Online retail is the one category where marketing performance is visible almost immediately, which is both the attraction and the trap. Because sales can be attributed daily, agencies in this space are judged on a return figure that is easy to quote and surprisingly easy to flatter: bid on your own brand name, take credit for repeat buyers, and the dashboard looks superb while incremental revenue barely moves. Choosing well means understanding which parts of the job actually create new customers, which parts merely record them, and what evidence separates the two. This guide covers scope, pricing and the specific checks worth running before you hand anyone your ad account.
What the retainer usually covers
Most ecommerce retainers combine four things in varying proportions. Paid acquisition: search, shopping and social campaigns, plus the feed management that quietly determines whether shopping campaigns work at all. Retention: email and messaging flows, which are usually the highest-margin work in the account and the most often neglected. Organic and content: category page structure, product markup and the editorial that earns non-branded visibility. And conversion work on the site itself, from product page detail to checkout friction. Feed quality deserves particular attention, since Google's ecommerce documentation and its product structured data guidance both set out what a product listing needs to be understood correctly, and a neglected feed limits every campaign that depends on it. Ask which of the four an agency actually staffs, by name and hours, rather than which appear on the capabilities slide.
How the pricing models differ
Three shapes dominate. Flat monthly retainers are the most predictable and keep incentives neutral, which matters when the right advice is sometimes to spend less. A share of ad spend is common and quietly rewards scale rather than efficiency, so ask what happens when spend should fall. Revenue or profit share aligns interests best in principle and is hard to administer in practice, because attribution across search, social, email and marketplaces is contested and returns and margin complicate everything. Whatever the model, separate the management fee from the media budget in the proposal, and get the minimum term and notice period in writing. Also establish who owns the ad accounts, the pixel and conversion history, and the email list. Accounts held by an agency cannot travel with you, and rebuilding the learning history is a real cost paid in the first months after a switch.
The evidence that separates good from plausible
Ask three questions that are hard to answer with a deck. First, show me a client where you reduced spend and explain why. The answer reveals whether the agency optimises for the client or the invoice. Second, how do you treat branded search in reporting, since counting brand clicks as acquisition is the most common way a return figure gets inflated. Third, what is your view of incrementality testing, and have you ever run one. You are not necessarily buying a testing programme, but an agency that has never asked whether its results would have happened anyway is measuring rather than proving. Then ask for two client stores you can visit. Look at their product pages, their category structure and their email flows if you can subscribe. What is live on a real store beats every screenshot.
Getting the first ninety days right
A well-run start looks similar across agencies. Weeks one and two: access, tracking verification and a feed audit, because broken tracking invalidates everything that follows and broken feeds cap the ceiling. Weeks three to six: restructure the campaigns that are clearly misallocated, fix the highest-value product data, and get the basic retention flows live if they are missing, since abandoned cart and post-purchase sequences are usually the fastest available revenue. Weeks seven to twelve: begin the slower work on category pages, content and testing, and establish reporting that separates new customers from returning ones. Insist that the reporting runs in your own analytics and ad accounts. Retailers who also sell services or operate locally should agree early how those enquiries are counted, so the same dashboard does not mix a store order with a booking request.
Questions people ask about e commerce marketing agency
What return on ad spend should we expect?
There is no universal figure, because it depends on margin, repeat rate and category. The useful target is contribution margin after ad spend, not a raw return multiple. An agency that quotes a benchmark before asking about your margins is selling reassurance.
Should one agency handle both paid and email?
It helps, because the two decide each other's economics: paid acquisition is affordable only at the repeat rate email produces. If you split them, make sure someone owns the combined number, or each will optimise its own report while the business result stalls.
How important is the product feed?
More than most buyers assume. Titles, attributes, availability and pricing accuracy determine what your listings can match and how they display. Feed work is unglamorous and frequently the highest-leverage thing available in the first month.
What should we keep in-house?
Ownership of accounts, data and the customer list, always. Beyond that, keep whatever depends on knowing the product intimately, which is often merchandising and photography. Outsource the disciplines that need daily practice across many accounts, such as bidding and feed management.