How to choose an
ecommerce growth
advisor.

I am an advisor, so read this with that in mind. What follows is the process I would want a brand to run on me, including the steps that would most likely disqualify me. That is deliberate. A selection process that a good advisor is happy to be put through is the only kind worth writing down, and if this page helps you decide I am the wrong fit, it has done its job.

Most brands do not run a process here at all. They ask two people they know, take the one who seemed sharpest on a call, and find out eighteen months later whether it worked. That is not a criticism, it is what happens when a decision arrives without a template. Hiring an agency has a template, the RFP, and most operators have run one. Choosing an individual advisor has no equivalent, so it gets made on vibes.

Here is the sequence, seven steps, in the order that actually protects you. Each step hands off to a deeper piece where one exists, because this page is the map, not the territory.

The order matters more than it looks. Six of the seven steps are cheap, and the one expensive step, the paid trial, only works if the five before it have narrowed the field to people worth paying. Run them out of order and you end up testing the wrong person carefully.

First, decide whether
you need an advisor
at all.

Outside judgment pays for itself in three situations, and only three. The first is a decision you will make once, where you have no pattern to draw on and the cost of getting it wrong is large: a replatform, a raise, an acquisition conversation, a category expansion. The second is a problem your team has been circling for two quarters without resolving, which almost always means the problem is framed wrong rather than under-resourced. The third is when you personally have become the bottleneck on decisions you no longer have time to think about properly.

It does not pay for itself when what you actually need is hands. If the diagnosis is clear and the work is execution, you need an operator on payroll or an agency with capacity, and an advisor will feel expensive and slow. It also does not pay when nobody internally has the authority to act on the advice. An advisor who reports into someone who cannot change anything is an expensive newsletter.

The situations where you need something else instead

The three-way choice between an advisor, an agency and a full-time hire is a genuinely different question with a genuinely different answer, and it is worked through properly in advisor, agency or full-time hire. If the honest answer is that you need your first senior operator on the team rather than outside counsel, that decision has its own shape, laid out in making your first senior operator hire. Resolve that before you go any further, because everything below assumes you have already decided the answer is an individual advisor.

Step 1: write down the
decision, not the
job description.

"We need help with growth" is the brief that produces a bad shortlist, because every advisor alive can say yes to it. It selects for the people most comfortable being vague, which is the opposite of what you want. A brief that nobody can decline is a brief that has told you nothing.

Write the decision instead. One paragraph, sent to three people, containing four things: where the business is now in real numbers, the specific decision or problem in front of you, what you have already tried and what happened, and what a good outcome looks like in ninety days. That is the whole document. It should take twenty minutes and it should be uncomfortable to write, because the discomfort is you discovering that the problem is less well defined than you thought.

Why the brief exists to help the wrong people say no

The point of a specific brief is not to inform candidates. It is to give the wrong ones a reason to withdraw. An advisor whose experience does not match the decision will usually tell you so if the brief makes the mismatch obvious, and the ones who do not tell you are exactly the ones you have now identified. This is the same mechanic that makes a good agency brief work, and the eleven-question version of it lives in the DTC agency RFP script, which is the sister page to this one. If you are selecting a firm rather than a person, start there instead.

Step 2: pick the model
before you pick
the person.

Audit, sprint, retainer or equity. The shape you choose changes who is even a candidate, which is why it comes before the shortlist rather than after it. An advisor who is excellent on a two-week diagnostic may be the wrong person for an eighteen-month embedded relationship, and vice versa, and you will not be able to tell the difference on a call if you have not decided which one you are buying.

The default answer for a first engagement is an audit or a scoped project, not a retainer. A retainer is a relationship, and you do not have one yet. Buying a relationship before you have tested the work is how brands end up nine months into something nobody wants to be the first to end. What each shape is actually good for is broken down in audit, sprint or retainer.

Equity deserves a specific warning. Trading advisory for equity can align incentives beautifully and it can also turn a three-month relationship into a permanent line on your cap table. The traps are worth understanding before you offer it, and they are catalogued in the advisor equity trap.

Step 3: where good
advisors actually
come from.

Almost always from operators who have worked with them. Not from a search, not from a marketplace, and not from a list. This is not snobbery about channels, it is a point about what each channel selects for. A directory or marketplace ranks on availability, responsiveness and willingness to be listed, and none of those correlate with whether someone can solve your problem. The best advisors are usually working, quiet, and at capacity, which is precisely why they are not optimising a profile.

The exact question to ask the person referring

"Do you know a good growth advisor" gets you a name. Ask this instead: "who helped you make a decision you were stuck on, and what did they actually change?" The second question is answerable only by someone who watched real work happen, and the answer tells you the shape of the engagement, not just the person. If the referrer cannot name what changed, they are recommending someone they liked, which is a different and much weaker signal.

What a real track record looks like when NDAs prevent naming clients

This is the question nobody addresses honestly, and it matters because most senior advisory work is under NDA. The advisors with the strongest track records are frequently the ones least able to describe it, and the ones with the freest hand to name logos are often naming projects they touched lightly.

Ask for the shape instead of the name. What did the business look like at the start, in real numbers. What decision did you influence, and what was the alternative being considered. What changed, and over what period. What would you do differently. An advisor with genuine experience can answer all four about an anonymised engagement in specific, checkable detail, because the specifics are what they remember. Someone who was peripheral will answer in outcomes without mechanisms, and the tell is that the story has no decision in it.

Then ask for a reference from an engagement that ended. Anyone can produce a happy current client. The engagement that finished, or finished early, is the only reference with real information in it.

Step 4: screen for
the one thing that
actually matters.

Have they run your specific problem, at roughly your scale, recently. That is the screen. Not years of experience, not the size of the logos, not the quality of the thinking on a call, all of which are easy to fake or easy to misread. The three qualifiers do most of the work: the same problem, because adjacent experience transfers less than people assume; roughly your scale, because the decisions at $5M and $50M are different decisions wearing the same words; and recently, because commerce moves fast enough that five-year-old channel experience is closer to theory than practice.

The traits and red flags worth checking beyond that screen are already catalogued in what to look for in a DTC consultant, and the difference between a platform specialist and a strategic advisor is drawn in Shopify expert versus consultant. I will not re-list them here.

The red flags, in one short list

Leads with a deck instead of questions. Cannot name a client they have said no to. Proposes a retainer before diagnosing anything. Describes results without describing mechanisms. Will not put a scope in writing. Answers "it depends" without ever saying what it depends on. Any one of these is worth a follow-up question. Two together is a pattern.

Step 5: the questions
to ask on the
first call.

Ten questions, and what a strong answer sounds like. The pattern to watch for across all of them: strong answers contain questions back at you, and weak answers contain case studies.

1. What would you do in the first thirty days, and why? Strong: a diagnostic sequence, and what each step would rule out. Weak: a list of deliverables.
2. Which of our numbers do you want before you answer that? Strong: three or four specific ones, named. Weak: "send me whatever you have."
3. What problem would you tell us not to work on? Strong: names one, and explains the opportunity cost. Weak: everything is important.
4. Who have you said no to, and why? Strong: a real example, described without disparagement. Weak: has never said no.
5. What is the last thing you were wrong about? Strong: a specific call, and what changed their model. Weak: a humblebrag.
6. How does this engagement end? Strong: has an answer ready, usually a capability handover. Weak: has not thought about it.
7. What do you need from us for this to work? Strong: access, a decision-maker, and a cadence. Weak: nothing much.
8. Where does your experience stop being relevant to us? Strong: draws the boundary clearly. Weak: claims no boundary.
9. What would make you decline this engagement? Strong: names conditions. Weak: none.
10. What did you change at a business like ours, and what was the alternative? Strong: a mechanism and a counterfactual. Weak: a number with no decision attached.

The two questions that separate operators from presenters

Numbers 2 and 3. An operator wants your data before they will commit to a view, because they have been wrong often enough to know that the answer lives in the numbers rather than the pattern. And an operator will tell you what not to work on, because prioritisation under constraint is the actual job. A presenter will answer both questions expansively and specify nothing.

What that first call should produce, and how to tell a good one from a pleasant one, is covered in what a first call with a growth advisor should produce.

Step 6: buy a small
piece of real work
first.

This is the highest-signal step in the process, and the one most often skipped. Before any retainer, buy a scoped piece of paid work: a diagnostic, an audit, a single decision worked through properly. Price it as a real engagement, not a discount. Expect it to cost somewhere between a fraction and a full month of the retainer you are contemplating.

A good advisor will usually suggest this before you do. That is itself a signal. Free strategy calls select for people whose business model is selling rather than advising, and the extended free discovery process is a sales funnel wearing a consulting hat.

What the trial actually tests, which is not competence

You will mostly know whether someone is competent from the call. What the trial tests is everything the call cannot: how they behave when they hit something they do not know, whether they change their view when your data contradicts their prior, how they write, how they handle your team, and whether working with them is a net addition or a net drain on the week. Those are the things that decide whether an eighteen-month relationship is good, and none of them are visible in a pitch.

What the engagement should cost across each shape, with real 2026 ranges, is in what a growth consultant costs.

Step 7: agree scope,
cadence and an exit
before you sign.

Four things belong in the agreement, and all four are easier to write before anyone is invested. The scope, stated as the decisions in play rather than a list of activities. The cadence, in hours and in named sessions. What you owe them, which is usually access to data and to a decision-maker. And the exit.

What "embedded" should actually mean in hours

"Embedded" is the vaguest word in this entire market and it is worth forcing into numbers. Ask what a normal week looks like: how many hours, in what shape, on what recurring calendar. An embedded retainer that turns out to be one call a month is not embedded, it is a subscription to an opinion. Write the cadence into the agreement so that nobody has to raise it awkwardly in month four.

The exit clause, and why a good advisor offers one first

Every engagement should have a stated way to end that does not require anyone to be unhappy first. A short notice period, a review date, and an agreed handover of whatever the advisor was carrying. A good advisor will usually propose this before you ask, because they have been on the other side of the relationship that outlived its usefulness and they do not want another one. An advisor who resists an exit clause is telling you that the retainer is the product.

Everything that happens after this point, the cadence, the reporting, the way to tell in month three whether it is working, is a separate discipline entirely and it is covered in how to work with a growth consultant. This page ends where that one starts.

What this looks like
when it actually
goes right.

A composite, drawn from engagements I have run and anonymised, because the specifics are under NDA. A brand comes in convinced its problem is paid acquisition, because that is where the pain is visible. The mandate is written as "our blended CAC has gone up 40% in two quarters and we do not know why." The first engagement is a two-week paid diagnostic, not a retainer.

The diagnostic finds that acquisition is fine and the real problem is that a repeat cohort quietly stopped repeating, which had been masked in the blended number for five months. Nobody had looked, because nobody owned the question. The advisor's whole contribution in that fortnight is one reframe and three numbers that nobody had put next to each other.

Then, and only then, a retainer starts, scoped to retention rather than acquisition. It runs for nine months and ends because the brand hires someone to own it permanently, which was written into the exit from the start. That is the shape of a good one: it changes what you are working on, it is smaller than you expected, and it ends on purpose.

The decision about whether outside judgment is worth buying at all is the prior question to all of this, and it is argued out in is a growth consultant worth it. If you are weighing an equity-based startup advisor rather than a paid ecommerce one, that is a different buyer with a different scorecard, set out in how to evaluate a startup advisor.