Thread by @minethatdata

1 - Why do I care about new customers when my industry obsesses with catering to best customers via omnichannel theory?

Good question!

It goes back almost 30 years.

2 - Ok, I was working at Lands' End in 1991. One of your Circulation Managers had a challenge. She had two segments of customers she mailed, and one she didn't. She knew how the mailed segments performed, but needed an estimate for the segment she didn't mail.

3 - Estimating segment performance for an un-mailed segment was easy ... I had done this work at the Garst Seed Company in the late 1980s. So I performed the exercise for her.

4 - And my results were accurate enough that I was asked to analyze other issues.

We had a division that mailed 2 catalogs in a quarter. What would happen to the two catalogs if a third catalog was mailed?

I recommended a strategy ... an A/B test.

5 - Say the first catalog generated $5,000,000 in sales and the second catalog generated $5,000,000 in sales. What would happen to the 2 catalogs if a 3rd catalog was mailed, and how much would the 3rd catalog generate?

6 - The results were kinda interesting.

Instead of generating $5,000,000*2 = $10,000,000, we generated $4,300,000*3 = $12,900,000.

The new catalog didn't really generate $4.3 million ... it took away $0.7 million * 2 = $1.4 million from the other catalogs.

7 - This meant that the new catalog was ($4.3 million - $1.4 million) / $4.3 million = 67% incremental. 33% of the sales of the new catalog were cannibalized from existing catalogs.

8 - This finding was really interesting. If this happened here, it might happen in many of our mailings.

In fact, the result replicated most of the time.

When we converted the results to profitability, we learned that we weren't making money on all of these new mailings.

9 - Well, this finding was controversial.

One faction of Leaders said the new catalog generated $4.3 million in sales.

Another faction of people, led largely by our test results, said the new catalog generated $2.9 million in sales.

10 - Eventually, I got promoted and led a team of 8 analysts. Most of my analysts studied test results. We consistently saw the same outcomes, and saw that our company could be more profitable if we followed a "data-driven" approach.

11 - There were a group of Executives responsible for business units ... those business units were responsible for mailing more catalogs to grow their sales. Naturally, those folks didn't want to mail fewer catalogs ... they wanted to mail more catalogs.

12 - Eventually, all this stuff came to a head. And on a Monday morning in December, the CEO that believed in my results was fired ... replaced by a new CEO from one of the business units that my team's test results suggested wasn't very profitable.

Oh oh.

13 - Within a month, the new CEO disbanded my team, moved 7 of my 8 analysts into the business units so they could support the business units in growth efforts (instead of analyzing test results that showed the business units weren't truly growing much).

I was finished!

14 - I likely deserved to be finished. I wasn't terribly kind about this issue, I thought the Leaders who disagreed with me weren't terribly smart.

So I deserved everything that I got.

15 - But at the time? That was a devastating career outcome!

16 - An interesting thing happened, however.

A Manager showed me a simulation he wrote (this was in 1995) in SPSS.

The simulation tool showed that if a business unit was not as profitable as it should be due to mailing too many catalogs, it could become profitable anyway.

17 - He showed me that if one of these business units invested in new customers, the new customers would eventually become loyal customers, and the loyal customers (even while generating marginal levels of profitability by being overmailed) would still be very profitable.

18 - The secret was to take some money away from mailing too many catalogs and invest it in customer acquisition ... then the new customers would become loyal and would generate a ton of profit.

19 - I internalized this strategy. It took me a full year to digest it, but I "got it".

20 - I left Lands' End for Eddie Bauer at the end of 1995 (you'd leave too if you were stupid enough to be caught up in the wrong end of a corporate coup).

And by late 1998, I was in charge of Circulation/Analytics. We had the same "over-mailing" issue.

21 - Except this time, I was in a position to do something about it.

I couldn't tell the EVP of the Home Division to mail fewer catalogs, but I could control "who" received the catalogs.

22 - So I told my team to mail all additional catalogs to prospects ... to individuals who had not bought from Eddie Bauer previously. We'd infuse the business with new customers who'd become loyal customers.

23 - The strategy worked! The Home Division achieved record profitability (a loss of $200,000 on an annual basis ... which the company celebrated as great performance ... but it was better than the $4,000,000ish that the division typically lost).

24 - I got to apply the results of the simulation at Lands' End in 1995 in real time in 1998-1999 and achieved much improved performance.

The simulation showed that the strategy worked.

25 - And ever since then, the strategy has worked.

If you want loyal customers in the future, you acquire customers today.

It's been the central theme of my Consulting work since early 2007. Clients have had considerable success following the thesis.

26 - And you'll have success as well.

If your company has customers with an average annual repurchase rate of < 40%, following a new customer thesis instead of the failed omnichannel thesis will yield considerable success.

27 - So give the concept some thought.

Could you invest money in new customers instead of bells & whistles designed to get existing customers to spend more, and as a result have more existing customers in the future generating profit for your brand?

Thanks,
Kevin