pricing.
In summary, to create a sustainable competitive advantage, focus on great customer service and a clean store environment. To build sales, differentiate with customer service and get customers in the door more often. Understand how to use data more effectively, avoiding the mistake of overreacting, and use your information more wisely than your competitor does.