Latest / MBA Training Marketing / Customer lifetime value: fix it by purchase cycle
Transcript
- 0:00Host: Welcome to the Leaders Insights Podcast.
- 0:08Host: Today's episode, calculating customer lifetime value when purchase cycles vary by category.
- 0:13Host: A furniture retailer told me last month they'd been valuing their customers using the same
- 0:17Host: lifetime value formula as their online grocery arm.
- 0:21Host: Someone who buys a sofa every eight years scored like someone buying milk every Tuesday.
- 0:26Host: How badly does that break?
- 0:28Expert: Catastrophically and quietly, which is the worst combination.
- 0:33Expert: Lifetime value, or the total profit you expect from a customer across the whole relationship,
- 0:39Expert: collapses the second you average across categories with different rhythms.
- 0:44Expert: That mattress buyer looks like a churned debt account by month three because they haven't
- 0:48Expert: come back.
- 0:49Expert: They haven't come back because nobody buys two mattresses in a quarter.
- 0:53Host: So the number isn't just wrong, it points you the wrong direction.
- 0:57Expert: Exactly the direction that costs you money.
- 1:00Expert: You look at the mattress customer, see no repeat purchase, and you flood them with win-back
- 1:04Expert: discounts.
- 1:05Expert: You're bribing someone who was never leaving.
- 1:08Expert: Meanwhile, your grocery customer, who's genuinely slipping, buying every two weeks instead of
- 1:13Expert: weekly, gets no attention because the blended average smooths their decline into invisibility.
- 1:20Host: Walk me through what category-aware actually means in practice, because it sounds like
- 1:24Host: an excuse to build 12 spreadsheets.
- 1:27Expert: It starts with one honest question per category.
- 1:30Expert: What's the natural purchase cycle?
- 1:32Expert: Grocery, days?
- 1:33Expert: Seasonal clothing, twice a year, spring and autumn?
- 1:36Expert: Mattresses, once a decade if you're lucky.
- 1:39Expert: You define churn, a customer walking away, relative to that cycle, not a fixed 90-day
- 1:45Expert: window.
- 1:46Expert: A grocery shopper silent for a month is a fire alarm.
- 1:49Expert: A mattress buyer silent for a year is Tuesday.
- 1:52Host: And the retailers with mixed baskets, someone buying groceries in a coat in a mattress from
- 1:57Host: the same brand.
- 1:58Expert: That's where it gets interesting and where most models cheat.
- 2:02Expert: You don't compute one LTV for that person.
- 2:04Expert: You compute a value stream per category and sum them each with its own cycle and its
- 2:09Expert: own churn definition.
- 2:11Expert: MIT Sloan Management Review has been decent on this, treating the customer as a portfolio
- 2:17Expert: of category relationships rather than a single lump.
- 2:21Expert: The coat relationship can be healthy while the grocery one is dying.
- 2:25Expert: A blended number hides that entirely.
- 2:28Host: Give me a figure.
- 2:29Host: What is getting this wrong actually cost?
- 2:31Expert: The retention economics are brutal.
- 2:34Expert: The oft quoted line is that lifting retention 5% raises profit by 25 to 95%, that traces
- 2:41Expert: back to Bain research from the 90s, still cited because it's roughly held up.
- 2:47Expert: On tooling, SEMrush pegs a chunk of retail marketing waste at misfired re-engagement
- 2:52Expert: campaigns.
- 2:53Expert: But flag it.
- 2:54Expert: They sell analytics software, so they're incentivized to make your targeting look broken.
- 2:58Expert: Cross check against your own campaign logs before you believe the scale.
- 3:02Host: Let's be uncomfortable.
- 3:04Host: Is this just complexity for its own sake?
- 3:06Host: A small retailer with three product lines.
- 3:09Host: Do they need this?
- 3:10Expert: If those three lines have different cycles, yes.
- 3:13Expert: And it's not hard.
- 3:15Expert: You're not building a machine learning model.
- 3:17Expert: You're grouping products by how often people buy them.
- 3:20Expert: Three buckets.
- 3:21Expert: Fast, seasonal, rare.
- 3:23Expert: The failure isn't insufficient math.
- 3:25Expert: It's applying one average where three would do.
- 3:28Host: And the small retailer feels this harder.
- 3:30Host: They can't afford to discount a customer who is going to buy anyway.
- 3:34Host: Where does the standard approach break most invisibly?
- 3:37Host: The thing nobody catches.
- 3:39Expert: The seasonal category, because it fakes churn twice a year.
- 3:43Expert: Your swimwear customer disappears every October.
- 3:46Expert: The naive model marks them lost, triggers a discount, and you've just trained a loyal
- 3:51Expert: seasonal buyer to expect a coupon before every summer.
- 3:54Host: You taught them to wait.
- 3:56Expert: That's revenue you set fire to, trying to save a customer who was coming back regardless
- 4:01Expert: of what you did.
- 4:02Host: So what's the one thing a listener does Monday morning?
- 4:04Expert: Pull your product catalog and tag every line with its natural repurchase interval.
- 4:09Expert: Days, months, years.
- 4:11Expert: Just that column.
- 4:13Expert: Then look at your churn rules and see how many categories you're judging against a single
- 4:16Expert: window.
- 4:17Expert: I'd bet it's most of them.
- 4:19Expert: That one column tells you exactly where your lifetime value numbers are lying to you, and
- 4:24Expert: you can fix it before you send another wasted discount.
- 4:27Host: The lie in the average.
- 4:28Host: Good place to end.
- 4:30Host: Sources for today's episode.
- 4:32Host: SEMrush vendor.
- 4:33Host: SEO analytics tools.
- 4:35Host: DigiDay MIT Sloan Management Review.
- 4:38Host: That's your five.
- 4:39Host: The full CMO track, structured no fluff, is at MBA-training.com.