Nothing about good go-to-market has changed. Know your customer, reach them efficiently, convert reliably, keep them, and run the whole thing on data with a rhythm. Every one of those is as true as it was a decade ago. What has changed, completely, is what it costs to act on them — and because execution has become dramatically faster and cheaper, the distance between companies running a sharp commercial engine and companies that are not is widening quickly.
It has never been easier to build impressive things that do not matter. Choosing what to build, and why, is the new scarce skill.
Two wrong reactions, both expensive
Most commercial leaders currently hold one of two positions, and both cost money.
Panic. Everything is about to be disrupted, the sales team is obsolete, and the answer is to buy something quickly. This produces licences nobody adopts and a budget line that renews out of inertia.
Dismissal. It is hype, the fundamentals are unchanged, and we will wait for it to settle. The first half of that is correct, which is exactly what makes the conclusion wrong.
The fundamentals genuinely have not changed. The execution of them genuinely has. A task that required a procurement cycle, an integration project and a six-figure commitment can now, in many cases, be built in days for a few thousand pounds. That does not change what good looks like. It changes how quickly a company can close the gap to it — which means the cost of not closing it has gone up.
Has your cost of customer acquisition moved in the last twelve months, up or down? The interesting finding is rarely the direction. It is how many teams cannot answer at all, which is a data problem rather than an AI problem, and it is the more urgent of the two.
Look upstream before you build anything
Ask a room of commercial leaders where they look first when growth stalls, and almost everyone says sales execution. Across more than a hundred commercial reviews, that is usually the wrong place. The root causes cluster upstream: proposition, ideal customer profile, and product-market fit drift.
This matters more in an era of cheap building than it did before, because cheap building makes it easy to industrialise the wrong thing. Automating outreach to a customer profile that has quietly stopped buying does not produce growth. It produces more efficient failure, at higher volume, with better reporting.
Three patterns recur, all anonymised from real reviews:
- The buyer profile changed underneath the company. The person who signs moved to a different function with different priorities and a different budget cycle. The pitch still addresses the old buyer, who no longer decides.
- The product was commoditised by free alternatives. What used to be worth paying for became adequate and free inside something the customer already owned. Value did not disappear; it moved.
- A subscription model that was right for a market of five accounts. The monetisation model fitted the market the company started in, and nobody revisited it when the market it now sells into turned out to look nothing like that.
None of the three is fixed by a tool. All three are found by asking the first of the five questions before the third.
Five questions before any build decision
A full commercial assessment runs to twenty-three topics, which is the right level of resolution for a diligence exercise and the wrong one for a leadership team deciding where to spend a quarter. It collapses to five questions that have not changed in decades and will not change soon.
| Fundamental | The question | What cheap execution unlocks here |
|---|---|---|
| 1 · Proposition, Pricing & PMF | Do the right customers still want what you sell, at that price? | Continuous signal on how buyers describe the problem, rather than an annual research exercise |
| 2 · Marketing & Demand | Do the right buyers know you exist? | Reaching a precisely defined audience without a media budget |
| 3 · Sales Execution | Do you convert interest into revenue? | Preparation and follow-up compressed, so selling time goes to selling |
| 4 · Retention & Expansion | Do you keep and grow what you win? | Early warning from usage and conversation data that nobody previously had time to read |
| 5 · People, Data & Cadence | Is the engine run with owners, data and rhythm? | One version of commercial truth, assembled automatically rather than by hand each month |
Run a build idea against these before funding it. If it does not attach to one of the five, it is a project rather than a priority, however impressive the demonstration.
Three things that are now buildable in days
Not tool recommendations — the specific products will change within a year. They are worth describing because each demonstrates a category of thing that used to require a budget cycle and now does not.
Reaching a conference audience without attending
Speaker and delegate lists are frequently public. Combined with a sequencing tool costing tens of pounds a month, an entire event audience becomes addressable. Set that against the five-figure cost of a seat at a major industry event and the comparison is uncomfortable — particularly for anyone who has spent that budget for years.
A commercial cockpit assembled rather than procured
Market-signal scanning, full touchpoint history, and drafted follow-ups that know what was discussed last time, in one place. Built in days. This is where the old best-of-breed advice — buy the best tool for each job and integrate them — has genuinely stopped holding, and it is worth saying plainly by anyone who used to give that advice.
A weekly five-minute briefing nobody has to write
Every customer conversation summarised automatically into a short internal briefing for the whole company. The value is not the summarising. It is that the rest of the business hears what customers are actually saying, weekly, without depending on anyone finding time to write it up.
The common thread is that each addresses a fundamental rather than a tool category, and each costs days and thousands rather than quarters and millions.
Four moves, in order
Knowing what is possible changes nothing on its own. Four moves convert it into something that happens, and the order matters more than any individual step.
- Own it. One gap, one number, one name. Not a committee, not a transformation programme. The most common reason this work stalls is that it belongs to everybody.
- Sequence it. Upstream first, and two things at a time. In practice most businesses complete one or two initiatives in a year unless somebody is dedicated to running the plan, so a list of nine is a list of seven disappointments.
- Budget it. Days and thousands, not quarters and millions. The cost structure has changed and most approval processes have not caught up, which means work that would sail through as an experiment gets treated as a capital project.
- Resource it. Decide what to keep in-house, what to grow, and what to bring in on demand. This is a design question rather than a hiring question, and it has changed shape in the last twelve months.
Underneath all four sits the unglamorous foundation: clean commercial data. Nobody gets budget for a data cleanup, which is why the practical route is to sequence a couple of early, visible wins first and earn the right to fix it.
The scarce skill moved
A year ago the constraint was technical. Knowing which tools existed, how to connect them, how to make the integration hold. That skill has become dramatically more available, and its scarcity value has fallen accordingly.
What is scarce now is domain expertise and hard critical thinking. AI output sounds right. Whether it is right takes judgement, and judgement comes from having watched enough commercial engines to know which answers are plausible and wrong.
This has a direct consequence for how teams are designed. The diagnosis is a joint exercise: commercial leadership sets the priority, builders make it real. Neither works alone. A builder without commercial judgement produces impressive things that do not matter. A commercial leader without a builder produces a list of things that never get made.
For anyone hiring into a commercial function this year, that is the shift worth internalising. The question is no longer whether someone can build it. It is whether they can tell you why it is worth building.
Why it pays, in the language of an exit
For anyone operating inside a hold period, this connects directly to the equity story, and the connection is sharper than it used to be.
Growth is the numerator. The machine is the multiple. Buyers increasingly pay for how growth is produced, not only how much of it there is. A business growing at the same rate as a competitor, but doing so through a repeatable, instrumented, well-owned commercial engine, is a different asset — and is priced as one.
Timing matters. This work compounds, so it is worth most when started early in the hold. Inside two years of an exit the filter sharpens considerably: the only things worth starting are the ones that will be visible in the numbers within twelve months, because anything unproven at the point a buyer underwrites it gets discounted rather than credited.
One thing to do on Monday
Not a programme. One action.
Find your reddest fundamental of the five. Put one number on it and one name against it. Then get your commercial leads and whoever builds things in your business choosing together, rather than in sequence.
That last part is the whole argument compressed. The gap between companies with a sharp commercial engine and companies without one has never been more affordable to close. It has also never been easier to spend a year closing the wrong one.
Frequently asked questions
Has AI changed the fundamentals of go-to-market?
No. Knowing your customer, reaching them efficiently, converting reliably, retaining them and running the engine on data are unchanged. What has changed is the cost and speed of acting on them, which widens the gap between businesses that execute well and those that do not.
Where should a commercial team start with AI?
Upstream, with diagnosis rather than tooling. Across more than a hundred commercial reviews the root cause of stalled growth usually sits in proposition, ICP or product-market fit, not in sales execution. Automating outreach to a customer profile that has stopped buying produces more efficient failure.
What does it cost to build commercial tooling now?
For the kind of examples described here, days and thousands rather than quarters and millions. The larger obstacle is usually an approval process still calibrated to the old cost structure.
What skills should commercial teams hire for now?
Domain expertise and critical thinking more than technical build skill. AI output sounds right; establishing whether it is right requires judgement. The best results come from commercial leaders and builders choosing together rather than sequentially.
How many initiatives can a commercial team realistically run?
One or two completed in a year, unless someone is dedicated to running the plan. Sequencing upstream first and limiting live work to two things at a time is what makes the difference between a plan and a list.
Ready to aim higher?
If you know the execution gap is closable but not which gap to close first, that is the diagnosis we run, across the whole commercial engine rather than the part that is easiest to automate.