Back to PlaybooksPlaybook 01

Nine stages. Only one of them is your problem.

Go-to-market is not one activity. It is nine, running in sequence. At any moment exactly one of them is costing you everything downstream. Here is how to find out which.

Author

Ben Sheppard

Read time

3 prompts · about 45 min end to end · ends in a branded document you can save as PDF

Just here for the prompt? Skip to it →

01

The trap

When leads thin out, the usual answer is more activity. More posts, more outreach, a new channel. A few replies arrive, no deals close, and the work moves to the next visible thing.

Go-to-market is not one activity. It is nine stages running in sequence, and at any moment one of them is the constraint. Work on the other eight can be competent and still change nothing. The founder who doubles posting may need to fix capture. The founder who rewrites the site may need to fix qualification.

The practical error is treating activity as a system. A system has a named stage, a handover to the next one, and a measure that shows whether it works. Find the earliest break, fix that one, then see what moves. Fix three things at once and an improvement teaches you nothing. Fix the first break and the next one becomes visible.

03

The nine stages

Think of it as a conveyor belt. Each stage hands something specific to the next, and a break anywhere stops everything downstream.

12345 FoundationResearchContentCaptureOutreach Build your ICP: define who buys from evidence.02 ICP Your voice: set the rules for how you sound.03 VOICE PLAYBOOK 04 6789 QualifyConvertExpandTrack THE QUARTERLY LOOP: refreshed from real outcomes THIS PLAYBOOK FINDS THE STAGE YOU NEED Foundation has its ICP and Voice playbooks. Content has its calendar playbook.
  1. 1. Foundation. Who actually buys, mined from evidence rather than memory.
  2. 2. Research. Where your buyers already spend attention, and what the market charges.
  3. 3. Content. One core idea a week, argued properly, in your voice.
  4. 4. Capture. Everyone who engages ends up in one list, with what they actually said recorded verbatim.
  5. 5. Outreach. Warm, signal-grounded first contact: a message that could not have been sent to anyone else.
  6. 6. Qualify. A gate with three outcomes, not two: qualified, declined, and marginal.
  7. 7. Convert. Diagnose, then prescribe. Every call ends with a named action and a date.
  8. 8. Expand. The revenue you have already earned, and the truest measure of whether delivery is any good.
  9. 9. Track. One chain: piece → visit → capture → qualified → call → won.

Three rules hold the whole thing together. Evidence over invention: every claim comes from something a customer actually said, did, or paid. Automate the rhythm, never the judgement: the system gathers and drafts; a human decides what sends. Fix one constraint at a time: because the belt is literal, and the constraint moves.

04

The ten-minute version

Nine questions. Answer them honestly. The value of this exercise is entirely in how uncomfortable it is.

Tick the ones you can answer yes to.

Most founder-led businesses fail at question 1 or question 4. Almost everyone fails at question 9.

05

The evidence version

The ten-minute version relies on your own judgement, and founders are generous with themselves. These three prompts audit the same nine stages against what is actually in your systems, and they are instructed not to be encouraging.

The final prompt returns a branded Inside Startups document you can save as a PDF.

Stage 01 of 03

Census and register

About twenty minutes. Counts what exists, quarantines junk, sets the evidence checklist, and picks the calls worth reading.

Prompt locked

Enter your name and work email above to unlock and copy the full prompt.

Stage 02 of 03

The evidence

About fifteen minutes. Reads the nominated calls, searches email and chat by intent, and prices every incident the record shows.

Paste the evidence block from the previous stage below this prompt before you run it.

Prompt locked

Enter your name and work email above to unlock and copy the full prompt.

Stage 03 of 03

The document

About eight minutes. This is the branded-document stage, producing the audit from the evidence blocks alone: the bill first, then the constraint, named in one sentence and argued.

Paste the evidence block from the previous stage below this prompt before you run it.

Prompt locked

Enter your name and work email above to unlock and copy the full prompt.

Run the stages in order, then compare the document's answer to the boxes you ticked. Where the two disagree, the evidence is usually right.

Bring the diagnosis to a working session. We will find the constraint together and leave with the first fix written down.

Book the session: €220

06

Why it works

Because it makes the invisible stage visible.

Every business already does some of these nine. What almost none do is check them in order, in one sitting, and name the earliest break. The diagnosis takes ten minutes and routinely overturns a quarter’s worth of planning.

And because it is sequential, it compounds. Fix the foundation and the content gets sharper for free. Fix capture and outreach stops being cold. Fix the gate and the close rate rises without a single change to how you sell.

07

The next step

This page tells you where to focus next.

Stage 1 is your constraint? Build your ICP.

Stage 3? Thirty days of content.

Anything else? Book a working session. Bring the nine questions and your honest answers.

€220 for the hour.

Book the session: €220

Or take the nine questions to your team first. The diagnosis is more useful when more than one person answers it.

Do I need to have done Playbook 02 first?

No. This is the page that tells you whether you need it.

How long does the diagnosis take?

Ten minutes on your own for the checklist, longer and more usefully with your team. The prompt version takes about an hour.

Why do only two stages have playbooks?

Deliberately. The foundation and the content engine are the two stages you can run well on your own with a good prompt. The other seven depend on each other and on judgement calls that go wrong without context, which is exactly what the working session is for.