Part four of a five-part series on building a B2B marketing function from scratch.

With the story straight and the plumbing fixed, we could begin to turn on demand activities. And the first rule, with a small sales team and a smaller budget, was simple. We could not afford to throw things at a wall and hope. Every pound of budget and every hour of selling time had to earn its place, not get burned on poor-fit leads that were never going to progress to close. So we built the opposite. A motion that aimed.

This series is about building a GTM from scratch. The story and the plumbing came first. This post is the demand engine, turning all of it into a predictable pipeline.

You do not sell to a TAM

Somewhere in every deck there is a slide with an enormous number on it. The total addressable market, billions of pounds, growing at some confident rate a year. It is a useful number for raising money and a useless one for building pipeline. You do not sell to a TAM. You sell to a finite list of actual people, at actual companies, with an actual problem, who can actually sign a contract.

That mindset shift is important. A TAM tells you the prize. It tells you nothing about who to call on Monday. The work is turning that abstraction into a named list, the specific accounts that fit and the specific people inside them who feel the pain and hold the budget, then earning their attention one at a time. Marketing that aims at a market hits nobody. Marketing that aims at people closes deals.

And even that list is not all live at once. At any moment only a small slice of any market is actually in-market, the accounts with a live problem and a reason to act now. The research puts it at roughly five percent in and ninety-five percent out, the 95-5 rule. That five percent is your active demand, and it is where this quarter’s deals are. The other ninety-five are not bad accounts, they are just not buying yet, so you stay visible to them and put your effort into the few that are in the window. Finding that five percent is the whole game, which is where signal comes in.

Signal, not spray

The motion was signal-led, and it started with the ICP from the positioning work, the specific accounts we genuinely solved a painful problem for. Everything pointed at reaching those people at the right moment with the right context.

The signal is the thing. Lead scoring on fit and intent, so a right-fit account showing real buying behaviour rose to the top and a tyre-kicker did not. Buyer-intent and website-activity signals telling us which target accounts were on the site, what they were reading and how often, which is the difference between someone idly browsing and someone quietly building a business case. Enrichment filling the gaps in the CRM automatically, so the team was not hand-keying job titles like it was 2009.

We ran this through an intent engine, signal-led tooling native to HubSpot, with TechTarget’s Priority Engine layered on top for third-party purchase intent. In tech and security it earns its place because it does not just name an in-market account, it surfaces the actual people researching the category.

And the part that pays you back is timing, not just fit. An account can match the ICP perfectly and still be the wrong account to chase this quarter, because they signed a multi-year deal with someone else last month, or because you have arrived too late to a process already sewn up. Fit tells you who could buy. Intent tells you who is in the window now, showing the behaviour of a team quietly building a case. That is where you spend, because that is where you can still change the outcome. Chasing a perfect-fit account that is locked in is just a slower way of wasting the budget.

Signal in action

A target account would start showing up, a few people from the same company reading the same pages, an uptick in intent across our category. The system flagged it, pulled together what we knew, and assembled a short pre-demo brief. Who they are, what they have looked at, what is probably on fire for them, the pillar that speaks to it. By the time a salesperson picked up the phone, they were not cold-calling a name on a list. They were walking into a conversation already half-warm, with context in hand instead of a name and a prayer. Less grunt work, more of the right conversations, with a team that could not afford a single wasted one.

The hardest part of a signal-led motion is not the technology. It is the discipline to ignore a good-looking lead that is off-ICP. A big logo downloads a guide and everyone gets excited. If they are not who you are for, you let them go, or at least you do not drop everything to chase them. Aiming only works if you are willing not to fire at the wrong target, however tempting it looks, and there is always a tempting wrong target.

Let the machines do the first mile

We put a couple of tools to work sharpening that signal, and then did something that is becoming normal now but still felt like an edge at the time. We put an agent on the first layer of SDR qualification, doing the initial pass before a human ever picked up the phone.

Think about what the first mile of qualification actually is. Checking a lead against the ICP. Confirming the basics. Working out whether this is a real opportunity or someone who downloaded a guide and will never buy. It is high volume, repetitive, and a genuine waste of a good salesperson’s day, and it is exactly the kind of work a machine does tirelessly, without getting bored or demoralised by the tenth dead end in a row.

So the agent handled the first pass, sorted the obvious noise from the genuine signal, and only the qualified, worth-a-human conversations reached the team. That did two things. It made a tiny team behave like a much bigger one, and it meant the humans spent their hours on the conversations that actually needed a human, which is where they were always going to add the most value. You do not replace the salespeople. You stop wasting their time and get them more efficient. Do that consistently and the sales team starts trusting what marketing sends them.

Test, learn, ramp

We did not turn the spend up until the systems felt more mature, and even then we started small, deliberately, to test the water rather than spray indiscriminately.

The tight budget honed our focus and forced discipline. Every pound had to earn its place, so we ran small, measured everything against the single source of truth we had already built, and only scaled what the numbers actually justified. A channel that worked got more. A channel that did not got killed quickly, before it ate a quarter’s budget proving a point we could have learned for a fraction. Test, learn, refine, then ramp, in that order, and never ramp on a hunch.

This is where the ‘boring’ foundation work from the last post pays you back. Because the pipeline definitions were clean and the numbers were real, we could see what each pound bought, attribute it honestly, and make the next call with confidence instead of crossing our fingers. Most teams turn spend up and hope. We turned it up and knew.

Fewer events, squeezed harder

The same discipline decides where you show up in person. Marketing earns much of its bad reputation at events, and sinks a lot of budget into them, so the move is to run fewer and get far more out of each one.

Say no more than you say yes

Say yes to every show sales asks for, book the stand, ship the banners, scan the badges, and you look busy while burning a budget with nothing to show for it. So ask the question. Why are we doing this event? “We did it last year” is not an answer. Is our ICP actually in the room, or just a crowd? Have we something real to say on the theme? Are our competitors there, and if so, have we a reason to stand next to them or a better one to be elsewhere? You can take a rival on at their own event, but only if you turn up with something worth hearing.

Sometimes the right call is to pass, and that is not the same as saying no to everything. It is choosing where a finite amount of energy goes and saving it for the battles that matter. A pass you can justify beats a yes you cannot. Have the confidence to say no, and back it with a reason rather than a shrug.

Twelve beats four hundred

Once you are choosing rather than reacting, the shape gets simple. Pick a handful of must-attend events, the ones within budget where you can genuinely make a dent, and run at those hard. A real presence at two or three, not a token one at eight.

Then do the thing most teams underuse. In your key regions, run small. Breakfast, lunch or dinner, ten or twelve of the right buyers around a table, a sharp topic, and a subject matter expert to lead it, someone who has done the work, has the stories, and can translate them into what they mean for the people in the room. Not a salesperson walking a product deck, because a pitch over dinner is not value, and every buyer at the table knows it inside a minute. Twelve of the right people beat four hundred badges scanned at a trade show, every time. It is a lift, you will graft with sales to fill the seats, and almost nobody does enough of it. Line up two or three themes and take them on the road, one region at a time.

Atomise everything

You built all that for a handful of rooms. Do not let it die there. Record the sessions, then do the thing that actually pays, atomise them. The content is built and the speakers are drilled, so the hard part is behind you, and one session breaks into a thousand smaller pieces: the full talk on a YouTube channel and a webinar channel, a section on the website if it fits, the sharpest two minutes snipped up for socials. A room of twelve becomes an audience you keep reaching, including the ninety-five percent who were not ready when you first showed up.

And this is not just an events trick. Every serious thing you make deserves the same treatment. A talk, a report, a customer story, a launch, each one is a dozen assets pretending to be one. Get in the habit of squeezing every last drop from every effort. Build once, use everywhere.

Outsource the engine room

One more decision shaped all of this, and it is the one people get most wrong. In our company “agency” was close to a dirty word, shorthand for budget spent and little returned. I understood the scepticism, but what existed when I arrived was not an SDR team. It was a glorified sales-ops function appending data to contacts, busy but not productive, motion without movement. So we handed early-stage pipeline generation to a specialist agency for a fast ramp. That bought us data, tools, reporting and scale we could never have built with three people, and it took the early-stage grind off a small, stretched sales team so they could focus on late-stage pipeline and closing. The plan was always to bring it in-house once we had the maturity and the volume to justify it. Economies of scale first, ownership later, a pattern worth holding onto for any function you cannot yet staff properly.

When an outsourced model does not work, the model is rarely the problem. Your GTM is, and the agency is the easy thing to blame. So before you point a finger, look at your own house. Are the processes actually in place? Have you trained them, or handed them a login and a target and wished them luck? Are you treating them as an extension of your team, with clear goals and a real shot at hitting them, or as a line item to shout at when the numbers miss? Outsourcing is not a silver bullet. You still have to do the hard yards to set the team up to win. Every agency I have ever brought in has added value. Coincidence?

That is the whole point of doing the foundation work first. When you finally spend, you can see exactly what it buys, ramp what works, and kill what does not, fast. Signal over spray, every time.

Next, building the brand that gets a small company into the room with the big players, and helps close the deal.

Godspeed.