Quick followup to last week. I told you we're cutting our biggest paid acquisition channel mid May as part of a bigger campaign we call Project 500 Bras. €65 Meta CAC. Just can’t accept that number.

Before we made the call, I ran the same five numbers I run on every advisory brand that walks into my inbox. If three look weak, the move is to fix retention first, before any paid decision.

Here are ours, in the worst consumer year DACH has seen in 20 years. So be gentle, working on it!.

Most brands quote a single repeat rate and feel good about it. Good start but doesn’t give you everything you need. The curve is important. At erlich, our current repeat order rate looks like this: 30 days at 9%, 90 days at 14%, 180 days at 17%, 365 days at 21%. Not happy with that yet, we want to be back at 30% after a year (been there before the crisis got worse). But the trendline is going up since we made a bunch of good decisions over the last year.

The slope between 30 and 90 is the most diagnostic. If it's flat, your post-purchase experience isn't working. If it's steep, your second-order trigger (replenishment, complementary product, content) is doing real work. Pull from your Shopify customer reports. Print it out, pin it to your fridge (ok maybe not, thats weird).

This is what one acquired customer is worth in revenue on transaction one. Ours sits at €102 net, which is super strong for our category, really proud of that. Underwear has structural problems, especially in 2026 in DACH: low AOV (people just checking you out and buying one slip from you), returns für complex products like Bras, size and fit is super super important as it’s directly tied to comfort and if it’s not comfy people are not coming back (i wouldn’t either).

€102 means we cross and upsell on the first basket better than most. If yours is below €70 and you're paying lets say €40+ in CAC, fix the basket before you touch retention. Bundles, free-shipping thresholds, complementary SKUs highlighted before checkout. Easier win than reworking a Klaviyo flow architecture.

The one almost nobody tracks, and the one where the model lives or dies. What percentage of customers who place a second order go on to a third? At erlich, our second-to-third retention runs 3-4 percentage points higher than first-to-second. That compounding is what makes the entire model profitable. Once we get someone past order two, they keep going and the cohort is getting better. If your second-to-third number drops off a cliff versus first-to-second, you have a product or experience problem that no marketing budget will fix.

The most important number. Ours sits at 1.4 right now, in the toughest consumer year DACH has seen in two decades. That means for every euro we spend acquiring a customer, we earn €1.40 back over their lifetime. Healthy DTC is usually quoted at 3:1. We're nowhere near that. The market is nowhere near that, especially since the energy cost explosion this year and everything that came with it. Anyone selling you 3:1 in DACH apparel right now is either applying linkedin logic or sitting on a one SKU subscription business (socks anyone?). Build it manually if your analytics plattform doesn't show it (it should): CLV divided by blended CAC. Update weekly, super easy with claude even if you only have excel files.

The question every retention metric eventually has to answer: does order one cover its own acquisition? When first-order CLV minus CAC minus COGS minus all variable cost is positive, you're golden. You can and should scale paid. Almost no apparel brand is there. We have reached that stage again after a long journey in the last months. But: If your first-order isn't profitable yet your order-2-to-3 retention compounds and your 12 month CLV/CAC is positive, you can still run a profitable brand. You just need the liquidity to keep products flowing, which is a newsletter on its own.

Because i know the pain of being choked by meta/google and i want you to have the knowledge to balance out your business. We can run that risky 500 Bra Campaign and cut paid because four of the five look solid going in: our curve has an ok slope, first-order CLV is strong, order-to-order compounds and we reached first order profitability again. The one that didn't: CLV/CAC at 1.4 (we want north of 2.0). The cut is an experiment to fix CLV/CAC by pausing our most expensive new customer channel and replacing it with… something else i’ll tell you next week ;) .

If yours look the other way (CLV/CAC fine but the curve is flat and the order-to-order step drops), leave paid alone. The fix lives in product or post-purchase, and your Meta budget won't move it.

Run these five numbers before any big paid decision. They take a thursday afternoon if your data is clean, a week if it's not. Worth it either way.

Have a wonderful weekend everyone,

Philipp

P.S. If your numbers are not where they should be and you want a sparring partner to look at them with you, just hit reply.

If you enjoy what you’re reading, support my little side gig and share this newsletter with a friend using this link here below. Means the world to me.

As a thank you, i’ll send you my actual AI ops stack. The skills and workflows I actually use every day at erlich textil and for my advisory clients.

Nobody became a developer to write documentation. But the docs still need to get written — PRDs, README updates, architecture decisions, onboarding guides.

Wispr Flow lets you talk through it instead. Speak naturally about what the code does, how it works, and why you built it that way. Flow formats everything into clean, professional text you can paste into Notion, Confluence, or GitHub.

Used by engineering teams at OpenAI, Vercel, and Clay. 89% of messages sent with zero edits. Works system-wide on Mac, Windows, and iPhone.

Try Wispr Flow free