Pricing conversations with ecommerce merchants go bad in a very predictable way. The merchant asks what it costs. You give a number. They say they had a quote for less. You either hold and lose the deal, or you bend and win a client who will treat every future invoice as negotiable. The problem was not the number. The problem was that nothing in the conversation before the number made the number mean anything.
Services sold to online stores are unusually easy to commoditise. Everybody sells SEO, ads, CRO, email and design; everybody describes them in the same eight words; nobody can prove anything in advance. When two offers look identical, the buyer has exactly one variable left to compare. That is how you end up in a race to the bottom you never chose to enter.
Anchor on the merchant’s arithmetic, not on your hours
A merchant does not run a website. They run a machine with three inputs — sessions, conversion rate, average order value — and a cost of goods that eats most of the output. Every service you sell moves one of those three numbers, and the merchant already knows roughly what each one is worth to them. If you price from your day rate, you are asking them to convert your effort into their outcome, and they will do that badly and in your disfavour.
So do the conversion for them, on their numbers, before you say a price. A store doing 40,000 organic sessions a month at a 1.4% conversion rate and a 60-euro average order is earning around 33,600 euros a month from organic. Moving conversion to 1.7% is worth roughly 7,200 euros a month; adding 15% to organic sessions is worth about 5,000. Against that, a 2,500-euro retainer is not a cost, it is a ratio. The number did not change. The frame did.
You can build that frame before the first call. Monthly traffic and organic keyword counts sit in the store record; average order value you estimate from category and catalogue, or you simply ask on the call. Precision is not the point. The point is that you arrive with the merchant’s own arithmetic already on the table, instead of arriving with a rate card.
Three delivery levels, three price bands
The cleanest pricing ladder in services has nothing to do with feature tiers. It is about who does the work. Do-it-yourself: you hand over the method — an audit, a documented playbook, a set of templates, a working session. Done-with-you: you sit alongside their team and drive, they execute. Done-for-you: you take the outcome off their desk entirely.
Price rises steeply from left to right, and not because the hours do. It rises because the buyer’s perceived certainty rises. A merchant buying done-for-you believes they will get your result. A merchant buying do-it-yourself knows they will get their own execution of your idea, which they privately suspect will be worse. You are being paid for the transfer of risk, not for the transfer of knowledge.
Most freelancers sell only the middle rung and then wonder why they cannot raise prices. Build all three on purpose. The do-it-yourself tier is your qualifier — a modestly priced paid audit that separates serious buyers from tyre-kickers and gets money moving in the relationship. The done-for-you tier is where your margin lives. The middle exists to catch merchants who have an internal team and a real budget but no direction.
The equation the buyer is actually solving
Strip away the theatre and every purchase decision is one comparison: is the outcome I believe I will get worth more than the price, plus the effort it costs me, plus the risk that it does not work? Three of those four terms have nothing whatsoever to do with your invoice.
That gives you three levers before you ever touch price. Raise belief in the outcome by being specific about the mechanism — not “we will improve your SEO” but “we will rebuild your 40 thinnest category pages against buyer-intent queries, in this order, starting with the eleven that already rank on page two”. Reduce effort by taking things off their plate: migrations, briefs, approvals, the interminable back-and-forth with their developer. Reduce risk with structure: a short paid pilot, a defined first deliverable, a clean exit.
Discounting is the only lever that costs you money, and it is the one most people reach for first. When a merchant hesitates, the honest question is which of the other three terms is too high — and it is usually risk. They do not think you are expensive. They think you might not work.
Take the risk instead of taking the discount
Risk reversal in services is more flexible than the refund guarantees people copy from consumer marketing. You have better instruments. A first-month scope that is small, fixed and unmistakably deliverable. A defined kill switch after 90 days with no notice period. A split between a setup fee and a monthly retainer so the merchant’s exposure at any one moment stays small. A written definition of what failure looks like, offered by you before they think to ask.
That last one is badly underused and disproportionately effective. Telling a merchant “if we have not moved these two numbers by month four, you should fire us, and here is everything we hand over when you do” does more for your close rate than 15% off ever will. It also filters out the buyers who wanted a discount rather than a result, which is worth more than the deal you might lose.
Be careful with pure performance pricing. It sounds like the ultimate risk reversal and it works well when the metric is clean, attribution is undisputed and the merchant’s operation is stable. On a store with a broken analytics setup, heavy seasonality and a catalogue that changes weekly, revenue share degenerates into a monthly argument. Sell a hybrid: a floor that covers your cost, and a variable component on a metric you both actually trust.
Segment your prices before the merchant segments them for you
One price for every merchant guarantees that you are too expensive for half your market and too cheap for the other half. The fix is not a discount policy. It is a small number of clearly different offers aimed at clearly different store sizes.
Store data makes that segmentation observable rather than guessed. Catalogue size tells you how much work a content or feed project really is. Monthly traffic and organic keyword coverage tell you whether there is anything to optimise or whether you are starting from zero. Platform tells you what the technical work costs — a theme change on one platform is an afternoon, on another it is a project. And the detected app stack tells you what they already pay for every month, which is the closest thing to a public statement of budget you will ever get.
A store running fourteen paid apps, a premium theme and a subscription billing tool is telling you it spends money on its storefront. A store on a free theme with two apps and 3,000 monthly sessions is telling you something else entirely. Neither is a bad prospect. They are different products at different prices, and mixing them is how agencies end up taking 80% of their revenue from 20% of accounts and all of their stress from the rest.
When they say it is too expensive
Almost every price objection is a certainty objection in costume. “It is too expensive” usually means “I do not believe this will work for me”, sometimes means “I cannot see how this fits my cash flow”, and only occasionally means “I do not have the money”. Those need three different answers, and the only way to know which one you are holding is to ask.
Ask it plainly: is it the amount, the timing, or the confidence? If it is the amount and they genuinely have less, sell them the smaller product rather than the same product cheaper — that is exactly what the do-it-yourself tier exists for. If it is timing, restructure the payments without touching the total. If it is confidence, go back to the mechanism and the risk reversal, because nothing you do to the number will help.
What you must not do is defend the price by listing deliverables. Adding items to justify a number teaches the merchant that the number is a function of item count, and the very next conversation will be about removing items. Defend the price with the arithmetic you opened with, or do not defend it at all.
Know your floor and hold it
Everything above is about the ceiling. The floor is arithmetic you owe yourself: what one client of this type actually costs you to serve, including the onboarding weeks that generate no output, the support hours, the account management, and the hard fact that you can only hold a certain number of accounts before quality drops.
Divide a realistic annual revenue target by the number of accounts you can serve well, and you have a minimum. Anything below it is a client you are subsidising, and subsidised clients are rarely the grateful ones. Turning them down is not arrogance; it is the only way the good accounts get the attention that produces the results you will later use to sell more work at better prices.
Then raise deliberately rather than reactively. New prospects get the new price; existing clients get it at renewal, with notice and a reason. And if a segment starts converting at a rate that feels suspiciously high, that is not a compliment. That is the market telling you it would have paid more.