A startup marketing consultant is bought for one of three quite different reasons, and confusing them is the usual cause of a disappointing engagement. Some founders want a strategist to decide who the product is for and how it should be sold. Some want an operator to run acquisition until the numbers justify a full-time hire. Some want a fractional executive to build a team and a plan a board will fund. Those are different people with different rates, and very few individuals do all three well. The market is crowded, entry costs nothing, and the loudest profiles are not the most useful, so the buyer needs a test that is harder to pass than a good conversation. This page sets out how to decide which of the three you need, what the engagement should produce, and how to check.
Which of the three problems you actually have
If you cannot state in one sentence who the product is for, what it replaces and why anyone would switch, you have a positioning problem and you want a strategist. That work is short, intense and expensive per hour, and it should end with a written document your engineers and your salespeople both recognise as true. If positioning is settled and nothing is growing, you have an execution problem and you want an operator who will actually run the channels rather than advise on them. If the company has product-market fit, revenue and a board asking for a plan, you want a fractional executive who can hire, budget and forecast. Be honest about which, because a strategist hired to execute will produce decks and an operator hired to think will produce activity. Write the problem down before the first call and see whether the candidate reframes it convincingly or simply agrees.
Consultant, agency or hire
A consultant is right when the problem is bounded, the answer needs judgement more than hands, and hiring the wrong permanent person would be expensive to unwind. An agency is right when the work is ongoing production across a channel and you need capacity rather than direction, which is why most early companies end up buying organic search as a service rather than learning it internally. A hire is right when the work is continuous, close to the product, and needs someone in the room every day. The costliest mistake is buying a consultant to avoid deciding, then extending the engagement quarter after quarter because ending it would mean admitting the decision is still open. Set a term, set what the term produces, and make the renewal a genuine decision rather than a default.
What a good engagement produces
Deliverables are the only thing that makes a consulting arrangement testable, so define them before the money moves. A positioning engagement should produce a written statement of segment, problem, alternative and differentiation, plus the messaging derived from it and the evidence behind each claim. An acquisition engagement should produce a working channel with its costs, its conversion rates and a documented process someone else can run. A fractional engagement should produce a plan, a budget, a hiring specification and a handover. In every case there should be a named artefact you still own after the consultant leaves, and access to every account created on your behalf. If the only artefact is a weekly call, you have bought company rather than work.
Testing a candidate without a bake-off
Ask for two specific things. First, a named client where you can speak to the operator who worked with them, not the founder who signed the contract, and ask that person what changed and what did not. Second, a short piece of live judgement: give the candidate your pricing page and your last month of numbers and ask what they would change first and why. Strategy sounds identical across candidates; a first move does not. Watch for claims that cannot be true, particularly guaranteed results on channels nobody controls. Google's own guidance on hiring search help warns that guaranteed rankings are not something an outside provider can promise, and the same scepticism transfers to any consultant guaranteeing an outcome that depends on an auction. Check that any performance claim in a pitch is one you could verify yourself; advertising rules apply to consultants selling services as much as to anyone else.
Questions people ask about startup marketing consultant
How long should the first engagement be?
Short enough to be a real test and long enough to produce something. A defined project of six to twelve weeks with a named deliverable works better than an open monthly retainer, because it forces both sides to agree what success looks like before anyone is invested in pretending. Renew deliberately if it worked. A retainer that rolls without a review turns into a subscription to a relationship rather than a purchase of work.
Should we pay for performance instead of time?
Rarely, at an early stage. Performance deals need a metric both sides trust, and most startups do not yet have clean attribution or a stable conversion rate, so the arrangement collapses into an argument about measurement. They also push the consultant towards whatever moves the metric fastest rather than what compounds. Pay for defined work, measure the outcome yourself, and renegotiate on evidence.
Is a consultant with no experience in our category a problem?
Not necessarily, and category experience is oversold. What matters more is motion: someone who has sold a similar product to a similar buyer at a similar price, whatever the category. A consultant who has only worked in consumer subscriptions will struggle with a long enterprise sales cycle regardless of how many decks mention your industry. Ask which motions they have run, not which logos they have.
What should we own at the end?
Everything created on your behalf: the domain, the analytics property, the search console account, the ad accounts, the documents and the raw files. Create the accounts yourself and grant access rather than accepting accounts the consultant owns. This is the single clause most often skipped and the one most often regretted, because rebuilding a year of measurement history is impossible.