Stop the guesswork. Build a forecast that gives the team something real to aim for and tells you early when something needs fixing.

Most ecommerce forecasts are wishful thinking. An optimistic number on a spreadsheet nobody believes or a hundred P&L line items that nobody looks at. And neither one helps you make the calls that matter.

So what does a good one look like? It's built on the way your business runs, it ties to the levers you can pull and every week it tells you whether or not you’re where you should be to hit the year end goal.

Why this makes a difference

Without a firm budget and forecast you're guessing. Stock's being ordered on a hunch. A hire's getting approved off the back of one good month. The media budget gets cut or doubled on the last fortnight instead of the trend sitting underneath it. And how many of those were decisions based on quantitative data?

But with one, the whole business settles down a bit. And that's worth more than people think. You know what to expect. The team knows what they're aiming at, and they're not guessing. Your targets turn into numbers the team can hit. And if there's a board, they stop asking you the same question every month.

What I see is that founders already know the forecast matters. They just haven't had the time, the framework, or somebody sitting next to them who knows how an online retail business behaves week to week, month to month, year on year.

How I think about forecasting

A forecast's a planning tool. Built off real history, averages, probability, and the levers you have control over. Being right about every month isn't the point. It’s having a target the team can plan against and a model where you can see things shifting early enough that you can still do something about it.

A few principles I work to:

  • I'd build it from history. Your last six, twelve or twenty-four months of trading will tell you more about the ebbs and flows of your business than any other tool - especially when paired with a well directed Ai tool.
  • I look at the drivers underneath the number. Sessions, conversion rate, average order value, repeat rate, media spend, organic growth... revenue's just the result of all those working together. The drivers are where the decisions live.
  • And then I'd make sure it’s used. A good forecast is a map of where we’re going and where we should be right now, today, this week in order to get there. It's a document open on your screen at the morning standup or weekly trade meeting.
  • Best, base, worst. I'd build all three. If the year starts going pear-shaped, what's the first thing you'd cut? Better to have answered that one at the start of the year.
  • Refresh it, don't rebuild it. The forecast is not an accounting function, it’s an operations tool. While the budget remains the baseline, a good forecast gets updated as the year goes. You learn, and it learns with you. That's the whole point of it.

AI's a big part of how I build and manage these. Data cleanup, scenario modelling, spotting patterns across categories that used to take days by hand. The model's still designed by a person who understands the business, and I'd be careful of anyone telling you otherwise. But the work behind it moves a lot faster than it used to.

What working with me looks like

It'll start with a clean look at how you've traded. Revenue by channel, margin by category, what the media's actually doing for you, all that sort of stuff. Then I'd build the framework, usually a working spreadsheet or a connected dashboard, that the team runs on week to week.

Then the harder part, and it's the one that decides whether any of this was worth doing. Getting it used. The forecast's the thing the weekly trade meeting runs off. Where it's off gets talked about out loud, and that's the interesting bit. Decisions get tied back to the model, so you're not arguing from memory. The budget stays where it is and the forecast moves as the year does.

And for brands without a finance person I'll run the cost side too. Where the profit's actually coming from, what your numbers are doing month to month, an easy to read one-pager for the board. Enough to give you real confidence in your numbers without hiring a finance team you're not ready for yet.

Who it's for

Good brands with good products, where the year still gets planned on instinct. Usually $3M to $20M. Brands heading into a raise, a big stock commitment or a new category who need to know what the year really looks like. And founders who're tired of being surprised at the end of the month.

Who it isn't for

Anyone wanting a pretty three-year model for an investor pitch. It's a working document. It's built from how you've really traded and it's designed to be opened every week.

Let's build a forecast that gets used

If the planning side of the business isn't where you want it, get in touch. The first conversation's free.

It'll usually start with a look at how you plan the year now, and what would have to change for that forecast to actually shape a decision.

But in the meantime, have a look at last year's plan. How close did it get? And more to the point, when did you first know it was off?

Contact me

Key points

  • I'd build it from your trading history, not from market guesses
  • Driver-based: sessions, conversion, average order value, repeat rate, media spend, all that sort of stuff
  • It's got to run your weekly trade meeting, not sit in a folder
  • Best, base and worst, built in from the start
  • AI does a lot of the building and the refreshing. The design's still a person