Short answer: Most small businesses pay between $500 and $2,000 a month for SEO. Published industry pricing surveys for 2026 put the broad market at roughly $500–$3,000 a month, with local-focused work averaging around $1,500 and agency retainers averaging closer to $3,000. Freelancers sit lower, near $1,300. Enterprise work runs $10,000 and up. The number that matters more than any of these is the one nobody prints: you will pay for roughly four to six months before the return starts, and that gap is the actual cost.
Kofi owns two dental clinics. He collected four quotes for the same brief, sent in the same email, on the same day.
They came back at $180 a month, $900 a month, $2,400 a month, and one that declined to give a number until after a discovery call. He assumed three of them were wrong.
All four were quoting honestly. They were also quoting four completely different jobs, none of which had been specified — because his brief said “we need SEO”, which is roughly as precise as telling a builder you need construction.
Illustrative composite — not a client, not a testimonial
This is the piece almost nobody writes, because printing real numbers means occasionally talking yourself out of a sale. Here they are anyway, with the uncomfortable arithmetic left in.
The ranges, without the throat-clearing
| Monthly spend | Who this is | Realistic scope |
|---|---|---|
| Under $300 | Almost nobody, honestly | A rank report and light edits. Not enough hours to fix and publish both. |
| $400–$800 | Single location, one core service | Technical foundation, keyword map, two pages a month, reporting |
| $900–$2,000 | Multiple services or locations | The above, four pages a month, local pages, competitor tracking |
| $2,000–$5,000 | Search is a primary acquisition channel | Weekly publishing, conversion tracking, quarterly strategy |
| $10,000+ | Enterprise, or genuinely brutal verticals | A team, technical SEO engineering, digital PR |
Prices are in USD because that is how the market quotes, including in markets that invoice in naira, cedis, shillings or rand. If someone quotes you local currency at a suspiciously round number, ask what exchange rate they used and when they last updated it.
What each price band actually buys
The reason quotes vary by a factor of ten is that they are pricing different amounts of human attention, and attention is the entire input.
Roughly speaking, a properly-run month for a small business contains: two to four researched pages, a technical pass over the site, position monitoring, and a report someone actually read before sending. That is somewhere between fifteen and thirty hours depending on how competitive the market is and how broken the site was to begin with.
At $180 a month, nobody is spending fifteen hours. What you are buying at that price is an automated rank-tracking email and, if you are lucky, someone occasionally changing a meta description. It is not a scam. It is just priced correctly for what it is, which is not very much.
What moves your number up or down
Two businesses of identical size get quoted $600 and $2,400 for what sounds like the same work. Usually four things explain the gap, and you can estimate your own position on each before anyone quotes you.
- How competitive your phrases are.“Emergency plumber” in a capital city is contested by businesses spending five figures a month. “Industrial valve recalibration” is contested by four companies, two of whom have not updated their site since 2019. The second is enormously cheaper to win.
- How broken your site is. A clean, fast site on a modern framework needs a week of technical work. A ten-year-old build with three plugins fighting over the same meta tags needs a month, and that month is billed before a word publishes.
- How many things you sell, in how many places. Every additional service and every additional location is another page that needs its own keyword, its own research, and its own maintenance. Cost scales with surface area, not with revenue.
- Whether anyone on your side can write. If someone in the business can produce a rough, factually correct draft, the provider is editing rather than researching from zero — and that genuinely reduces the bill.
Notice that only one of those four is about your budget. The other three are facts about your situation that exist before you talk to anybody, which is why sensible providers ask about them on the first call rather than leading with a price sheet.
The arithmetic nobody shows you
Here is the calculation that should decide this, and it takes ninety seconds.
- What is a new customer worth to you over their lifetime? Not the first invoice — the whole relationship.
- Divide your proposed monthly spend by that number.
- That is how many customers a month this has to produce to break even.
Kofi’s clinics: an average patient is worth about $600 over a few years. At the $900 quote, SEO needs to produce 1.5 new patients a month to break even. Not fifteen. One and a half.
That is a very different conversation from “is $900 a lot of money”. For a business where a customer is worth $40, the same $900 needs twenty-three new customers a month, and the answer is probably no — or at least, not yet, and not at that spend.
Now the part that gets left out. Add the ramp. Months one through three produce close to nothing regardless of who you hire, because new pages take time to be crawled, understood, and trusted. So your real question is not “can I afford $900 a month”. It is:
Can I afford to spend roughly $3,600 over four months, see almost nothing, and not panic in month three?
Month three is where most SEO engagements die. Not because the work failed, but because the flat months arrived exactly on schedule and nobody had warned anyone they were coming.
Why the $150 option is expensive
Cheap SEO is rarely a smaller version of good SEO. It is usually a different activity that shares a name, and it tends to produce one of three outcomes.
- Nothing happens. The best case. You spend $1,800 over a year and end up where you started, minus $1,800.
- Volume without judgement. Forty thin pages appear, targeting phrases nobody searches, several of them competing with each other. Cleaning this up costs more than the original engagement.
- Borrowed links. Positions improve for a while and then fall off a cliff. Recovery, where possible, takes months.
The general rule: if the price is well below what fifteen hours of skilled work costs in your market, ask specifically what is being automated and who is checking the output. There are good answers to that question. There is also no answer, which is itself an answer.
Retainer, project, or hourly
| Model | Typical range | Right when |
|---|---|---|
| Monthly retainer | $400–$5,000/mo | Ongoing content and ranking work, which has no end date |
| One-off project | $1,500–$10,000 | Technical audit, site migration, a rebuild — things that finish |
| Hourly consulting | $75–$300/hr | You have a capable team and need direction, not delivery |
| Performance-based | Varies wildly | Almost never — see below |
Performance-based pricing sounds like the buyer-friendly option and usually is not. It pushes the provider toward whatever moves the agreed metric fastest, which is rarely what grows your business. Pay for rankings and you will get rankings for phrases nobody valuable searches. The metric will be met. You will not be busier.
Currency, terms, and hiring abroad
Three practical points that get skipped in every pricing guide written for a North American audience, and which matter a great deal if you are invoicing in naira, cedis, shillings or rand.
Fixing the rate matters more than fixing the price. A USD retainer that felt comfortable in January can be materially more expensive by August without anybody changing the contract. Either agree a local-currency figure that is reviewed on a stated schedule, or agree the exchange rate source in writing. Do not leave it to be discovered on an invoice.
The cheaper international quote often is not. An overseas agency quoting below the local market usually achieves it by running the same template across every client. You can tell within one call: ask them to name a regulatory or market detail specific to where you operate. If the answer is generic, your content will be too — and content that reads as imported does not convert, however well it ranks.
Watch the payment terms, not just the number. Quarterly-in-advance at a small discount is a worse deal than monthly-in-arrears at full price if there is any chance you will want to stop in month three. The discount is being sold to you in exchange for your exit.
When you should not buy this at all
Genuinely, do not spend money on SEO this quarter if:
- Your website cannot take a booking or an enquiry properly. Fix the conversion path first. Traffic to a broken form is expensive nothing.
- Nobody searches for what you sell. Some businesses are genuinely demand-generation businesses. If the search volume is not there, no amount of optimisation invents it.
- You need customers within eight weeks. That is an ads problem, or a phone problem. SEO cannot help you at that timescale and anyone who says otherwise is taking your money.
- You cannot commit past month four. Stopping at month three is strictly worse than never starting, because you paid for all of the cost and none of the return.
Five questions before you sign anything
- How many pages publish each month, and who writes them?
- Is the technical audit included, or extra?
- What is the minimum term, and what does the exit look like?
- When we stop, what do we keep? (Correct answer: everything.)
- Which month do you expect to look flat, and what will you tell me when it does?
That last question is the one that separates providers. Anyone who has run this work honestly will answer it immediately and specifically. Anyone who has not will tell you their process is different.
Kofi took the $900 quote, on a month-to-month arrangement, after doing the 1.5-patients arithmetic on the back of an appointment card. That is the right way to make this decision — not by asking whether SEO is expensive, but by working out exactly how little it has to do to be worth it.