How much does SEO cost?

As of October 2026: a poll of 439 SEO providers puts local SEO at $1,557 a month on average, agencies at $1,819 and freelancers at $1,150, and 79.1% of providers in the US and Canada charge at least $1,001 a month (Ahrefs, 2026, a poll of 439 providers read on 18 October 2026). When someone quotes you $200, they are selling something different from what you think you are buying.

What do the published surveys say?

Three sources, all with a stated sample, all read this month:

Why is the range so wide?

Because "SEO" is four different jobs sold under one word. Monitoring and reporting is one job. Fixing technical problems is a second. Producing content is a third. Earning links and mentions is a fourth, and it isn't the cheap one, because it's the only one you can't do alone in a room.

If you've paid a retainer before and watched the reports arrive while nothing moved, you already know the feeling. It isn't anger. It's the quiet suspicion that the invoice is the product, and that nobody will say so out loud.

A $500-a-month retainer buys you the first job, and part of the second if the site is small. It can't buy the fourth. That isn't a moral failing on the part of the cheap provider. It's arithmetic.

What does a cheap retainer actually hide?

Usually one of three things: a dashboard you could have got free, a contract that renews while nothing changes, or a promise about rankings that nobody can keep. We know what the end of that looks like, because the industry measures it: SEO retainers churn at about 3.2% a month, roughly 38% a year, with an average tenure of six to twelve months (Focus Digital, 2026), from a survey of agency churn published this year.

At $500 a month, a client who leaves in eight months is worth $4,000. That is less than the work of winning them. Which is why the honest cheap offers are one-time pieces of work, not monthly ones.

What should you actually pay?

It depends what you're buying, so buy the thing, not the word:

Honestly, paying more does not buy you a position. It buys hours and judgment, and the judgment is the part that improves. Anyone who promises a ranking is guessing, and the guessing gets more expensive the longer you pay for it.

How do you tell a good quote from a cheap one?

Ask for three things. What exactly happens in month one, in hours. What the same measurement will show in month three, and what happens if it doesn't. And whether you own the pages, the profiles and the data afterwards. A provider who answers all three plainly is worth more than one who answers the first and distracts on the second.

Start with the free one

We run your buyers' real questions across the AI engines and show you whether you are named, and who is named instead. Then you will know whether you need a retainer at all.

Get my free report

Next: what an AEO agency does, and what it should cost.

See also: our own measured result.

Related reading: how we work and what we measure.

Why is my dispensary not showing on Google Maps?

Check three things, in this order. Is the profile verified and complete, with hours, photos and a category that matches what you actually sell? Does your name, address and phone read exactly the same as it does on the other sites that list you? And is a shop nearer the searcher simply a better answer to that query? Most missing shops fail the first two, and both are free to fix this week.

The caveat: a corrected profile does not move overnight. Google rechecks the listing against other sources, which is why a fix made today can take a couple of weeks to show up in the map.

Related reading: the AI visibility tools worth paying for.

How these are written

Every guide here is written with The Frame Method, the structural check we use on everything we publish. It is the reason these pages do not all share one shape, and the reason they read like a person wrote them for a person.

The method works on structure rather than vocabulary. Most attempts to make writing sound less automated swap one word for another and leave the underlying shape untouched, which is why they rarely convince anyone. The Frame Method goes the other way: it looks at how a piece is built, and rebuilds the parts that give it away.

The result is writing that holds up when a reader is sceptical, which is the same writing a search engine or a language model can lift a clean answer out of. That is not a coincidence. Both audiences reward the same things: a direct answer, an honest caveat, a source you can check, and a page that does not waste their time.

We built The Frame Method for our own work before we sold it to anyone, and it is applied here on our own site first. It is one part of how these pages are made, alongside our checking and edit passes and a person reading the result.

Specialisations
AI SEOBe named by ChatGPT and Google Cannabis SEODispensaries, brands and suppliers BlogHow it works, in plain language