What we run
What we take over when we take the seat.
Most brands do not have an eCommerce problem. They have an ownership problem. The store reports to one person, Amazon reports to another, the agency reports to whoever booked the call, and nobody is accountable for whether the business actually grew.
We take that seat. We run five areas as one business, we own the P&L line they roll up to, and we hand it back working in three to twelve months. This page is what that covers, how we measure each one, and what is usually broken when we arrive.
01
The Shopify Storefront
We build, launch and replatform on Shopify, then run CRO, merchandising, site conversion and subscription. Including the retention program that decides whether acquisition pays back.
What we run
Theme and build, replatform from whatever you are on now, CRO and testing, merchandising and collection strategy, product page and checkout conversion, subscription, and the retention program that decides whether acquisition pays back.
What is usually broken
The site converts well enough that nobody looks at it. Meanwhile the collection pages are built around internal category names, the product page answers none of the three questions a first-time buyer asks, subscription exists but nobody has touched the cancel flow, and the email file is large and unsegmented. None of that shows up in a traffic report. All of it shows up in contribution margin.
How we measure it
Conversion rate by device and by traffic source, not blended. Revenue per session. Repeat rate and time to second order. Subscription retention by cohort. We report what a change did, including the ones that did nothing.
02
The Marketplaces
We run Amazon 1P and 3P, Walmart, TikTok Shop and Instacart: listings, content, PPC, retail media and operations. Run as part of the same business, not one competing with your site.
What we run
Listing architecture and content, A+ and brand store, search and PPC, retail media, promotions and deal calendar, inventory and replenishment, case pack setup, and the operational work nobody wants to own: suppressions, category approvals, chargebacks, and brand registry.
What is usually broken
Amazon and DTC are pricing against each other with nobody arbitrating. The pack you sell on Amazon is the same pack you sell on your site, so the customer picks whichever is cheaper that week and you lose either the margin or the relationship. The agency reports ACoS while total contribution goes sideways. Nobody has reconciled what a 1P chargeback actually costs against what 3P would cost instead.
How we measure it
Contribution after fees, ads, returns and chargebacks, by channel. Share of voice on the terms that matter, not all of them. Price parity exceptions, tracked weekly.
03
The Demand
We run paid media across Meta, Google, TikTok and Amazon, managed to contribution margin and payback rather than a ROAS screenshot. CRM, email and SMS, retention and lifecycle. Demand you create and demand you capture, priced as one budget.
What we run
Paid social and paid search, Amazon-sponsored placements, creative testing and briefing, audience and budget allocation across channels, Klaviyo email and SMS, flows and campaigns, segmentation, and the retention program that decides what a new customer is worth.
What is usually broken
Every channel reports its own attributed revenue and the sum is larger than the business. Creative is the actual constraint but the reporting is all about bidding. The email program is three flows and a weekly send. Acquisition and retention are budgeted by different people, so nobody can say what a new customer is actually worth or what it is worth paying to get one.
How we measure it
Contribution margin and payback window, not attributed ROAS. New customer cost against first-order and ninety-day contribution. Email and SMS revenue as a share of total, and what that costs to produce.
04
The Answer Layer
Answer engine optimization (AEO), generative engine optimization (GEO) and AI chat discoverability. When a customer asks ChatGPT, Gemini or Perplexity what to buy, we make sure your brand is in the answer. Most brands cannot tell you whether they show up today.
What we run
A baseline of what the assistants say about the brand today and who they recommend instead. Product and brand content rewritten to answer the questions buyers actually ask, with the schema markup and FAQ architecture that makes it citable. The off-site authority signals answer engines draw from: reviews, comparisons, press, retailer and marketplace listings. And an llms.txt and content layer built for machines as well as people.
What is usually broken
Nobody has asked ChatGPT, Gemini or Perplexity what they say about the brand, so nobody knows the answer is wrong, out of date, or missing. The product content that exists was written for a category page, not for a question. The signals answer engines trust are thin or inconsistent, and a competitor with a worse product is being recommended because their answer was easier to find.
How we measure it
Share of answer: how often the brand appears when the buying questions in its category are asked, tracked monthly across the major assistants. Which sources are being cited, so we know which assets are doing the work. Referral traffic and orders from AI surfaces, reported alongside organic search rather than hidden inside it.
05
The Unit Economics
We own the P&L line: contribution margin, pricing and pack architecture, forecasting, and agency and vendor management. These are the decisions that connect the other four.
What we run
Contribution margin by product, pack and channel. Pricing and pack architecture across DTC, marketplace and retail. Forecasting and demand planning. Promo economics. Agency and vendor management, including hiring them, briefing them, holding them to numbers and replacing them when they do not deliver.
What is usually broken
Nobody can say which SKU makes money after shipping, returns and ad cost. The promo calendar was built on what worked last year. The forecast is a growth percentage applied to last year's number and everyone knows it. Four agencies each have an explanation, none of them is wrong, and none of them is accountable.
How we measure it
One P&L line, one team answering for it. Contribution margin by SKU, pack and channel. Forecast accuracy, tracked and reported against what we said last month.
Why these five are one job
Anyone can leave you a plan. We leave you a business.
Pricing decisions made in the marketplace channel change what the storefront can charge. Pack architecture decides whether paid media can hit payback. The retention program decides what you can afford to pay for a new customer. The answer layer decides whether any of that demand finds you when a customer asks a machine instead of a search engine. Split these across five vendors and every decision routes through someone who owns one of the five and answers for none of them.
How the engagement works
Embedded. Temporary on purpose.
- The seat
- From as little as five hours a week, embedded. In your Slack, on your calls, in your dashboards.
- The term
- Three to twelve months. Temporary on purpose. If you still need us in year two, we'd take the compliment.
- Ownership
- We make the calls. We set the plan, control the spend inside an agreed budget, make the pricing and promo calls, and report the result without a filter. If every decision still needs your sign-off, you do not need an operator. You need a vendor, and not a strategic partner like us.
- The exit
- We hand back a business that is documented, staffed or vendored correctly, and forecastable.
- Three ways in
- Audit, to find out what is actually wrong and what it is worth fixing. Retainer, where we run one or two channels and own their results. We Own The Seat, where we run the full digital business with full P&L and vendor authority.
Questions we get
Asked on most first calls.
Can you run just one of these five?
Yes. That is the Retainer. We run one or two channels and own their results, and we will tell you on the first call if the problem you described is actually somewhere else.
How do you own a P&L at as little as five hours a week?
Because the job is decisions, not hours. Setting the plan, controlling the spend, making the pricing and promo calls, and holding the vendors to numbers is senior work, not volume work. The volume work belongs to the agencies we manage.
Do we have to fire our agency?
No. We manage the agencies you have. We brief them, hold them to numbers, and tell you plainly when one is not worth what it costs.
What is the Answer Layer?
Answer engine optimization (AEO), generative engine optimization (GEO) and AI chat discoverability: the work that determines whether your brand is in the answer when a customer asks ChatGPT, Gemini or Perplexity what to buy. It is a different discipline from traditional SEO, it compounds, and the assets you build are ones you own.
How fast do we know if this is working?
The Audit gives you a real answer in two weeks. Inside an engagement, the first things we change are usually pricing exceptions, budget allocation and the retention program, and those move within a quarter.
Show us your business and frustrations. We’ll show you how we correct it.
Send the store, the marketplace accounts and the ad accounts. In two weeks you get a real answer: what is broken, what it is costing you, and what is worth fixing first.
If nothing is worth fixing, we will tell you that too.
Show Us The Business together@M38partners.com
You will hear from us within one business day.