Is SEO Worth the Money? An Honest ROI Breakdown

SEO is worth the money when your payback month arrives before your cash runs out. Published break-even medians run six to twelve months. A local service business paying $2,150 a month typically crosses into profit around month nine, then compounds. It is a bad buy when demand, close rate, or runway is missing.

TL;DR

• The deciding number is the payback month, not the ROI percentage everyone quotes
• Your true cost runs 40% to 70% above the retainer once content, tools, and your own hours are counted
• AI Overviews cut clicks hard for informational publishers and barely at all for local services
• Filter branded traffic out before you credit anything to SEO, or the whole calculation inflates
• A worked example at $2,150 a month crosses into profit in month 9 and returns about 190% over 18 months
• Five conditions make SEO a bad buy, and three of them are timing problems that resolve

One liner: Run the arithmetic on your own numbers before you read anyone’s proposal.

Key Numbers

What “Worth It” Actually Means

In short: SEO is worth the money when the return arrives before your patience and your cash run out. That makes the payback month the deciding number, not the ROI percentage everyone quotes.

Search “is SEO worth the money” and you’ll hit the same statistic within about thirty seconds: a median SEO ROI of 748%. It’s a real figure from a real research panel. It also comes with conditions almost nobody repeats. First Page Sage attaches that number to thought leadership campaigns, meaning deep keyword research, six to eight authored pages a month, and generative optimization, with an average break-even at nine months and best-fit clients whose customer lifetime value clears $10,000 [1]. A $600 monthly local retainer is a different product. Quoting the 748% next to it isn’t a forecast, it’s a category error.

Here’s the part the percentage hides. Take two businesses that both land at 400% ROI over 24 months. Identical outcome on paper. The first breaks even in month 7 and spends the next 17 months in profit. The second breaks even in month 16, watches money leave the account every month with nothing coming back, and cancels the retainer in month 11. That business paid for a 400% return and collected none of it. Same percentage, opposite result, and the difference is timing.

Three numbers get mixed up constantly, so it’s worth separating them:

Retainers make this a recurring cash question rather than a one-time purchase. Around 78.2% of SEO providers bill through a flat monthly fee [2], so you’re committing to a cost that repeats whether or not the returns have started.

So rewrite the question. Not “is SEO worth the money.” Ask what would have to be true for search to pay you back inside twelve months. The next seven steps work that out.

Step 1: Add Up the Real Cost, Not Just the Retainer

In short: Your retainer is not your SEO budget. Content production, internal hours, tools, and amortized setup fees typically add 40% to 70% on top of the invoice, and that total is the denominator every ROI calculation below depends on.

Start with what the market charges, then stop trusting the average. A poll of 439 SEO professionals put the mean monthly retainer at $2,917, with agencies at $3,209 and freelancers at $1,348 [3]. Clutch, drawing on a much larger review base, landed near $3,199 a month. SE Ranking surveyed agencies and found 64% charging under $1,000 [4]. Those numbers disagree by more than three times, and all of them are correct. They’re measuring different slices of a market that folds solo freelancers and enterprise consultancies into one word.

So build your own number instead:

The line owners miss is the third one. Approving drafts, sitting on calls, pushing developer tickets through, answering questions only you can answer. Four hours a month at a $75 loaded rate is $300 that never appears on any invoice. Ten hours is $750.

Run it on a real quote. A $1,200 retainer covering four pages a month, plus two extra pages at $250, plus six internal hours at $75, plus a $200 tool stack, plus a $6,000 audit spread across the first year, comes to roughly $2,650 in month one and about $2,150 a month after that. The quote said $1,200. The denominator is closer to double.

The retainer is not your only option

An agency is one of three routes, and the other two have prices too.

One caution that runs against our own interest: spending under roughly $500 a month rarely buys a program. It buys a monthly report. Research on SEO services found that owners spending below that threshold were 75% more likely to be dissatisfied with results than those spending more [13]. If a quote looks far below every range above, ask what specifically gets produced each month before you take the savings, and consider whether the do-it-yourself route would serve you better than a thin retainer. For a fuller breakdown by scope and provider type, see how much an SEO expert costs.

Write your monthly total down. Step 3 divides by it.

Step 2: Discount Your Traffic Estimate for AI Overviews

In short: Cut any traffic forecast you’ve been shown, but cut it by your vertical rather than by the headline number. Informational publishers lost close to half their organic clicks. Local service businesses lost under 7%.

Every projection an agency shows you rests on click-through rates by position. Those rates changed, and the change is measured rather than guessed.

Ahrefs re-ran its study in December 2025 across 300,000 keywords and found that the presence of an AI Overview correlates with a 58% lower click-through rate for the top-ranking page [5]. That finding doesn’t stand alone. Seer Interactive measured a drop between 49.4% and 65.2%. Kevin Indig found above 50%. Authoritas landed at 47.5%. Pew Research Center, which sells nothing in this market, found users clicked 8% of the time when an AI Overview appeared against 15% when it didn’t [7]. Four independent methodologies, one band.

Most of that work is correlational. One randomized field study went further and removed top-position AI Overviews for some participants. Outbound clicks nearly doubled for that group, which isolates cause instead of association [8].

So the effect is real. The question is whether it’s your effect.

Is SEO dead? Not evenly

The damage sorted itself by what kind of query a business depends on. A 2026 panel measuring twelve industries found media and publishing absorbed the steepest organic decline at 47.3%, while local services sat at the bottom of the table at 6.8%, real estate at 9.2%, and e-commerce and retail at 11.6% [6]. The reason is mechanical. An assistant can finish a general-knowledge question. It cannot finish a booking, a quote, or a directions query.

That last row matters more for small businesses than any other number in this article. Roughly 46% of all search queries carry local intent, based on a figure Google published in 2018 and still cited as the best available estimate [14]. A large share of search demand is the kind assistants have not absorbed.

If you sell a local service, the honest discount on your forecast is single digits, not half. If you run a content site funded by informational traffic, the honest answer to whether SEO is worth the money may be no, and Step 7 covers that case.

There’s a second adjustment, and it runs the other way. The same panel found AI Overviews cut a page from 87 clicks per thousand impressions down to 36, but the visitors who still clicked stayed 23.0% longer and converted to leads at a higher rate [6]. Fewer sessions, better sessions. For an ROI calculation that’s a smaller top number and a larger conversion rate, not a straight loss.

Check your own numbers instead of trusting the averages

Open Google Search Console, set the performance report to compare a recent quarter against the same quarter in 2023, and look at impressions next to clicks. Stable or rising impressions paired with a falling click-through rate at an unchanged average position is the signature. A page holding position one that fell from 6% CTR to under 1.5% has been summarized.

That percentage is your discount. Apply it in Step 3.

Step 3: Run the Formula on Your Own Numbers

In short: How to calculate SEO ROI: multiply organic sessions by conversion rate, close rate, and gross margin per deal, subtract your total cost from Step 1, then divide by that cost. The formula takes a minute. Sourcing the inputs honestly is the work.

Here it is:

ROI = ((Sessions x Conversion rate x Close rate x Margin per deal) – Cost) / Cost

Every ranking page prints some version of that. Almost none of them tell you where the numbers come from, which is where the answer gets decided. Four of the five inputs have a default that quietly inflates the result.

The branded-traffic filter matters more than the other four combined. If someone types your company name into Google and lands on your site, SEO didn’t earn that customer. Your truck, your referral, or last year’s marketing did. Agency ROI reports that skip this filter credit organic search with people who already knew the business existed, and on an established brand that can be most of the traffic. In Search Console, add a query filter excluding your company name and its common misspellings, then re-read the number. It usually drops hard, and what’s left is the traffic SEO actually built.

Margin is the second correction. A contractor closing a $10,000 job at 35% gross margin puts $3,500 in the business, not $10,000. Run ROI on revenue and you’ll overstate the return by roughly three times. Use the margin figure and the answer stays honest even when it stops flattering the channel.

Declare your attribution model before you run anything. Multi-touch is closer to reality than first-touch, because organic frequently assists a sale that closes through a phone call, a referral, or a retargeting ad. Whichever you pick, pick it once and hold it, since switching models mid-analysis is how a mediocre channel starts looking excellent.

One honest limitation. In 2026 the formula undercounts. When an AI assistant names your business inside an answer and the reader contacts you later without ever landing on your site, that influence never appears as a session. Nobody has a clean way to measure it yet. It means a marginal ROI figure deserves the benefit of the doubt, and it does not rescue a badly negative one.

Run the formula. Hold onto the number. Step 4 turns it into a date.

Step 4: Turn ROI Into a Payback Month

In short: Published break-even medians cluster between six and twelve months, with local campaigns at the fast end and competitive or high-value verticals at the slow end. Find your projected payback month, add three, and ask whether your cash survives that long.

An ROI percentage describes a finished campaign. A payback month describes the one you have to fund.

Here’s what the published benchmarks say, each with the scope it was measured under:

Read those as benchmarks, not forecasts. All of them come from research published by companies that sell SEO, including ours by extension, and none of them knows your close rate. Most retainers assume a six to twelve month minimum term [4], which tells you the industry’s own working assumption about when this starts paying.

The mechanic underneath is what matters. Your cost starts in month one and stays flat. Your return starts near zero and compounds, because rankings build on rankings and content keeps earning after it’s published. Maturity is part of why the front end is slow: the average page sitting on the first results page is roughly five years old [12], so a page published this month is competing against pages that have been accumulating signals for years. So the gap between the two widens before it closes. The deepest point of the hole usually falls somewhere in months four through seven, when you’ve paid five or six invoices and the traffic curve has begun to move but the revenue curve hasn’t caught it yet.

That shape is normal. It’s also exactly when retainers get cancelled.

Go back to the two businesses from the top of this article, both landing at 400% ROI over 24 months. The one that broke even in month 7 saw the line cross while it still had patience. The one that broke even in month 16 was staring at the deepest part of the hole in month 11, saw eleven invoices paid and not much back, and cancelled. It was three months from the curve turning. The percentage never told it that, because a percentage has no dates in it.

That gap between what owners expect and what happens is measurable. A 2026 study of small business digital marketing found that 68.8% of SMB owners expect significant SEO results within three months or less [11]. Set that against break-even medians of six to twelve months and the cancellation pattern stops looking like impatience. It looks like a forecasting problem that nobody corrected at the start.

So run this test before you sign anything. Take your projected payback month, add three for slippage, and ask whether your cash position survives that many months of full cost with partial return. If it doesn’t, the honest answer isn’t that SEO is a bad investment. It’s that this isn’t the right quarter to start one, and Step 7 covers what to do instead.

One more thing the benchmarks agree on: peak results land in year two or three, not year one. That cuts both ways. The good part is real, and it arrives after twelve months of paying for it.

Be skeptical of anyone promising positive ROI in 90 days on competitive queries. Nothing in the measured data supports it.

Step 5: A Worked Example, Month by Month

In short: A local service business paying $2,150 a month crosses from negative to positive in month 9 and finishes 18 months at roughly 190% ROI. Every input below is a stated assumption, not measured client data.

Numbers first, so nothing hides. This example uses the cost stack built in Step 1, the local-services traffic discount from Step 2, the input rules from Step 3, and it reports its crossover the way Step 4 asked for.

Is SEO worth it for a small business? Here’s one, in full

One rule does most of the work here. Only sessions above the 180-a-month baseline count. That traffic existed before anyone was paid, so crediting it to the retainer would inflate the result before the arithmetic starts. Every figure below counts the lift and nothing else.

At those assumptions each incremental session is worth $12.25 in gross margin. Here’s how it runs:

The line crosses in month 9, at a net of $118. That’s the whole payback event, and it’s an underwhelming moment. Eight months of paying, and the reward is being level.

Notice where the hole bottoms out. Month 5, at negative $7,820. That’s the deepest this business ever goes, and it happens while traffic is climbing and nothing looks broken. An owner reading their bank balance in month 5 sees the worst number of the entire campaign four months before the payoff.

Over 18 months the return works out to about 190% on $39,200 spent. Below the 748% figure everyone quotes, and that’s the point. This example assumes an average close rate and a mid-market job value, with no heroics anywhere.

Change one input and watch what happens

Drop the close rate from 25% to 15% and hold everything else steady. Crossover moves from month 9 to month 13. The 18-month return falls from about 190% to about 74%.

Nothing about the SEO changed. Same traffic, same conversion rate, same spend. What changed was how well the business handled the phone. That’s the lever most owners never look at, and it moves the answer more than the retainer does.

Run your own numbers through the same shape before you decide the channel is worth it or isn’t.

Step 6: Run the Same Money Through Paid Ads

In short: Paid search buys leads this week and stops the day the card does. SEO buys nothing for months and keeps working after you stop paying. The right answer depends on when you need the customers, not on which channel is better.

“Is SEO worth the money” is incomplete as a question. Worth it compared to what? Doing nothing, hiring a salesperson, and running ads are three different comparisons with three different answers. Ads are the one most owners are weighing, so run the same arithmetic through both.

SEO vs paid ads, side by side

Before going further, one number needs correcting. You’ll see it everywhere: organic leads cost about $31, paid leads about $181, usually credited to HubSpot. The figures trace to First Page Sage’s cost-per-lead-by-industry research, not HubSpot, and the underlying panels run from January 2021 through January 2024 [9]. That’s pre-AI Overview data being quoted as a current fact in 2026. The direction it points is probably still right. The precision is not, and the organic side of that ratio has not been re-measured since assistants started answering queries.

What is current is the trajectory on the paid side. An analysis of more than 3,000 home services search campaigns found cost per lead rose for 69% of advertisers year over year, at an average of 10.51%, against a cross-industry average of 5.13% [10]. That figure is specific to home services, and the mechanism behind it is not. Paid search is an auction, and auctions get more expensive when demand grows. Organic moves the other way, since a page that ranks keeps ranking without a per-click charge.

That’s the structural difference. It’s also why the comparison shouldn’t be a coin flip.

Paid wins outright in four situations. A brand new business with no domain history and no rankings to build on. A seasonal or time-boxed offer that has to convert inside eight weeks. A market you’re testing before committing to a twelve-month program. And any business whose cash can’t survive the payback month from Step 4. In all four, ads are the correct purchase and SEO is the wrong one for now.

Most businesses that can fund both run them together, using paid to cover the gap while organic matures, then tapering the ad spend as rankings carry more of the load. That’s less satisfying than picking a side, and it’s what the arithmetic usually supports.

Step 7: Five Conditions That Make SEO a Bad Buy

In short: Three of these are timing problems that resolve. Two are structural and don’t. If any of the five describes your business right now, the honest answer to the title question is no.

An SEO company wrote this list, which is worth stating out loud. The incentive isn’t as crooked as it looks. A client who signs under any of these five conditions cancels inside six months, blames the agency, and tells other business owners SEO doesn’t work. Nobody wins that sale.

Here are the five.

1. Your cash can’t survive the payback month plus three.
The test from Step 4. In the worked example the deepest point was month 5, at negative $7,820, with the crossover four months later. If a stretch like that would force you to cancel partway through, you’d be buying the cost without the return. That doesn’t mean doing nothing. It means the do-it-yourself route priced in Step 1 is the better use of the money this year. Not yet, rather than not ever.

2. Your traffic depends on informational queries that assistants now answer.
From the vertical table in Step 2. Media and publishing lost 47.3% of organic sessions, and the decline hasn’t plateaued. If your model is general-knowledge content earning ad or affiliate revenue, the ground under the channel moved. Structural.

3. Nobody is searching for what you sell.
SEO captures existing demand. It doesn’t create it. Run two checks in a keyword tool before you go further. First, your service plus your city, which tells you whether people search for the thing you actually sell. Second, the problem your customers have before they know your category exists, since a business selling something genuinely new often has no category demand while the underlying problem has plenty. If the first comes back empty and the second comes back full, SEO still works, and it works through content about the problem rather than pages about the product. If both come back near zero, ranking first for nothing is still nothing. Structural, though a genuinely new category can grow into demand later.

4. Your close rate is the real bottleneck.
From the sensitivity check in Step 6. Dropping the close rate from 25% to 15% pushed the crossover from month 9 to month 13 and cut the 18-month return from about 190% to about 74%, with no change to the SEO at all. If leads already arrive and don’t convert, more traffic buys more of a problem. Fix the follow-up first, then the return on every channel improves at once. Not yet.

5. You need customers inside eight weeks.
Seasonal offers, a lease that runs out, a launch with a date on it. Step 6 covers what to do instead. Not yet.

Three of these five say wait. Two say the channel doesn’t fit the business. Either way, knowing which one you’re in is worth more than a proposal.

How to Check the Math on an Agency’s Proposal

In short: Six of these seven questions test the arithmetic. The last one tests the seller, and it’s the one that tells you the most.

You now have everything you need to read a proposal properly. Here’s what to ask, and what the answers should sound like.

1. Is this traffic projection filtered to non-branded queries?
Good answer: yes, here’s the filter and here’s what the number was before and after.
Worrying answer: any version of “all organic traffic counts.” That projection includes people who already knew your name.

2. What AI Overview discount did you apply, and at what rate for my vertical?
Good answer: a specific percentage, with a reason it’s that number for this industry.
Worrying answer: none applied, or click-through curves that predate 2024. That’s a forecast built on a search engine that no longer exists.

3. Which click-through-rate dataset did you use, and from what year?
Good answer: a named source with a date, and an acknowledgement that the figure is a range.
Worrying answer: a single confident percentage. Published position-one click-through rates range from roughly 14% to nearly 40% depending on the dataset, and two agencies quoting different studies will hand you forecasts that differ by three times for the same keyword.

4. What close rate and margin are you assuming, and where did you get them?
Good answer: from your CRM, segmented by lead source, with margin rather than revenue.
Worrying answer: an industry average. Then the projection describes a hypothetical business, not yours.

5. Which month does this break even?
Good answer: a month, with the cumulative cost and return on either side of it.
Worrying answer: an ROI percentage and no date. That’s the flattering half of the picture.

6. What’s included, and what gets billed on top?
Good answer: an itemized scope naming pages per month, tools, and whether the audit is separate.
Worrying answer: a single monthly figure with “full-service” attached and no line items.

7. What would make you tell me not to buy this?
Good answer: specific conditions, readily named, resembling the five in Step 7.
Worrying answer: nothing would. A provider with no disqualifying conditions has one product and sells it to everybody who calls.

Question 7 carries the most signal because it’s the only one that can’t be answered with a spreadsheet. Anyone can produce a filtered projection when asked. Naming the customers you turn away costs something.

One more thing to watch in the reporting. Impressions, keyword counts, and domain rating are inputs, not outcomes. Ranking fourth for a term nobody searches is a real report of nothing. Ask what the monthly report contains, and check that leads appear somewhere in it alongside the rankings. Ask about citation visibility in AI answers too. When an assistant names your business in a response, that is reach the session count cannot see, and it belongs in the report next to rankings and leads rather than in a separate conversation.

So, is SEO worth the money? It’s worth it when your vertical still gets clicks, your demand exists, your close rate works, and your cash reaches the crossover month. When those four hold, the arithmetic in Step 5 tends to land somewhere between decent and excellent. When one of them doesn’t, no proposal fixes it. Run your own numbers first, then go read the proposal knowing what it should say.

Frequently Asked Questions

How do I calculate SEO ROI?

Multiply your organic sessions by your conversion rate, your close rate, and your gross margin per deal. Subtract your total SEO cost, then divide by that cost. The formula is short. The honesty lives in the inputs. Filter sessions to exclude branded queries, count only real conversions rather than newsletter signups, pull the close rate from your CRM by lead source, and use margin instead of revenue. Skipping any one of those four inflates the result.

How long does SEO take to pay off?

Published break-even medians cluster between six and twelve months, with local campaigns at the faster end and competitive verticals at the slower end. Treat those as benchmarks rather than promises, and note that the research comes from companies that sell SEO. The worked example in this article crosses from negative to positive in month 9. Peak returns typically arrive in year two or three, not year one.

Is SEO dead in 2026?

No, but the damage is uneven. Ahrefs found that AI Overview presence correlates with a 58% lower click-through rate for the top-ranking page as of December 2025, and independent studies from Seer, Authoritas, and Pew Research land in the same band. The split by industry matters more than the headline. A 2026 panel measured media and publishing down 47.3% in organic sessions, while local services fell 6.8%, because map and proximity queries still require a click. If your business depends on informational traffic, the ground moved. If you sell a local service, it barely shifted.

When is SEO not worth the money?

When your cash cannot survive the payback month, when your traffic depends on informational queries that assistants now answer, when nobody searches for what you sell, when your close rate is the real bottleneck, or when you need customers inside eight weeks. Three of those are timing problems that resolve later. Two are structural and do not. Knowing which one applies is more useful than any proposal.

Is SEO or paid advertising better for a small business?

Neither is better. They run on different clocks. Paid search delivers leads within days and stops the day you stop paying. SEO delivers nothing for months and keeps working after the spend ends. Paid is the right buy for a new business with no domain history, a seasonal offer, a market test, or thin cash. Most businesses that can fund both run them together and taper the ad spend as organic rankings mature.

How much should a small business spend on SEO per month?

Survey data puts the average monthly retainer near $2,917, with agencies around $3,209 and freelancers around $1,348, though published averages disagree by more than three times depending on who was surveyed. Most small and midsize businesses land between $500 and $5,000 a month. Below roughly $500 you are usually buying a monthly report rather than a program, and research found owners spending under that threshold were 75% more likely to be dissatisfied. Remember that the retainer is not the full cost, since content, tools, and your own hours add to it.

Can I do SEO myself instead of paying someone?

Partly, yes. Claiming and maintaining your Google Business Profile, writing pages that answer real customer questions, fixing broken links, and keeping your site fast are all things an owner can do without an agency, and they matter. What tends to need help is technical diagnosis, competitive content strategy, and earning links. Freelancers average around $71.59 an hour and an in-house specialist costs roughly $71,000 a year, so if your budget is under a few hundred dollars a month, doing the basics yourself is a better use of the money than a cheap retainer.

Why does SEO take so long to work?

Search engines need to crawl new pages, and rankings for competitive terms depend on signals that accumulate rather than switch on. The average page ranking on the first results page is roughly five years old, so a page published today competes against pages that have been earning links and engagement for years. Traffic also arrives before revenue does. Someone who reads your page in month 4 may not need your service until month 7, which stretches the gap between when the work happens and when the money shows up.

What should an SEO report include?

Organic sessions filtered to non-branded queries, conversions, and leads or revenue attributed to organic search, alongside the rankings. Citation visibility in AI answers belongs there too, since an assistant naming your business is reach the session count cannot see. Impressions, keyword counts, and domain rating are inputs, not outcomes. A report that shows only those is describing activity rather than results.

Definition Bank

Entity Cards

Google AI Overviews

SEO Retainer

Payback Period

Sources

1. First Page Sage, SEO ROI Statistics 2026, accessed August 2026.
2. Ahrefs, SEO Pricing Survey, 439 SEO professionals, 2026.
3. Ahrefs, SEO Pricing Survey, average retainer and hourly rates by provider type, 2026.
4. SE Ranking agency pricing survey and Clutch SEO pricing data, as reported 2026.
5. Ahrefs, AI Overviews Reduce Clicks, December 2025 update, 300,000 keyword study.
6. First Page Sage, AI Impact on Website Traffic Report, 2026.
7. Pew Research Center, click behavior with and without AI Overviews, 2025.
8. Randomized field study on AI Overview removal and outbound clicks, as reported by Search Engine Journal, April 2026.
9. First Page Sage, Average Cost Per Lead by Industry, organic versus paid, panels January 2021 to January 2024.
10. LocaliQ, analysis of 3,211 home services search advertising campaigns, 2025.
11. HigherVisibility, Small Business Digital Marketing Study, 2026.
12. Ahrefs, How Long Does It Take to Rank, page age analysis.
13. Backlinko, SEO Services Statistics, spend threshold and client satisfaction.
14. BrightLocal, Local SEO Statistics, citing a Google figure published 2018.