I built the tool I needed to run two businesses — and help run a few others.

My name is James. I've been running ecommerce businesses for nine years. I've made every ad mistake there is, hired agencies, fired agencies, built spreadsheets I was proud of — and then watched those spreadsheets become the problem. Along the way I started helping other founders with their numbers. That's when it became obvious: the platforms are not your advisors. They're your media buyers. And a media buyer doesn't care about your margin.

Helping others was brutal on my time. Every store needed its own export, its own pivot table, its own moment where I'd stare at a screen trying to remember why a campaign that looked great last Tuesday was bleeding money today. I'd get deep into the data — genuinely deep, the kind of hyperfocus that means two hours later I've built a beautiful analysis that answers a question nobody asked. The weed-pulling. The rabbit holes. I know them well.

What I actually needed — for my stores and for the founders I was helping — was something that could do the daily read-across for me. Pull the Shopify numbers. Pull the Google numbers. Pull the Meta numbers. Reconcile them into one version of truth. And then tell me the one thing I need to act on today.

Not a list of 40 things. One thing. With context.

The problem with every tool I tried — and I tried most of them — is that they start from the wrong place. They show you platform data back to you in a nicer font. Meta shows you Meta numbers. Google shows you Google numbers. Neither platform has any incentive to tell you that you're over-indexed on one channel, or that your true marketing efficiency ratio is actually 1.8x, not the 4.2x ROAS they're reporting.

I spent four months in a server-side tracking rabbit hole trying to solve attribution properly — better signals, cleaner data, finally know exactly which channel drove which sale. We were mid-Black Friday when I made the call to walk it all back. Four months of work, reversed in an afternoon.

What I came away with isn't a better attribution model. It's a conviction: attribution is consistently inconsistent. The platforms will always disagree. The last-click model will always lie. Meta and Google have been arguing over who drove that sale since the first pixel fired, and they will be arguing about it long after we're gone.

Let Meta chase Meta attribution. Let Google chase Google attribution. Your job is to know your number — how much you can spend to acquire a customer and still make money — and then trust the platforms to do what they're actually built to do: find more people who look like your best buyers. Stop fighting them. Feed them.

The dashboards that claim to solve this usually pull it together visually, but the numbers underneath are still wrong. They use the revenue figure the platform reported, not Shopify's actual settled orders. They use your supplier price as the margin input, not your real landed cost per unit. They don't account for discounts on individual orders — so their margin looks right on full-price days and silently wrong on any day you run a promotion. They don't know which orders are new customers and which are repeat. They don't factor in your COGS, your fixed overhead, or what a customer is actually worth to you over time.

So when they tell you your MER is 40%, what they really mean is: based on approximate numbers that exclude most of your costs, you're profitable. Which is almost the same as not knowing.

I'd been asking this question for nine years across two businesses. I never found a tool that answered it cleanly. So eventually I built one.

I'm not a software company founder. I'm an ecommerce founder who learned enough to build what he needed. That's important context — because everything in Merchint exists because a real store needed it, not because a product manager thought it would be a nice feature.

The profit dashboard exists because I needed to know, in real time, whether the money coming in from Shopify was actually covering the money going out to ads plus the cost of goods. The frozen historical costs exist because I got burned once when a supplier changed a price and retroactively made three months of profitable-looking reports into lies. The new customer detection exists because I learned the hard way that a store can look like it's growing when it's really just churning.

And the AI daily action — the one recommendation you get each morning — exists because my brain does not handle a dashboard full of levers well. Give me ten things to optimise and I'll end up changing all of them, forget which one I changed, and then have no idea what moved the needle. One action. That's it. That's the whole design.

The inventory intelligence exists because I held cloth nappy stock for nearly three years through the post-COVID demand collapse. Every month of holding was capital that couldn't move into activewear — a market that was growing while nappies were collapsing. Clearing at cost, even at a per-unit loss, would have freed that capital years earlier. I didn't have the signal to see it clearly enough to act. The daily action isn't always "increase your ad budget." Sometimes it's: your sell-through rate on this category has been flat for six weeks — a promotional email will move it faster and cheaper than any campaign you could run right now.

Growing an ecommerce business is a capital allocation problem. Profit flows into three decisions every day: inventory to order, ads to run, and stock to clear. Most tools only help you with the middle one.

Here's what I've realised after nine years: the founders who win at paid ads aren't the ones who are best at Meta or Google. They're the ones who understand their numbers well enough that the platforms work for them, not the other way around.

Every paid ad you run teaches the algorithm something about your customer. If you're running the right ads with the right budget ceiling — one grounded in your actual margin, not a gut feel — you're compounding that learning. You're feeding the algorithm signal that makes the next campaign cheaper. You're not fighting the platform. You're training it.

The founders who don't have that clarity are always in reactive mode. Pausing campaigns because the ROAS dipped. Restarting campaigns because it came back. Interrupting the algorithm's learning cycle every time they panic. Paying a premium — in both cost and time — to stay exactly where they are.

Merchint is designed to break that cycle. Give you the number. Give you the context. Give you one clear action. And then get out of your way so you can run your business.

This isn't a venture-backed analytics company trying to own the ecommerce data stack. It's a tool built by a founder, for founders. And that's exactly what it'll stay.