Store owners buy agency help at a predictable moment: revenue is real but growth has flattened, paid acquisition has got more expensive, and nobody in house has time to work out which of those two facts is causing the other. That is the moment an agency pitch lands well, because it promises to take the whole problem away. The buyers who do well from here are the ones who split the problem back apart before signing, because an e-commerce programme is really three purchases wearing one invoice: media you spend, labour you rent, and technical work on your own store that you keep whatever happens next. This page sets out what each part costs you, what an agency genuinely controls, and the small set of documents that make two proposals comparable.
Three purchases, one invoice
Ask any shortlisted agency to quote the three lines separately. Media is the money that leaves your account and goes to Google, Meta, Amazon or a retail network, and it is not agency income. Management is the labour: strategy, build, creative iteration, reporting, and it is charged either as a flat retainer or as a share of media. Store work is everything done to your own site, from product page structure and feed hygiene to speed and checkout friction, and it is the only part that keeps paying after the contract ends. When a single monthly figure covers all three, you cannot tell whether a proposal is expensive labour with modest spend or the reverse, and you certainly cannot tell what you are left holding if you leave. A provider who will not break the quote out is telling you something about how the relationship will be reported later.
Why a share of spend changes the advice you get
Percentage of media is the oldest pricing model in performance marketing and it is not dishonest, but its incentives are visible and worth naming out loud. If income rises with the budget, the recommendation to raise the budget arrives more easily than the recommendation to cut a channel that is not working. Flat retainers invert that pressure: labour is capped, so the temptation is to spend less time rather than more money. Neither model is disqualifying and both are common. What matters is that you know which one you are buying, that you agree in advance the point at which spend gets reviewed rather than increased, and that the person reading your reports is not the person whose fee moves with the answer. Ask what happens to the fee in a month where the honest advice is to pause a channel.
The product data layer nobody sells you
The least glamorous input to e-commerce visibility is structured product data, and it is where a competent agency earns its money quietly. Google publishes explicit product structured data requirements covering price, availability, condition and review markup, which is what allows a listing to carry a price and stock state in search results rather than a bare blue link. The same underlying feed quality drives shopping campaigns, marketplace listings and most on-site merchandising. A store with inconsistent titles, missing identifiers or stale availability will underperform in paid and organic at the same time, and no amount of creative testing fixes it. When you compare agencies, ask who owns the feed, whether they will fix it or merely report on it, and whether the fixes land in your store or in a tool you stop paying for the day you leave.
Reviews, claims and what the rules require
E-commerce marketing runs on social proof, which is exactly the area regulators have been tightening. The FTC's endorsement guidance is clear that a material connection between a reviewer and a seller has to be disclosed, and that testimonials cannot imply typical results the seller cannot substantiate. In 2024 the FTC announced a final rule aimed squarely at fake reviews and testimonials, including reviews written by insiders and suppressed negative feedback. If an agency proposes incentivised reviews, seeded testimonials or before and after claims about a product outcome, the liability lands on the merchant, not the vendor who suggested it. Get the substantiation standard written into the contract, and ask to see how a candidate has handled a client whose best selling claim could not be supported. This is also the question that separates a specialist e-commerce marketing company from a generalist who treats claims as copywriting.
Questions people ask about e-commerce marketing agencies
Should I hire one agency for everything or specialists per channel?
It depends on how much internal coordination you can carry. One agency across paid, email, and store work reduces the number of meetings and makes attribution arguments internal rather than between vendors. Specialists usually go deeper in a single channel and cost less per channel, but somebody on your side has to own the overall plan. Below roughly a small in house marketing team, a single accountable provider is usually the calmer choice.
What is a fair minimum engagement?
Ask for the minimum rather than guessing at it, because it filters a shortlist faster than any other question and takes one email to answer. Where an agency publishes a starting retainer, treat it as a floor rather than an estimate. Our index records the minimums agencies state on their own sites with the date checked, which lets you see whether your budget is inside the market before you sit through a pitch.
How long before an e-commerce programme shows results?
Paid channels give a readable signal within weeks because you control the spend and the traffic starts immediately. Organic and store work take longer and Google's own guidance warns that some changes take effect quickly while others take months to show. A proposal that promises the organic curve on a paid media timeline is either compressing the truth or planning to buy the result and call it growth.
Who should own the ad accounts and analytics?
You should, in every case. Create the accounts under your own billing and grant the agency access, rather than accepting accounts the agency created on your behalf. It costs nothing on day one and it is the difference between changing providers in an afternoon and losing your campaign history, conversion data and audience lists at the worst possible moment.