Gnawing on Numbers

Occasionally a client will send a spreadsheet of company statistics and ask me to comment on what I see.

I usually look and see ambiguous statistics but I certainly don’t want to say that.

Discussing business numbers with people is like discussing religion. No matter what you say, you’re unlikely to change their intrinsic beliefs, so I always approach these conversations carefully.

“What do you see?” I ask.

“Well, last year 68 percent of our customers were repeat customers and 32 percent were new customers. Now we’re selling 63 percent repeat customers and 37 percent new customers.”

“What do you think this tells us?”

“It tells us your ads are working!” the client says excitedly.

“Perhaps it does,” I say. “But it could just as easily indicate that our competition is growing stronger or that we have somehow offended or disappointed our old customers.”

My client gave me a confused look, so I continued, “If a smaller percentage of our business is repeat customers, couldn’t this mean that fewer customers are choosing to buy from us again? Couldn’t it indicate that we’ve disappointed them somehow?”

The confused look became a worried look. “But our sales volume has never been higher.”

“I know that,” I said. “But that could mean that we’re bringing in new customers fast enough to disguise the very serious problem that we’re losing our old customers to someone else. After all, you said yourself that our percentage of repeat customers is down.”

“Do you think we have a problem with our old customers?” the client asked, now truly worried.

“Not at all,” I smiled. “I’m just saying that nothing can be learned from the numbers you gave me.”

Not everything that can be measured has meaning.

Many of you are now recoiling in doubt and disbelief. I get that. Like I said, talking about business numbers is like talking about religion.

Here’s how I finished that conversation: “If a company sells a product or service that people buy once a year, what percentage of their customers will be new customers in year one?”

“One hundred percent,” said my client with confidence.

“And if our sales volume doubles in year two and exactly 50 percent of the customers are new customers, what percentage of customers did we retain from year one?”

The client thought for a moment, then said, “If business has doubled and one half of our customers are new and the other half are repeat, this means that one hundred percent of last year’s customers chose to buy from us again.”

I continued, “Sales in year three are exactly triple the sales of year one. One third of the customers are new and two-thirds are repeat customers. What does this tell us?”

Another moment of thought, he answered, “We have 100 percent retention of customers from the first two years.”

That’s when I said, “But someone is likely to point out that your percentage of new customers is falling and they’ll likely interpret this to mean that your ads aren’t working. After all, your sales volume grew 100 percent in year two but only 50 percent in year three and your percentage of new customers has fallen from 100 percent to only 33 percent. You’re now doing triple the volume you were doing just two years ago but these numbers would seem to indicate that you’ve got serious problems with your advertising.”

The client began to smile again, so I continued, “Oh, and I forgot to tell you that this company increased their prices by 12 percent at the beginning of year two, so none of what we just calculated is accurate. And that company has only been in business for 3 years! Your company, on the other hand, has been in business since 1939 and you sell a product the average person buys every 13 years and lots of old customers have died or moved away and new people have moved to town and some of your old competitors have gotten more aggressive while others have gone out of business and we need to factor in the percentage of sales opportunities your salespeople are closing and yes, you’ve also got a brand new ad campaign. If we take all that into consideration – assuming all the data is available and can be trusted – how are we going to calculate it and what do you think we’re going to learn?”

He smiled as he ceremoniously tore up the spreadsheet and said, “We’re making a lot of money and I like the ads.”

“Good. Let’s go have lunch.”

So we did.

When I got back from lunch, two other clients had emailed spreadsheets to me and asked me to comment on what I saw.


Roy H. Williams