How to Measure SEO ROI for a Service Business (2026)

An SEO retainer is the only line item on your P&L where the vendor grades their own homework. The agency picks the metrics, builds the report, and every month the rankings chart points up and to the right — while you stare at the schedule wondering which of those rankings ever became a booked job. This guide is the correction: the exact math, from call tracking to close rates to lifetime value, that tells you whether SEO paid for itself.

We publish it because it is the same scorecard we ask clients to hold us to — here is how you will know whether we earned the retainer. If you are still deciding what to spend, start with how much SEO costs a service business. If you are already paying and the answer is fuzzy, keep reading. The section on when the math says fire your agency is the one your current provider hopes you skip.

How do you calculate ROI on SEO?

SEO ROI = (revenue attributed to organic search − total SEO cost) ÷ total SEO cost × 100. A $2,500/month retainer costs $30,000 a year, so $30,000 in organic-attributed revenue is 0% ROI — break-even — and $90,000 is 200%. For a service business, "revenue attributed" means closed jobs traced back to organic leads. Not sessions, not rankings, not a tool's "estimated traffic value."

The formula is trivial. The inputs are where every measurement effort dies, because a lead-gen business has no checkout: revenue arrives by phone call, form fill, and booked estimate, often weeks after the click. The real work of measuring SEO ROI is building the trace from search to signed invoice. Five numbers make it possible:

• Total SEO cost — retainer plus tools, content, and your own hours (covered next)
• Organic leads per month — calls and form fills traced to organic search specifically
• Lead-to-close rate — the percentage of those leads that become paying jobs
• Average job value — revenue per closed job, ideally per service line
• Customer lifetime value (LTV) — repeat and referral revenue over 3–5 years, the multiplier most owners forget

What counts as the true cost of SEO?

True SEO cost = retainer + content you buy separately + tools + your own time. For most service businesses on a $2,500/month program, real annual cost lands around $32,000–$38,000, not $30,000. If you spend three hours a month approving content and sitting on strategy calls, and your operator time is worth $150/hour, that is another $5,400 a year the ROI math has to clear.

• Line item | Typical annual cost | Usually counted?
• Agency retainer at $2,500/mo | $30,000 | Yes
• Call tracking + analytics tools you pay for directly | $600–$2,400 | Sometimes
• Photos, video, review software | $0–$2,000 | Rarely
• Owner and staff hours (3 hrs/mo at $150/hr) | $5,400 | Almost never

Undercounting cost flatters ROI, which is exactly why agencies quote "return on retainer" instead of return on total cost. Run the math on the full number. And if a provider quoted you dramatically less than these figures, read what cheap SEO services actually deliver before celebrating the discount.

What numbers do you need before you can measure anything?

Three business numbers gate everything: lead-to-close rate, average job value, and lifetime value — and they come from your job records, not from any marketing tool. Most owners can pull them in an afternoon: count last quarter's inbound leads, count how many became paying jobs, and average the invoices.

Run an illustrative example — arithmetic, not a promised outcome. A plumbing company closes 40% of inbound calls at an average $650 per job. Fifteen organic calls a month is 6 jobs and $3,900 in monthly revenue against a $2,500 retainer: thin, on first-job math. But if the average new customer returns twice over five years and refers one neighbor, effective value per new customer is closer to $2,000 — and those same 6 jobs are worth $12,000. LTV is the difference between "barely break-even" and "best channel we have."

How do you attribute revenue to SEO when leads come by phone?

Install call tracking with dynamic number insertion (DNI): it shows a unique phone number to each website visitor and records which channel — organic, paid, maps, direct — produced every call. That is the only reliable way to attribute phone revenue to SEO. Asking "how did you hear about us?" at the front desk fails, because callers answer "Google" whether they clicked an ad, a map listing, or an organic result.

Phone attribution is not a detail; for service businesses it is most of the picture. BIA/Kelsey estimates phone calls influence $1 trillion in U.S. consumer spending, and Invoca's benchmark analysis of 60+ million calls found 37% of home-services calls driven by digital marketing are genuine leads — and 46% of those leads convert on the call itself.

The minimum stack is small: DNI call tracking, GA4 conversion events on every form and booking widget, and a lead-source field in whatever you use to track jobs — a spreadsheet works. Tag every lead once at intake, reconcile against closed invoices monthly, and the ROI formula finally has real inputs. A setup that only counts form fills is ignoring the channel where your best leads live.

What is a good ROI for SEO?

A defensible target for a service business: organic-attributed revenue of at least 3x total SEO cost by the end of year two, counting first-job revenue only, with LTV upside on top. Published benchmarks run higher — First Page Sage's 2026 SEO ROI report, covering campaigns from Q1 2021 to Q3 2025, shows industry-level ROI between 317% (ecommerce) and 1,389% (real estate), with break-even between 5 and 14 months.

Treat benchmark tables as context, not a contract: they average mature campaigns, carry survivor bias, and half the stats listicles ranking for this query recycle the same handful of vendor figures. Your honest benchmark is your own paid-channel math. If organic's cost per booked job undercuts your Google Ads cost per booked job by year two — and keeps falling while paid costs keep rising — SEO is doing its job. That comparison gets a full breakdown in SEO vs Google Ads: which comes first.

How long does it take to see ROI from SEO?

Expect break-even at months 6–12 and clearly positive ROI in year two; First Page Sage's industry data puts break-even between 5 months (construction) and 14 months (legal services). First organic leads typically show up at months 4–6, which means the first two quarters must be judged on leading indicators, not revenue.

Leading indicators predict revenue: indexation rate, ranked-keyword growth, impressions, first tracked calls. Lagging indicators are the revenue itself. Judging a 90-day-old program on lagging indicators guarantees a false negative; judging a 12-month-old program on leading indicators is how bad agencies survive. Month-by-month expectations are mapped in how long SEO takes to work.

When does the math say you should fire your agency?

Fire your agency when leading indicators are flat at month 6, or when cumulative organic-attributed revenue has not crossed cumulative cost by month 18 in a normal competitive market. Those are the two math triggers. Everything else — slow replies, ugly reports — is annoyance, not evidence. Here is the checkpoint schedule:

• Month 3 — process check: technical fixes shipped, pages indexed, content published on the promised cadence. No revenue expected yet. Fire only for inactivity.
• Month 6 — leading indicators: impressions and ranked keywords climbing in Search Console, first organic calls in the tracking dashboard. Flat lines here mean the program is not working, and waiting longer will not change that.
• Month 9–12 — lagging indicators arrive: organic leads costing less than paid leads, a visible trace from call to closed job.
• Month 18 — the ROI verdict: cumulative organic revenue should exceed cumulative total cost. If it has not, and the agency's only answer is "SEO takes time," the math says leave.

Some failures override the calendar. If your agency guarantees rankings, fire them now — nobody controls Google, and anyone guaranteeing rankings is lying about something. If they refuse to install call tracking, they are choosing not to be measured. If the monthly report leads with sessions and "keyword movement" but cannot name the calls and booked jobs organic produced, you are being managed, not served.

And notice what a 12-month lock-in contract actually does: it suspends every checkpoint above until the agency has been fully paid. Lock-ins protect the agency from this exact math — that is their function. It is why Foundgrove runs month-to-month with no minimum and no lock-in: if the checkpoints fail, the same spreadsheet that hired us can fire us.

How do you measure ROI on GEO and AI-search leads?

Same formula, one extra attribution layer. AI-referred visitors show up in GA4 as referral traffic from domains like chatgpt.com and perplexity.ai, and callers increasingly answer the lead-source question with "ChatGPT recommended you" — so add an AI-assistant option to your lead-source field and count those jobs like any other. The revenue math — leads × close rate × job value — does not change.

What changes is the budget line. Many agencies now sell GEO (generative engine optimization) as a separate retainer, which means a second cost your ROI math must clear. Foundgrove includes GEO/AEO in the base retainer from day one, so AI-search leads land on the same $2,500/month cost line as organic. For whether the channel deserves priority in your vertical at all, see when GEO is worth the investment.

How should you use this math on your next agency call?

Bring three demands to any provider, including us: call tracking installed in month one, a monthly report that leads with calls and booked jobs rather than rankings, and terms that let you act on the checkpoints — month-to-month, cancel anytime. An agency that resists any of the three is telling you, in advance, how the engagement will end.

This scorecard is the operating agreement behind our SEO services for service businesses: retainers from $2,500/month, month-to-month with no lock-in, GEO/AEO included from day one, no ranking guarantees ever, and you own the site, the content, and every scrap of tracking data — so the measurement stack survives us if the math ever says it should. Want the numbers before you spend anything? The free audit is a 10-minute personal video teardown of your site, delivered within 2 business days — no card, no pitch. Get my free audit