The sequence, and what depends on what
Build a B2B marketing strategy in this order: gather evidence from won and lost deals, define segments and an ideal customer profile, settle positioning and test the message, set objectives with baselines and dates, choose channels from where the audience already is, split budget between capturing existing demand and creating future demand, agree the measurement plan, and fix a review cadence. The order is not a preference — each stage produces the input the next one needs.
The dependencies matter, because most stalled strategy work has skipped one. You cannot choose channels before you know the segment, because viability depends on how much of it you can address. You cannot set an objective without a baseline, or you are naming an aspiration. You cannot allocate budget before positioning is settled, because you do not yet know what the money is buying. And you cannot judge any of it later unless the measurement definitions were agreed first.
Realistic elapsed time for a first attempt is six to ten weeks, most of it waiting rather than writing.
Stages one and two: buyer evidence, then segments you can count
Weeks one to five. The first stage replaces internal opinion about why people buy with what buyers actually said. Four sources, in descending order of value: eight to twelve interviews with recent buyers, won and lost, asking what changed that made them look, who else was involved and what nearly stopped it; twenty to thirty recorded sales calls, listened to rather than summarised, whether through Gong or Fathom or the recordings sitting unwatched in your conferencing platform; the closed-lost reasons in the CRM, re-coded by you, because price usually means value not established; and support tickets, which show what buyers misunderstood before buying.
The second stage turns that into segments. Build the ideal customer profile from firmographics — industry, employee count, revenue, geography — and technographics, then validate it against your best existing customers by retention, expansion and margin rather than by logo value. Size it with a business database such as Dun and Bradstreet, ZoomInfo, Apollo or Cognism; that count decides whether broad media is arithmetically sensible at all. Use domains rather than company names as match keys, and decide in advance how subsidiaries and acquired brands are handled.
Artefacts: verbatim quotations by theme, a re-coded loss tally, the words customers use, two to four counted segments, and an exclusion list. A strategy that excludes nothing has not decided anything.
Stage three: positioning, and a message test that can fail
Weeks five to seven. Positioning answers four questions in a form a stranger can repeat: for whom, against which alternative including doing nothing, what claim, and on what evidence. The fourth is where most attempts collapse — an unsupported claim is a slogan, and buying committees contain people paid to find them.
Two tests, both of which must be capable of failing. The substitution test: replace your company name with your closest competitor's, and if the statement is still true you have written a category description rather than a position. Most B2B boilerplate fails this in the first sentence. The read-back test: read the positioning to five people inside the segment who do not work for you, then ask them to say what you sell and who it is for. If three of five get it wrong, the problem is the words.
Run both before commissioning design work. Artefact: one page. If it takes more, the decisions have not been made.
Stage four: objectives a finance director would accept
An objective needs four components: a baseline, a target, a date and an owner. Missing any one makes it a wish. Grow awareness is a wish. Move unaided awareness among maintenance engineers in four named states from a measured baseline to a stated figure by the end of the third quarter, using the same survey instrument, is an objective — because it can fail.
The specific B2B difficulty is sample size. If you close thirty deals a year, quarterly targets stated in closed deals are noise: two deals slipping across a quarter boundary swamps any real effect. Set objectives at the altitude where you have enough events to see a change — reach and share of category search at the top, qualified opportunity creation in the middle, revenue and win rate over four quarters rather than one.
Write down at the same time what you are not optimising this year, or every metric quietly becomes a target and nothing moves enough to notice. Three objectives is the practical ceiling.
Stage five: choosing channels from where the audience already is
Do not start from a list of channels and ask which to use. Start from three pieces of evidence: the search demand that already exists, judged by who plausibly types a query rather than by volume totals; what interviewees said about where they read, ask and check — trade publications, associations, technical committees, distributors; and who customers name unprompted when asked whom they trust.
Then apply two filters. Can this channel address a meaningful share of your segment, and can you afford enough repetition in it to be remembered rather than merely present. As an illustration of how quickly small audiences collide with platform mechanics, LinkedIn's guidance recommends audiences above 50,000 for Sponsored Content and above 15,000 for Message Ads, and a matched audience built from an uploaded list resolves to far fewer targetable members than the file has rows once records are matched and the required location facet applied. Short named-account lists frequently cannot run at all.
Artefacts: a channel plan naming what runs, at what weight and in what sequence, plus a written kill list of channels you are not using.
Stage six: the budget split, and the ratio you should not quote
Money in B2B marketing does two jobs. It captures demand that already exists — search, retargeting, the website itself, sales enablement — and it creates the memory that makes you a candidate later, through reach, brand work and public relations. Both are necessary; the argument is only ever about proportion.
Handle the commonly cited ratios honestly. The brand-versus-activation splits quoted in B2B discussions originate in Les Binet and Peter Field's IPA Databank analysis, distributed for business markets through the LinkedIn B2B Institute. That work is substantial and case-based, but largely weighted toward consumer cases, and the B2B extension rests on a much smaller sample. It is a reasonable prior, not a measured B2B optimum — and presenting it as one to a finance director who later checks the source costs more credibility than having no number at all.
A defensible method instead: fund the demand-capture floor first, meaning the activities that measurably lose money when switched off. Decide the remainder against your own replacement cycle and market size, then instrument it so that in two years you argue from your own data. Budget the unglamorous lines too: research, the website, data hygiene.
Stage seven: agreeing measurement before anything launches
Write the measurement plan while the strategy is still a draft, and have sales and finance sign it. It needs the leading indicators that count as early evidence, the expected lag between spend and revenue signal, written definitions of a qualified lead, an opportunity and a source, the named reporter, the frequency, and the decision each report informs. A report that informs no decision will be discontinued within a year, and should be.
Three traps are specific to long cycles. Attribution across quarters, where the influencing activity and the closed deal fall in different reporting periods and each side of the business claims the version that suits it. Small samples, where ordinary variation reads as a trend. And self-reported source data, which usually disagrees with analytics — keep the open how-did-you-hear-about-us field anyway, because it often names the channel analytics cannot see.
Where the audience is large enough to divide, hold something out: a region, a segment, a tranche of accounts. A holdout is the only structure that answers whether the spending caused anything.
The document that gets written, filed, and never opened again
The most common failure is not a wrong strategy but a correct one nobody consults. The symptoms are consistent: a deck presented once; a channel plan that no longer describes what is running by month three; nobody able to state the positioning without looking it up; and a document written by an agency, so the people executing never made the decisions.
The causes are structural, so the fixes have to be:
- A one-page version that fits on a wall — segments, positioning, three objectives, channels in and channels out. The long document is the appendix.
- A named owner accountable for the objectives, not for producing the document.
- A quarterly review with the measurement plan's real numbers present, and permission to change a decision. A strategy that cannot be amended will be ignored instead.
- A decision log recording what changed, when and on what evidence.
- An annual repeat of the evidence stage, because loss reasons move.
If you take two things from this sequence, take the first stage and the last: listen to buyers directly, then put the review meeting in the calendar before you write anything.
Frequently Asked Questions
How long does it take to build a B2B marketing strategy?
Six to ten weeks for a company doing it properly for the first time, and most of that is waiting rather than writing. Buyer interviews take two to three weeks to schedule and complete, and CRM data usually needs cleaning before it can be read.
If a strategy is delivered in a fortnight without recent win-loss research, the evidence stage was skipped and every later decision rests on internal opinion.
What is the difference between a marketing strategy and a marketing plan?
The strategy is the set of decisions: which segments, what position, which objectives, which channels are in and which are deliberately out. The plan is the schedule that follows — campaigns, owners, dates, budget by month.
Most documents labelled strategy are plans with a mission statement attached. If it contains no decision about what you will not do, it is a plan — useful, but impossible to evaluate, because there is no stated bet to be right or wrong about.
How much should a B2B company spend on marketing?
The percentage-of-revenue benchmarks in circulation are not comparable with each other. They come from different survey populations, define marketing spend differently — some include salaries, some do not — and cover categories with very different growth assumptions.
Reason structurally instead. Spend is driven by how many organisations you must reach, how often they buy, what a contract is worth across its term, and how much ground you are taking from incumbents.
Who should own the B2B marketing strategy?
Whoever is accountable for the objectives it sets, which means an internal marketing leader rather than an agency. External help is useful for the evidence stage and for challenging comfortable assumptions, but if the decisions are made outside the company the people executing never weighed the trade-offs.
Sales needs to co-sign two things: the segment definitions and the definition of a qualified opportunity. A strategy sales did not agree to produces leads nobody follows up.