Thread by @minethatdata

Jacob asks a wonderful question. We like Jacob!

Allow me to explain a situation that keeps popping up.

https://twitter.com/JacobLDavis/status/1230637937290813441

1 - Let's say that you acquire 100 customers via branded search terms that you paid for.

You spent $2,500 on 5,000 clicks. Each purchaser generated an AOV of $100. 35% of sales flow through to profit.

Profit = 5,000*0.02*100*0.35-$2,500 = $1,000.

Profit/Cust = 1000/100 = $10.

2 - On the surface, you made a thousand dollars on your campaign. You did good! And you made $10 profit for every customer you acquired. That's also good.

3 - Your CLV / LTV metrics tell you that the branded search customer (paid) is worth $15 profit in year one, $10 profit in year two, and $5 profit in year three.

So far, so good.

4 - Of course, you have CLV / LTV for all of your sources ... and you know that any search customer (branded or not) is not worth as much as are customers from other sources. So while your effort is profitable, it's not as profitable as other sources.

5 - Search buyers tend to have lower CLV / LTV because they are buying for a need they have right now ... they're different ... oftentimes seasonal buyers without commitment to more frequent orders.

6 - But there's another big issue at play.

If you didn't pay for the branded term, you cannot assume that you wouldn't get the customer to purchase.

Often, the customer still purchases ... search is just part of a "journey". Often, the cannibalization rate is +/- 50%.

7 - In other words, if you didn't pay for the branded term you'd still get 50% of the orders you get if you do pay for the term.

Here's where life gets interesting.

8 - If we run the p&l assuming that half of the orders would have happened anyway, then we have to apply full cost against half the orders.

Ready?

Profit = 5,000*0.02*100*0.35*(50%)-$2,500 = ($750).

Profit/Cust = ($750)/50 = ($15).

Oh oh.

9 - In other words, since half of the orders would have happened anyway, we learn that we actually lost $750 on the campaign, and we actually lost $15 per new customer ... chewing up a full HALF of future profit.

This turns out to be a big mistake.

A BIG mistake!!!!

10 - This is what gets Executive Teams stirred up.

The ones who understand the relationship between short-term profit (i.e. measured via Google Analytics), cannibalization (measured via attribution / tests), and long-term value ... those folks get very frustrated.

11 - They get frustrated that there isn't one set of metrics available ... they get frustrated that the vendor loudly advocates for spending money ... they get frustrated that marketing has short-term metrics (positive) and analytics has cannibalization data and LTV (negative).

12 - They get frustrated that "camps" emerge ... vendor-aligned employees advocate for spending money, analytics folks think everybody is dumb, and those who measure CLV / LTV want money invested elsewhere.

13 - They get frustrated that there isn't a holistic view of the future that shows what the business looks like 5 years from now if investments are made or not made ... heck, sometimes losing money today is a GOOD idea to protect the health of a business (ask Amazon).

14 - So that's the kind of stuff that, when I sit in an Executive Meeting, I get to see first-hand.

None of this stuff is easy, of course.

Trying to paint a holistic long-term story of what the business "will" look like if you make certain decisions helps alleviate tension.