How to Judge an Ecommerce SEO Agency Before You Pay One

I read the pages ranking for “best ecommerce seo agency” before writing this. The highest‑placed article is a ranked list published by an agency that put itself at number one and openly says it is not a neutral directory. The two service pages under it are sales pages. Between all three there is no price, no contract term, and no description of a store that should not hire anybody.

That absence is not an oversight. Ahrefs puts this query at roughly 350 searches a month in the United States, a keyword difficulty of 7, and a cost per click of $9.00 — cheap to rank for, expensive to buy. The person typing it has a budget and is deciding where to put it, which is the last person an agency wants to hand a checklist that might disqualify them. So here is that checklist, written by someone who is one person and will tell you when that is the wrong shape for the job.

What does an ecommerce SEO agency cost?

Nobody on page one will tell you, and the reason is that the right number depends almost entirely on how much non‑branded organic revenue is actually at stake. Size the retainer against the revenue it is meant to move, not against the agency’s headcount.

The arithmetic takes two minutes and you can do it before you speak to anyone. Open your analytics and find monthly revenue from organic search. Then split it: the part that arrived on searches containing your brand name, and the part that did not. Only the second number is in play — the first is demand you already created and would largely keep without any agency at all.

Say organic brings $40,000 a month and a quarter of that is non‑branded. Use your own figures; the shape is what matters. That puts $10,000 a month in the addressable column. A program that lifts it by half is worth $5,000 a month once it lands, and it will not land for months. Against that, a $3,000 retainer is a reasonable bet and a $9,000 one is not, and you have worked that out without a single proposal in front of you.

The four pricing models, and what each one rewards

Four models are in common use and each one quietly rewards a different behaviour. None is a trap; the trap is buying one without knowing what it pays the agency to do.

The last row is where most disputes start. If you sign a performance deal, define the metric as non‑branded organic revenue in writing, or you are buying back traffic you already had.

What separates an ecommerce SEO operator from a generalist?

An ecommerce SEO operator works on the catalog; a generalist works on the blog. That is the whole distinction, and it survives contact with every proposal I have been asked to review.

A store with 800 products does not have 800 pages. It has products, variants, collections, filtered versions of collections, paginated versions of those, tag pages, internal search results and vendor pages — and Google will find most of them. The work that matters is deciding which templates should exist, which should be consolidated, what happens to a product URL when the product is discontinued, and how a collection page earns a category term. It is unglamorous, mostly deletion, and it is where the revenue is.

A generalist proposal is recognisable in one line: the deliverables are a number of articles per month and nothing about the catalog. Blog content has a job in an ecommerce program, but it is rarely the first thing to invest in, and an agency that opens there has either not looked at your store or cannot do the other work. Google’s own Search Engine Optimization (SEO) Starter Guide is explicit that this is slow work — “Some changes might take effect in a few hours, others could take several months” — which is exactly why the order you do it in decides the outcome.

The five questions that expose a generalist

Ask these on the first call, before anyone sends a proposal. You are not testing whether they know the answer; you are testing whether they ask to see the store first.

1. “What should happen to my filtered collection URLs?” A real answer names crawling and indexing before it names rankings, and refuses to commit until somebody has seen how your filters build URLs. A generalist says they would optimise them.
2. “What do you do with a product page when the product is discontinued for good?” Redirect, keep and reroute, or retire — there is no single right answer, only a wrong one, which is not having thought about it. For a store that churns seasonal stock it is the biggest recurring SEO decision you have.
3. “Which of my pages would you delete?” Everyone will tell you what they would add. An operator has a view on subtraction, and will want the data before naming anything specific.
4. “Show me an account where rankings improved and revenue did not.” Not a trap. It establishes whether they have ever looked at that pairing, because it happens constantly and an agency that has never seen it is measuring the wrong thing.
5. “Will you report on branded or non‑branded performance?” If the monthly report is total organic sessions, you will be shown your own brand equity every month and invited to call it progress.

On that last one, know what the numbers mean before they are presented to you. Search Console’s documentation on what impressions, position and clicks are defines average position as a relative ranking averaged across all impressions — so a batch of new pages surfacing at position 40 for a few thousand queries can move the account average on a slide without any of it being worth money.

When is an agency better than a consultant?

An agency beats a consultant whenever the constraint is production capacity rather than judgement. That is a real and common situation, and pretending otherwise would be a sales pitch rather than an answer.

• The work needs many hands at once. A migration with a hard date, a catalog rebuild, or a content program running across several categories in parallel. One person cannot be the bottleneck on a deadline that is not moving.
• You have no developer. Most SEO findings are somebody else’s tickets. An agency with its own build team closes the loop; an advisor hands you a list and the list sits there.
• Multiple markets or languages. Volume of routine work, and routine work is what teams are for.
• You need continuity through your own staff turnover. An agency survives its own people leaving as well as yours, which is a genuine advantage nobody names out loud.
• Procurement requires it. Insurance thresholds, vendor onboarding, a finance department that will not raise a purchase order for an individual. Unglamorous and decisive.

A consultant wins when the constraint is knowing what to do rather than doing it, when you want the person who diagnosed the problem to be the person still on the account in month nine, and when the budget is small enough that an agency would staff it juniors‑first. The fuller version of that trade‑off is in in‑house versus agency versus consultant.

Who should not hire an ecommerce SEO agency?

Four situations where the correct move is to keep the money, and any agency worth hiring will tell you if one of them applies to you.

Where I am the wrong choice

I am one person, so the first bullet in the agency list above is the honest disqualifier. If you need a content program producing serious volume every month, a build team on standby, or three markets covered at once, hire an agency and do not let anyone talk you out of it.

What one person buys instead is that whoever looked at your catalog is still on it a year later, and that nothing gets quietly reassigned to the person who joined in March. Shurhold has been a client for 7 years, which is the only kind of evidence that means much here — not a case study, a renewal, repeated.

One more thing worth saying plainly, because the rankings screenshot is the standard closing move in this industry. I have an account where organic rankings improved sharply and the organic revenue has not followed yet. That is on the results page, in those words, because the alternative is showing a position chart and letting you assume it meant sales. Ask every agency you meet for their equivalent, and watch what happens.

One thing worth doing before any of those calls: know what is actually wrong with the store. SEO Karbon audits a Shopify store from the outside, free and without a login, and ranks the findings by the revenue at risk. Walking into a pitch already knowing your own weakest dimension changes the conversation, because you can hear whether the proposal addresses it or avoids it.

What should I check before I sign?

Five things, and they take one email to establish.

1. Who implements. Whether the agency changes the theme and the product data itself or files tickets for your developer. That answer decides how long everything takes and it is rarely in the proposal.
2. Who owns what. Content, the Search Console and analytics properties, and any tooling. All of it in accounts you own, with the agency invited in.
3. What the report contains. Non‑branded clicks and revenue, named as such. Agree it before month one so there is nothing to argue about in month six.
4. Notice period, and what happens at the end. A long initial term is defensible for slow work; automatic renewal with a long notice window is not.
5. Any guarantee at all. Google’s guidance on whether you need an SEO puts it bluntly: “No one can guarantee a #1 ranking on Google. Beware of SEOs that claim to guarantee rankings, allege a special relationship with Google, or advertise a priority submit to Google.” Treat a ranking guarantee as the end of the conversation.

Google also suggests asking for examples of previous work and what results to expect in what timeframe. Whether that timeframe comes back in months, stated without hedging, tells you most of what you need. The rest of what I publish about Shopify SEO starts from the same place: the cheapest honest fix first, whether or not it looks like strategy.

Questions people ask