From your question to a decision you can defend

Seven steps. Not every decision needs all of them and the smaller ones move quickly, but this is the shape of the work.

  1. 1

    Frame

    Define the decision that's actually being made. What you're optimizing for, which levers you control, what's constraining you, and which unknowns could move the answer. A lot of the value shows up here, because plenty of decisions look hard until someone writes down what is really being chosen between.

  2. 2

    Model

    Turn the business economics into something quantitative, connecting the decision to its likely consequences. A hire becomes ramp time and capacity and pipeline conversion. A price change becomes retention and mix and collected cash.

  3. 3

    Forecast

    Project what happens under each course of action. Usually that means ARR, cash, runway and profitability, plus whatever operating measures the specific decision turns on.

  4. 4

    Optimize

    Look for a better combination than the options you started with. This is the step most scenario analysis skips. The strongest answer often isn't Plan A, B or C, but something assembled out of pieces of each.

  5. 5

    Stress test

    Try to break the recommendation. Slower growth, rising acquisition cost, churn drifting up, a hire that takes longer to ramp. A recommendation is more useful once you know how easily it falls over.

  6. 6

    Recommend

    Say what I'd do and why, along with the risks and the assumptions the answer leans on. If the honest answer is that it's close either way, that's what you get told.

  7. 7

    Monitor

    Name the numbers and the thresholds that would change the call. A decision that's right in March can be wrong by September, and you want to hear that from a tripwire rather than in hindsight.

Why the optimize step matters

Scenario analysis usually compares a handful of plans somebody wrote down in advance. Hire two people. Push marketing. Sit on the cash. Those comparisons are useful as far as they go.

The better answer is often a combination nobody put on the list. Hire one person in April, put another $75,000 into the channel that's still paying back, push the product work to Q3, and hold runway above fifteen months. That plan never appears if you only score the options you started with.

What's involved on your end

It starts with a conversation about the decision you're contemplating. Nothing to prepare for that one. I want to understand what you're weighing and what's riding on it before either of us talks about data.

What I need after that depends on the decision. Some questions are answered from exports you already have. Others need a real pull, like customer-level subscription history or contract terms going back a couple of years, and that can take some work on your side to assemble. I'll tell you what the request looks like before you commit, so you can weigh the effort against the question.

Where I can work from what already exists, I will. Your books don't need to be clean first, and I'd rather see them as they are than wait while someone tidies them up.

You don't need perfect data

Good decisions need evidence. They don't need you to pretend every input is known exactly. Where something important has never been measured, it goes in as a range rather than a guess dressed up as a fact.

Rather than assuming acquisition cost is $8,143, we might carry $7,000 to $10,000 and ask whether the recommendation survives the whole span. If it does, you can act on it. If it flips somewhere in the middle, that tells you which number is worth going and measuring properly.

The goal isn't precision. It's knowing whether the answer holds up.

What you end up with

A recommendation you can act on, with the reasoning attached. Alongside it you get the options priced out against each other, an honest look at what happens if things go badly, the point at which the answer would flip, and the handful of numbers worth keeping an eye on afterwards.

The model is the evidence. The deliverable is the decision.

What I don't do

I'm not your bookkeeper and I'm not your accountant. Bookkeeping, tax filing, audit work and system implementations stay where they are, and I work alongside the people doing them. I'm also not trying to replace your forecast. If you have a good one it becomes an input.

Questions people ask

How is this different from traditional FP&A?

FP&A gives you the foundation. Reporting, budgeting, forecasting, knowing where you stand. This builds on that to evaluate a specific decision. The question changes from what is likely to happen, to what we should do given our objectives and what's actually available to us.

We already have a forecast. Does that make this redundant?

No, it usually makes it faster. A decent forecast becomes an input rather than something I rebuild. The extra layer is taking the choices in front of you and working out which one the forecast actually supports.

I don't have clean books. Is that a problem?

It's the normal starting point. Part of the early work is figuring out what your numbers are telling us and where they need a second look. You don't need to tidy anything before we start.

What will you need from me?

Less than you'd think, and it depends on the decision. I work out what has to be modeled and ask only for the evidence that needs. A couple of years of monthly history helps on the bigger questions, but plenty of decisions work with much less. Where something has never been measured we use a range instead.

Is this just about cutting costs?

No. The objective might be growth, cash, profitability or long term value, and you set it. Sometimes the analysis says spend more. Sometimes it says the thing you were about to fund is fine but the timing is wrong.

Do you do bookkeeping or tax?

No. I work alongside your bookkeeper and your accountant rather than replacing either of them. They keep the records right. I use them to answer questions about what to do next.

How much of my time does this take?

Not much. You forward access and exports at the start. We talk at the beginning and again at the end, and I handle everything in between. There's no homework and nothing to prepare for calls.

What if the answer is that my idea is bad?

Then you've saved the money, which is most of the value. An analysis that only ever agrees with you isn't worth paying for. I'll show you where it breaks and what would have to be true for it to work.

What does it cost?

It depends on the decision and how much modeling it needs. We'll talk through it on the call, because the right scope looks different for a single hiring question than for a full operating plan.

Got a decision in front of you?

Tell me what you're weighing and I'll help you reach the decision that works best for you.

Discuss your decision