Marketing complex B2B services: why ads fail and what works instead

The most honest marketing meeting I ever sat in lasted twenty minutes and ended with a question nobody could answer.

We were in a general contractor’s meeting room. Marketing presented the report: impressions, clicks, cost per lead, a handful of website enquiries. The commercial director looked at the slide and asked: “Fine. How many of these became contracts?” Silence. “None. But we have two projects that came through the conference in Kraków.” When was that conference? Two years ago.

That silence is where marketing complex B2B services actually lives. You spend the budget today, the result arrives eighteen months later, through a different channel, and the link between the two is very hard to prove.

I did not learn this as an outside consultant. I learned it from the inside: I worked inside engineering companies and general contractors in strategic business development roles. I have had the same conversation with marketing teams in Poland, Ukraine, the EU, North America and the Middle East. The problem was the same in every one of them, and so was the answer.

Why ordinary marketing breaks when you sell complex B2B services

The tools most companies use were built for a different kind of sale. Here is where they come apart, point by point.

The product does not exist until the contract is signed. You are not selling an item off a shelf. You are selling a commitment to do something nobody has done in these conditions, at this budget, on this ground, inside this regulatory frame. An ad has to show “what”, and the “what” has not been defined yet. It is born during the negotiation.

Demand is episodic, not continuous. A plant expands production once every seven years. A developer enters a new segment once every three. Advertising works where demand exists daily and you only need to intercept it. Here there is no such moment in the calendar, it arrives when the board signs off the capex, and you are the last to hear about it.

The market is microscopic. In a specific niche, cold storage design, say, or heavy-duty industrial flooring, a country may hold 150 to 400 buying organisations and a few thousand people who genuinely influence the decision. Ad platforms cannot target at that resolution. You pay for impressions to tens of thousands, of whom thirty are relevant. And those thirty are not clicking banners, because they are in a meeting.

A committee decides, not a person. Chief engineer, technical director, procurement, legal, finance, sometimes the owner, sometimes a lender or an insurer. Gartner has described the same pattern for years: a typical B2B purchase involves a buying group of six to ten people, each with their own logic and their own fear. The technical director fears it will not work, procurement fears overpaying, legal fears liability. One message does not close all of them.

The cycle is measured in years. From first conversation to signature: several months to three years. In that time the client’s budget changes, your head of marketing changes, and the attribution window in your analytics expires long before the decision is made. In tendered work there is one more thing: by the time the RFP is issued, the buyer already had a shortlist of firms they trust. You earn a place on that list over years, not campaigns.

The problem nobody likes to say out loud

The hardest part is not that the channels do not work. It is that proving they work is very hard.

Two equally bad things follow.

First: marketing starts reporting whatever it can count, reach, impressions, followers, number of articles published. It creates the illusion of activity. The managing director reads the report and sees numbers that mean nothing to the P&L.

Second, and worse: when costs come under pressure, the marketing budget is cut first, because it is the only one that cannot defend itself with numbers. The consequences surface eighteen months to two years later, when the pipeline is empty and nobody can say why.

I have watched both. I have also watched a team that spent two years building industry presence lose most of it in a single quarter of cuts, because conference invitations go to people, not to companies, and the people had left.

Expertise as the only working asset

When you cannot sell the product because the product does not exist yet, you sell the only thing that does exist at the moment of the conversation: a demonstrated ability to solve the problem.

That means expertise. Not as a slogan on a website, but as a specific, checkable, narrow competence held by specific people.

Look at it from the buyer’s side. A client cannot assess the quality of the service until it has been delivered, and by then they have spent millions. So they assess what they can assess now: does this team understand my problem better than I do? Every article, every talk, every answer to a technical question is a free sample of the work to come.

Most firms make the same mistake at this point.

Trust is built by narrowness, not by breadth.

“Five construction trends for 2026” gives a reader nothing, anyone could have written it, and the reader can tell. But a piece on “how we calculated settlement of a piled foundation on soft soils under winter concreting, and where our first iteration was wrong” is the one a chief engineer saves, forwards to a colleague, and remembers the author’s name from.

Something else works against intuition: admitting difficulty and error gets you further than demonstrating perfection. Everyone in this segment is an adult and everyone knows that something goes wrong on hard projects. A firm that explains how it responded when things went wrong looks safer than a firm whose portfolio contains only victories.

Channels for marketing complex B2B services, ranked by real return

The order below came out of practice. It holds up in Europe and North America alike, with an adjustment for culture.

1. Offline: industry events, talks, technical committees

The biggest return and the least attention paid to it. Not an exhibition stand, but a talk. The difference is large: a stand makes you one of forty suppliers in the hall; a presentation makes you the only person your target audience listens to for forty minutes.

The same category includes technical committee work, contributing to standards, membership of professional bodies, judging awards, guest lecturing. It is slow and returns nothing inside a quarter, but it builds a position competitors cannot buy.

2. LinkedIn, from personal profiles, not the company page

The company page is close to dead in this segment: people do not read companies, they read people. A chief engineer’s profile with two thousand relevant connections outperforms a company page with ten thousand followers, half of whom are recruiters.

In practice that means marketing services the personal profiles of your experts rather than an abstract “company”. Yes, with the risk that an expert leaves. That risk is manageable, and cheaper than the alternative.

3. The website as a technical library, not a blog

Nobody reads company news. What does get read: methods, breakdowns, calculations, checklists, worked details, explanations of codes and regulations. Material somebody opens because they have a specific problem on their desk.

It is also the only channel that works at the moment the client finally starts searching, and they are not searching for “general contractor Warsaw”, they are searching for the wording of their own technical problem.

4. Closed formats: workshops and joint technical sessions

Ninety minutes with the client’s engineers over their actual drawing beats a year of banners. It is almost a sale, but formally not a sale yet, which is exactly why it works.

5. Trade press and co-authored publications

An article in a professional journal, written by your engineer together with the client’s engineer, is simultaneously content, a case study, third-party proof, and a strengthening of the existing relationship.

The hard part: making it work operationally

This is where most such programmes die, on daily routine, not on strategy.

Engineers do not want to write. And they should not have to. The right model: the marketer does not write instead of the expert and does not ask for a draft. The marketer runs an interview. Forty minutes recorded with an engineer at the end of a project stage yields an article, three posts and the outline of a talk. The expert spends forty minutes rather than three evenings, which is why they agree to a second one.

One expert, one topic. Not “our technical director writes about everything”. Each carrier of expertise should own one recognisable, narrow area. It is easier for them and it registers better with the market.

The content plan follows the industry calendar, not the marketing calendar. You do not invent topics for the quarter, you look at which conferences, tender cycles, regulatory changes and seasonal peaks fall in the next six months, and you prepare for those.

Every project must leave a trace. A practical rule: a project is not closed until someone has written down what can be shown to the market. Otherwise a year later you have twenty delivered projects and nothing to publish.

Business development: turning a talk into pipeline

Marketing creates the pretext. Business development turns it into pipeline. Without that link, the whole programme becomes expensive public education.

What that means in practice, using a conference as the example:

Before. Get the attendee list, or at least the company list. Pick 8–12 targets. Find out who will physically be there. Write in advance, not “come to our stand”, but “we are presenting on X, you have exactly this problem, do you have 20 minutes after the session?” Half will say yes.

During. The goal is not a hundred business cards; it is 5–10 substantive conversations, each ending with a specific technical problem written down. A card without a problem attached to it is litter.

After. Follow up within 48 hours, and not with “great to meet you” but with an answer to that problem: a calculation, a reference to a standard, an example from practice. That is the second free sample and it is what opens the next meeting.

Separately: business development should be feeding marketing its topics. The people who hear the market’s objections and questions every day are the best source of a content plan that exists inside any company.

How to measure when attribution breaks down

Give up on calculating the ROI of an individual post. Measure differently: by account and by leading indicators.

The evaluation horizon is not a quarter but 18–24 months. Agree it with the board at the start, in writing, together with the interim indicators. Otherwise the programme is shut down in month ten, two months before the first contracts land.

The most expensive mistake: “everyone already knows us”

Everything above breaks against one belief I have heard in every second company: “Everyone knows us. We built there, there and there. It is a small industry, everyone knows who we are.”

It is almost always false. And more dangerously, it is a falsehood that looks true for exactly as long as things are going well.

Where the belief comes from. When orders are flowing, when word of mouth works, when existing clients fill the pipeline by themselves, it feels like the system is working. What is actually working is inertia from previous years plus a handful of personal relationships. You look at a full order book and draw the wrong conclusion: the market knows us.

What is actually true. “Everyone” is usually 10–15 people: two or three current clients, a few former colleagues, a couple of designers. The committee that will decide your next large contract consists of different people. The client’s technical director changes, and the company “everyone knows” is off the shortlist. People rotate through this industry faster than reputation accumulates.

And separately: a completed building does not speak for itself. It stands there, it looks good, and it carries neither your name nor any explanation of the non-trivial problems you solved on it. A portfolio is evidence, but only for someone who has already started checking you out. It brings nobody in on its own.

So the moment you realise the market does not know you well enough is already the moment it is too late. The pipeline has dipped, you start acting, and between the first action and the first result lies a year and a half of emptiness. That year gets crossed on discounted pricing, the most expensive way to survive there is.

What to do about it. Accept that this is not a project but a mode of operation. Not a campaign switched on when numbers dip, but a steady, unglamorous activity that continues through the good years too, especially through the good years, because that is when you have both the resources and something worth talking about.

In practice that means a small but non-negotiable minimum, delivered regardless of workload: how many talks per year, how many pieces per quarter, how many meetings with non-clients per month. The numbers can be modest. What matters is that they do not depend on whether the quarter is good or bad.

And yes, you will not be able to connect a colleague’s conference talk to a contract that lands two years later. I understand that perfectly and I am not going to pretend a report exists that proves it. But I have seen enough companies on both sides of that choice to say it plainly: it is worth doing. The difference between a systematic and an unsystematic approach is not budget and not tooling. It is whether the work stops when business is good.

Common mistakes

1. Giving marketing a lead-count KPI. It produces cheap, irrelevant enquiries and poisons reporting for years.
2. Selling from the podium. A talk that turns into a company presentation destroys trust faster than not speaking at all.
3. Writing broadly “to reach more people”. Broader reach means less trust. The logic here runs backwards.
4. Publishing in the company’s name rather than people’s names. Anonymous expertise does not earn trust.
5. Outsourcing writing to an agency with no expert inside the process. A specialist reader spots non-specialist text in the first paragraph.
6. Separating marketing and BD organisationally without joining them by process. Marketing creates the pretext, BD never picks it up, the pretext burns.
7. Expecting results within the quarter. And therefore stopping exactly when it starts to work.
8. Switching marketing on only when the pipeline dips. The most expensive of them all.

Conclusion

In a business where the service is complex, customised and expensive, marketing cannot be about capturing attention. It can only be about demonstrating competence, consistently, publicly and specifically.

It is slower than advertising, produces no attractive dashboards and consumes the time of the most expensive people in the company. But it is the only thing I have seen work, in Poland, Ukraine, Western Europe and North America, in firms of every size.

And it does not switch on when needed. That is exactly why this work must not be paused in the year when business is good: the result of that year shows up not this year but two years out. Never stop doing it. Especially when it feels like you could.

Your expertise is your product. The only question is how many people outside your office know about it, and whether they are the same people who will be deciding next year.

If you are planning entry into a new segment or a new market right now, the same logic applies, I wrote about it in more detail in “Entering a new market: start with people”. The ways we can work together are described on the services page.