The average SEO pitch to an ecommerce merchant is a technical document delivered to somebody who does not care about technical documents. Crawl budget, canonical tags, internal linking depth: all real, all irrelevant to the person deciding whether to sign. The merchant is running a business with a margin, and every euro of your retainer comes out of it.
This is not an argument for dumbing anything down. Merchants who run stores at scale are perfectly capable of following a technical explanation. It is an argument for changing which end of the chain you start from, because the end you start from decides what the conversation is about.
Merchants buy revenue per session, not rankings
Ask a merchant what their SEO is worth and most will not know. Ask what a session from Google is worth and many can tell you within a few cents, because they know their conversion rate and their average order value. That number is the entire conversation, and it belongs on the table in the first five minutes.
Frame everything in it. A ranking improvement is an abstraction; 4,000 additional monthly sessions on buyer-intent queries, multiplied by their revenue per session, is a forecast they can check against their own accounts. You will be wrong sometimes, which is why you give ranges and state your assumptions — but a wrong forecast that engages with their arithmetic beats a technically correct audit that does not.
This also changes what you propose. If revenue per session is the unit, then queries that convert are worth several multiples of queries that merely inform, and the plan should reflect that ratio brutally. Merchants notice when a proposal is mostly blog posts, and they are right to be suspicious about it.
Know the trajectory before the call
The single most useful thing you can bring to a first conversation is a picture of the store’s organic history that the merchant has not drawn for themselves. With roughly 170 monthly data points of organic traffic and positions per store, the shape is visible instantly: a steady climb, a long plateau, a cliff in one specific month, or a slow bleed nobody noticed.
Each shape is a different conversation. A cliff means something happened — a migration, a redesign, a platform change, an algorithmic hit — and the pitch is diagnosis and recovery, which is urgent and therefore easy to fund. A plateau means the current approach has hit its ceiling, and the pitch is a new surface: categories they do not have, intents they do not cover. A slow bleed is the hardest sell, because nobody is panicking yet, and the job is to make the trend line visible enough that they start.
A steady climb is a real answer too. Sometimes the honest read is that the store is doing fine and does not need you this quarter. Saying so buys more credibility than any deck, and merchants who are growing tend to remember exactly who told them the truth when they eventually plateau.
Category pages are the product; blog posts are the accessory
Ecommerce SEO money sits in the pages that can take an order. Category and collection pages, filtered views that map to how people actually search, and product pages for terms with genuine demand. Informational content supports those pages; it is not a substitute for them, and a proposal built primarily on it is usually a proposal built around what is easy to deliver rather than what is worth buying.
The diagnostic is straightforward and you can do most of it before the merchant hands you anything. Compare catalogue size against organic keyword footprint. A store with 900 products and a few hundred ranking keywords has a coverage problem: whole intent clusters have no page pointed at them. A store with 60 products and thousands of ranking keywords has the opposite situation and needs conversion work, not more content.
Then look at intent. Rankings concentrated on informational terms with a thin transactional footprint is a common and expensive pattern — traffic that looks healthy in a monthly report and does almost nothing to revenue. Rebuilding around purchase-intent queries can reduce sessions and increase revenue at the same time, and you should warn the merchant about that in advance, or the first report meeting will be a difficult one.
AI visibility belongs in the conversation now
Merchants are asking whether assistants recommend them, and most agencies still answer with either a shrug or a buzzword. Treat it as a measurable surface instead: how often the brand is mentioned in generated answers, and which of its pages get cited. Both are countable, and they behave differently from classic rankings.
The practical point for a pitch is that the two surfaces are correlated but not identical. A store can rank respectably and still be close to invisible in generated answers, usually because its pages are structured for a crawler rather than for extraction — no clear specifications, no comparison content, no direct answers to the questions buyers actually type. That gap is a concrete, sellable piece of work with a before-and-after you can show.
Do not oversell it. Nobody has a decade of evidence about what moves AI citation counts, and merchants can tell when a channel is being pitched with more certainty than it deserves. Position it as a measured experiment with a proper baseline, not as a guaranteed channel with a guaranteed return.
The proposal: baseline, hypothesis, horizon
Three elements make an ecommerce SEO proposal signable. A baseline: the current numbers, stated by you, from data the merchant can verify — traffic, keyword footprint, the trend shape, current AI mentions and cited pages. If your baseline is wrong, they will discover it in week two, so pull it from something real rather than from a screenshot.
A hypothesis: what you believe is limiting the store, in one sentence, and what you intend to do about it in a defined order. Not a list of thirty tasks. “Your catalogue outgrew your category structure two years ago; we rebuild the 40 highest-demand missing categories, starting with the eleven where you already rank on page two.” A merchant can argue with that sentence, which is precisely why it converts.
A horizon: when each thing should show, with the uncomfortable parts said out loud. Technical fixes in weeks, category work in two to four months, authority work longer than either. Merchants have usually been burned by somebody who promised results in thirty days, and a proposal that says “nothing meaningful will move before month three” reads as competence rather than weakness.
Structure and price the engagement honestly
The three-month contract is where ecommerce SEO relationships die. It is long enough for you to have spent all the setup effort and short enough that nothing has compounded, which produces a merchant who concludes SEO does not work and an agency that never got paid for the interesting part of the job.
Split it. A paid diagnostic first — fixed price, two to four weeks, real access, ending in a plan the merchant owns whether or not they continue. Then a retainer with a defined initial horizon and a clean exit. The diagnostic filters out merchants who wanted a miracle for 300 euros a month, and it starts the relationship with a delivery instead of a promise.
Price the retainer against the arithmetic you opened with, and vary it by store size using observable things: catalogue size, traffic band, platform, whether they have internal developers. A 40-product store and a 4,000-product store are not the same job, and pretending otherwise is how agencies end up losing money on their largest accounts while congratulating themselves on the logo.
The objections you will actually get
“We tried SEO and it did not work.” Usually true, and usually because it was three months of blog posts. Ask what was delivered and on which pages. The answer is nearly always informational content and no structural work, which lets you draw the distinction without criticising the previous provider — merchants dislike agencies who trash their predecessors more than they dislike the predecessors.
“How long until we see results?” Give the honest horizon, then give them something to see before it arrives. A small technical fix or a single rebuilt category page that moves inside six weeks is worth a great deal politically, because it gives the merchant something to show internally while the slow work compounds in the background.
“Can you guarantee positions?” No, and say so flatly. Then offer the guarantee you can actually honour: a defined scope, a defined review point, a clean exit, and everything you produce staying theirs regardless. A merchant who still insists on position guarantees after that is telling you they will be a difficult client, and that is useful information delivered early and for free.