How to Choose a Digital Marketing Consultant (Without Getting Burned)

Ask what they would kill, not what they would launch

Anyone can propose new channels. A new channel is exciting, it fills a deck, and it delays the moment anyone has to admit something is not working. The harder discipline is subtraction, so the most revealing question in a first conversation is what they would stop doing in your account.

A consultant who has actually operated will answer this quickly, because they have had to kill their own campaigns before. Someone who has only advised will hedge, because stopping something is the only recommendation that can be proven wrong within a month.

Insist on seeing the constraint before the plan

A funnel has one binding constraint at a time. If your landing page converts at 0.4%, more traffic is money set on fire. If you convert at 4% but only 300 people see the page, better copy changes very little. These need opposite plans, and no one can know which you have before looking at your data.

Be suspicious of a proposal that arrives before anyone has opened your analytics. A pitch written without your numbers is a template, and you will be paying retail for something assembled wholesale.

Separate strategy from execution in the contract

The common failure is paying a strategy fee and receiving a document. The second failure is paying an execution retainer and receiving activity with no thesis behind it. Structure the engagement so the same person is accountable for both, or make the handoff explicit and paid separately.

A clean shape for a first engagement looks like this:

• Weeks 1–2 — diagnosis. Full funnel audit, a written constraint, and a ranked list of what to fix. Fixed fee, delivered whether or not you continue.
• Weeks 3–10 — build. Ship the fixes. Defined deliverables, not hours.
• Month 4 onward — compound. Retainer tied to a testing cadence and a monthly review, cancellable with 30 days' notice.

Check whether they can build what they recommend

Strategy that cannot be implemented is a wish. If a consultant recommends a landing page rebuild, server-side tracking, or a new checkout flow, ask who does that work. If the answer is always an unnamed third party, you have acquired a middleman and a margin, not a solution.

This matters more in smaller companies, where there is no internal team to absorb the handoff. The gap between the deck and the deployment is where most marketing budgets quietly die.

Agree the number before you start

Fix one primary metric for the engagement and write it down. Qualified leads per month, cost per acquisition, contribution margin — the specific metric matters less than the fact that it is singular and agreed. Engagements that track eight metrics are engagements where nobody can fail.

Then agree the review date. Ninety days is usually right: long enough for compounding to show, short enough that a bad fit is not expensive.