Truth

Wave 1

Truth.

Every engagement starts here, and it starts the same way: before we build anything, we work out the numbers as they actually are. What a customer really costs you by channel, once returns and discounts come out. What a customer is really worth, by cohort, over time. Where margin goes between the checkout and the ledger. And where you show up, or don't, when a machine assembles the shortlist your buyers see. It all lands on one page, the founder signs it, and then we build.

We do it first because the numbers a growth plan stands on usually give a little when you press on them. A blended acquisition cost can hide one channel quietly paying for another. Churn counted at the cancel button gets counted weeks after the customer actually made the decision. Contribution margin that lives in a month-end spreadsheet isn't a number anyone defends in a meeting. That's what happens when a brand grows faster than its instruments, and you can't fix a leak you can't name.

The baseline is also how we earn an opinion about your business. We'd rather show you the read than ask you to trust the pitch.

The first deliverable is a page of numbers you'd sign.

Truth versus the waves next to it

Truth finds things and names them. The waves after it do the fixing and the keeping, and holding that line is what stops a diagnostic from swelling into a rebuild nobody asked for.

What good looks like.

Truth ships five things, and each one ends in a decision rather than a deck.

1. One page of numbers the founder signs

The baseline itself: your real CAC by channel after returns and discounts, contribution margin per order, repeat rate by cohort, the churn definition and the day it fires, and where you stand in the answers your buyers actually read. Ten numbers, maybe twelve, each with its source and its definition sitting right next to it.

The signature matters more than it sounds. Once the founder signs the page, every debate downstream gets shorter, because the team stops arguing about what's true and starts arguing about what to do. That trade alone is worth the wave.

2. The reads, each built to answer one question

The page comes together from small, scoped reads, each with a clock on it and a specific question to answer:

• The visibility read. Where you stand in the answer engines: how often you get cited next to named competitors, in the questions your buyers actually ask.
• The retention read. Where customers actually leave, when the signal first shows, and what a cohort is worth, all computed on your own order history.
• The cost-to-serve read. What an order really costs by SKU, warehouse, and carrier, and where the spread is widest.

Not every engagement runs every read. The problem you walk in with picks the first one, and the first one usually finds the second.

3. The skip list

Every read ends with a list of things not worth doing at your stage: the attribution suite you don't need until your definitions settle, the personalization platform that's premature before the basic flows exist, the replatform that's really a configuration problem. The truth about how to get where you're going includes the parts to skip, and we'd rather tell you now than sell them to you later.

4. The holdout habit, declared here

From this wave on, anything we ship gets measured against a holdout: a group of customers who didn't get the change, so the lift is proven instead of asserted. We declare it in wave one because it disciplines everything after. A team that knows its work will face a control group scopes differently and ships differently, and that includes us.

5. A cadence, because numbers drift

The baseline isn't a one-time event. Channels shift, cohorts age, the answer engines re-rank, so the page of numbers gets re-read on a schedule your team keeps, and any plan that contradicts the current page has to say why.

How it runs. Picture a growth-stage, subscription-led brand, we'll call it Marlowe. Growing, but straining: the founder can quote revenue from memory but acquisition cost only as a blend, and churn means "cancelled," counted weeks after the customer went quiet. Truth redefines churn on order gaps so it fires while there's still time to do something about it, splits acquisition cost by channel and takes it after returns, and puts one finding on the table nobody expected: the channel everyone had been celebrating looks very different once returns come out of it. The growth plan changed before anyone built anything.

The threads, in this wave.

Woven in.

The models are what make a fast read possible at all: scoring order history, tracking citations across the engines, reconstructing margin from messy exports. They run during the read without making an appearance, and they're also where the discernment shows. The skip list exists because we've watched brands buy instruments before they had definitions.

Your team.

The baseline is written so your team can re-run it without us, every number with its source, every definition in one place you own. Months from now, when someone asks what your real CAC is, the answer shouldn't be "ask the agency." It should be a page your team keeps current.

The coda.

You can't grow a number you can't see. Get the truth down first, and everything after it moves faster.