Why isn't our go-to-market working, and how do we tell which layer broke?

When a go-to-market stops working, the layer that broke is almost never the layer you're looking at. A go-to-market runs on five of them: who it's for, what you're the alternative to, the words that carry it, how you sell, and the proof you put in front of people. They fail downward. A crack in the second layer arrives at your desk as a problem in the fourth and fifth, because those are the only two layers instrumented well enough to complain.

That's why the response is usually a channel experiment, a new sequence, a pricing test, or two more reps. Each of those is a real fix for a real problem, just not for the problem you have. Six months later the numbers look the same and the team is more tired.

The diagnostic that finds the real break is cheap and takes about a day. Test the layers in order, from the top, and stop at the first one that fails. The first failure is the one to fix. Everything under it is a symptom you'd otherwise pay for twice.

What are the five layers of a go-to-market?

Every go-to-market, however it's written down, resolves into these five. They sit in dependency order, and each one inherits whatever the layer above it got wrong.

• Market. Which specific companies, at which moment, with which trigger. Not a vertical and a revenue band, but a describable situation a buyer is in.
• Position. What a buyer would do instead of buying you, and why you're the better answer for the situation in layer one. Your real competition usually includes doing nothing, which most positioning ignores.
• Message. The words that carry the position into a room, a page, and a deck. This is where a correct strategy quietly becomes an unrepeatable one.
• Motion. How the message reaches people and how a deal actually moves: channels, sequences, the sales process, the handoffs.
• Proof. Case studies, references, ROI math, security reviews. The evidence that makes a believable claim credible.

Failure travels one direction. If the market layer is wrong, your positioning is aimed at people who never had the problem, your message lands flat in front of them, your motion burns through a list that was never going to convert, and your proof describes customers your prospect doesn't recognize. Run that top down and it's obvious. Run it bottom up, which is how most teams experience it, and it looks like five separate problems appearing at once.

Why do we keep debugging the wrong layer?

Because of where the instruments are. Your motion layer has a CRM, a pipeline report, conversion rates by stage, and a weekly meeting built around it. Your proof layer has win-loss notes. Above those, the measurement thins out fast, and the market and position layers have nothing at all. No dashboard has ever shown a leadership team that their position is the constraint.

Call it dashboard gravity. Attention falls toward whatever is measured, and the measured layers sit at the bottom of the stack. A team with a broken position layer will reliably spend two quarters optimizing the two layers underneath it, because those are the only ones that produce a number to move. The work is diligent and the instinct is good. The altitude is wrong.

We ran into our own version of this, and the numbers are worth showing because they're ours rather than a case study we're allowed to describe loosely.

We track how AI engines answer buyer questions in our category. Across a 30-day window ending August 10, 2026, covering 4,010 chats and three engines, PitchKitchen showed up in 21.99 percent of answers on value proposition development, at an average position of 1.1. We win that room outright. In the same corpus, same window, same engines, we showed up in 8.91 percent of answers on go-to-market positioning, at an average position of 3.9.

Our first instinct was to look at the motion layer: publish more, publish faster. That would have been months of volume aimed at a break sitting two layers up. The actual problem was that our pages spoke about narrative identity while buyers in that room were typing go-to-market. Message layer. Nothing below it was going to fix it.

How do we run the diagnostic in a day?

Work down the stack and stop at the first failure. Each test is deliberately blunt, because a blunt test you'll actually run beats a rigorous one you'll schedule and cancel.

• Market. Pull your last 20 closed-won deals and your last 20 closed-lost. Can you describe, in one sentence, the situation the won accounts were in that the lost ones weren't? If the honest answer is company size, this layer is broken and everything below it is guesswork.
• Position. Ask your five best salespeople what the customer would have done if they hadn't bought. Five different answers, or four answers that skip doing nothing, means you're positioned against a competitive set your buyers don't actually hold in their heads.
• Message. Ask three people in three functions to explain what you do in one jargon-free sentence, separately, in writing. Compare them. Three different companies on the page means the message layer broke, whatever your strategy deck says.
• Motion. Look for the stage where deals reliably stall. A single stall point is usually a genuine motion problem. Stalls scattered across every stage point back up the stack.
• Proof. Ask whether your case studies describe the trigger from layer one. Proof that features a customer your prospect can't see themselves in is decoration.

Most teams that run this in good faith stop at layer two or three. That matches what we see in diagnostics: the strategy is usually sound and the articulation of it never survived contact with a live room. That specific failure has its own shape, and we've written about what makes a go-to-market narrative fall apart in live sales conversations separately. If the three-sentence test above came back messy and you want a sharper read on that one layer, how do I know if my B2B messaging is broken, not just underperforming? goes deeper on the difference between a message that's weak and one that's wrong.

What does a broken position layer actually look like?

It rarely looks like losing. It looks like winning slowly. Deals take three calls to explain rather than one. Buyers compare you to companies you don't consider peers. Your champion loves you and can't carry the argument into a room you're not in. Discounting creeps up without a pricing decision ever being made.

From the motion layer those all read as sales-execution problems, which is why the next move is so often more headcount. We've covered that specific fork in should we fix our messaging or hire more salespeople when the pipeline stalls?, and the related question of whether the slowdown is a story problem at all in is our slow growth a messaging problem or a product problem?.

There's a second tell worth checking now, and it takes four minutes. Ask ChatGPT or Gemini what your company does, then ask which companies a buyer in your category should consider. If the answers are generic, or shaped like a competitor, that's the position and message layers being read back to you by a machine that had no stake in the outcome. If you want that read done systematically against your homepage, our Brand Signal Score scores the narrative clarity, trust, AI-readability, and conversion of a single page against a 19-criteria rubric.

What do we fix first once we know which layer broke?

Fix the highest broken layer and leave the rest alone until it's done. This is the part teams resist, because the lower layers are where the deadlines live and someone has already committed to a campaign. Rebuilding a position while simultaneously running the motion built on the old one produces a quarter of contradictory activity, and the data it generates can't be trusted either way.

Sequence beats effort here. A corrected market layer changes who the position is aimed at, a corrected position changes every sentence in the message layer, and a corrected message changes the sequences, the deck, and the homepage. Work top down and each fix makes the next one smaller. Work bottom up and you'll rewrite the same assets three times, which is the quiet reason our product launches fall flat so often: the launch is the motion layer firing on top of a break nobody went looking for.

One more thing worth saying plainly. A broken upper layer isn't evidence that the strategy was stupid. Markets move, and a position that was exactly right at $3M in revenue frequently stops being right at $15M, because the buyer changed and the document didn't. The failure is rarely in the original thinking. It's in the absence of anything that forces the thinking to be revisited when the company underneath it changes.