B2B Marketing Strategy

A set of decisions, not a list of activities

A marketing strategy is a small set of decisions that constrain everything the team does afterwards: which businesses you are trying to win, which you will decline, what you will claim about yourself, and which activities you have chosen not to fund. The campaign calendar, the channel mix, and the content plan are execution that follows from those decisions.

What gets called a strategy is usually one of four other things — a budget split across channels, a list of tactics with owners and dates, a set of numeric targets, or a deck of aspirations about market leadership. Each is a legitimate artefact. None constrains anything. One test separates them: if the document does not rule something out, it is not a strategy. A plan that says yes to LinkedIn, search, events, content, and account-based programmes has made no decision; it has divided a budget.

A real strategy is therefore uncomfortable to sign off, because somebody's preferred channel gets defunded and somebody's favourite segment gets deprioritised. If it cleared review without an argument, it probably says nothing.

Why the structure of business buying changes the answer

B2B strategy is not consumer strategy at smaller scale. Four structural features change what a sensible plan looks like.

• The market is often small and countable. If a few thousand companies in the world could plausibly buy what you sell, reach and frequency arithmetic behaves nothing like a consumer category. You can name your market, which means you can also get it wrong in a way that spending more will not fix.
• The buyer is a group. Gartner's 2025 survey of 632 B2B buyers describes buying teams ranging from five to sixteen people across as many as four functions. A strategy pitched at one job title is invisible to most of the people who can stop the purchase.
• Cycles outrun reporting periods. Spend in one quarter becomes revenue in a later one, which makes quarter-by-quarter optimisation actively misleading.
• Deal values are high and volumes low. A handful of wins can make a year, so an average computed on twenty deals is noise wearing a decimal point.

Together they push a B2B strategy toward naming a specific set of companies, being patient about payback, and distrusting any metric derived from few deals.

Building the segment from won and lost deals

An ideal customer profile (ICP — the description of the kind of company you win, keep, and expand) should come from evidence you already own: your own closed-won and closed-lost records, not a workshop.

1. Export every deal from the last two to three years, won and lost, with industry, employee count, revenue band, geography, corporate structure, and the software the company already runs.
2. Add the outcome data your CRM buries — retention, expansion, support load, discount given, cycle length.
3. Find the segments where you win at above-average rates, at above-average prices, and keep the customer afterwards. All three, not one.
4. Read the losses properly. Not the dropdown reason a rep picked, the narrative. Losses cluster by segment more reliably than wins do.

What emerges is usually uncomfortable: the segment you win in is smaller and less glamorous than the logos on the aspirational slide. An aspirational ICP is the most expensive error in the whole document, because messaging, channel choice, content, and sales hiring all inherit it.

Write the exclusions as explicitly as the inclusions. "We do not sell to companies under 200 employees" is more useful than any positive description, because a junior person can enforce it on a Tuesday.

Positioning a competitor could also sign

Positioning is the claim you make about what you are and who you are better for. The test is whether a competitor could put their name on your statement without changing a word. Most B2B positioning fails immediately: the enterprise-grade platform that helps teams work smarter, built for scale, trusted by industry leaders. Everyone in the category says it.

A claim that works has three properties. It is contestable — a competitor would dispute it rather than nod. It is falsifiable — some evidence would prove it wrong. And it names a trade-off: you are better for a particular kind of buyer because you are worse at something else. "The only system in this category built for regulated multi-entity groups, which makes it a poor fit for a single-site business" is positioning. "Powerful yet easy to use" is a wish.

The trade-off is the part teams delete, because it reads as admitting weakness. It does the opposite. A stated limitation is the most credible sentence on a website, and it disqualifies bad-fit enquiries before sales spends a quarter on them.

Choosing channels from where the audience already is

Channel selection starts with an unglamorous question: where do these specific people already spend attention, and what can you buy or earn there? Not which channel the team enjoys, and not which one produced a good quarter eighteen months ago.

For most B2B categories the honest list is short — search demand around the problem and the category, LinkedIn because firmographic and job-function targeting genuinely exists there, trade publications and their newsletters, a small number of conferences the buyers actually attend, and the customer's own peer network. Anything else needs a reason, and "a competitor is doing it" is not one.

Budget then splits between building future demand and capturing current demand. Brand work pays off later, when a buyer enters the market; activation converts people already looking. The commonly quoted 60:40 brand-to-activation ratios trace to Binet and Field's IPA Databank analysis, which is largely consumer-weighted with a much smaller B2B extension. Use them as evidence that long-term work is systematically underfunded — that argument is well made — not as a measured B2B optimum. Your own split should follow your category's purchase cycle, your share of existing category search, and how much pipeline already arrives inbound.

Where strategies go wrong

Five failures account for most bad strategies, and each substitutes an easy decision for a hard one.

• Tactics chosen before decisions. The document opens with a channel list because channels are concrete and choices are contentious. Fix the order: segment, claim, exclusions, then channels.
• A segment defined by wishful thinking. The ICP describes the customers an executive wants rather than the ones the company wins. Rebuild it from deal data and write down the disqualifiers.
• Positioning by committee. Each reviewer adds a qualifier; by round three the claim is inoffensive and interchangeable. Give one person the pen and protect the trade-off.
• Channels chosen from internal skill. The team runs paid search well, so the strategy is paid search. Defensible for a quarter, indefensible for a year.
• No written exclusions. With nothing declared out of scope, every request gets absorbed and the plan becomes a queue.

A sixth is worth naming separately: a strategy nobody can connect to the commercial arithmetic — price, deal size, accounts needed — is a marketing document rather than a business one, and it loses the first budget argument it meets.

One page, and how to review it without rewriting it

A strategy that does not fit on one page is not being used. The page carries six things: the segment and its exclusions, the claim, the three or four channels you are funding, what you have decided not to do, the handful of measures you accept being judged on, and the review date.

Measurement should follow the purchase cycle, not the reporting calendar. If a typical purchase takes seven months, a quarterly read on pipeline by channel is mostly noise. Report leading indicators monthly — qualified enquiries by segment, the share of enquiries that fit the profile, branded search volume, meetings booked with in-profile accounts — and lagging indicators on a rolling twelve-month basis. Treat self-reported source data as a directional cross-check on analytics, never as arbitration.

Then separate two review questions. Are we executing? That is monthly or quarterly, and it is about delivery. Is the strategy wrong? That is annual, reopened early only by a change in the facts: a competitor at a materially different price, a win-rate shift visible across several quarters, a product change, or a regulatory change. A strategy rewritten every quarter is not a strategy; it is a mood.

Frequently Asked Questions