How Does an SEO Guarantee Work? Measuring Your Way to Page One

The short answer: a measured share of page one, not a ranking

Here is the whole promise, in the words we register and publish:

• The sentence. "We guarantee up to 50% of the key phrases you could rank for will rank on the first page of Google."
• The list. "We build a list of every variation your customers could type to find a business like yours, and we go after all of them." At the start we agree the list with you in writing.
• The three places. Your Google Business Profile in Google's local map results, page one of Google's organic results, and AI suggestions and answers.
• The schedule. At least 20% on the first page by day 90, 40% by day 180 and 50%+ by the end of the first year.
• Page one in each place. Each milestone share is measured separately in each of the three places. A phrase that shows only in the map results counts toward the map share, not the organic share. In the AI answer, the share is counted over the phrases for which Google shows one.
• Qualification. The first page guarantee is offered after an assessment, and only if you qualify.

"Up to 50%" and the year-one floor of 50%+ describe one finish line: the first page guarantee tops out at the year-one milestone, while our work keeps going toward every phrase on your list.

What Google says about SEO promises, and why this one is different

Here is Google's line, word for word, from its advice on hiring an SEO:

We agree with it, and a first page promise should never suggest otherwise. The same bullet tells owners to beware of SEOs that claim to guarantee rankings, claim a special tie to Google, or sell a paid fast lane into it [1].

Google's advice goes on: "An SEO audit should be about giving you realistic estimates of improvement, and an estimate of the work involved." [1] It tells owners to walk away from anyone who promises the top result, and to ask any SEO: "What kind of results do you expect to see, and in what timeframe? How do you measure your success?" [1] A measured-share promise answers each question: the result is a share of page one, the timeframe is 90 days, 180 days and one year, and success is a search of every phrase on your list from your market on each date.

In a Google podcast transcribed by Search Engine Journal, John Mueller called ranking or traffic promises "usually a red flag", because no one can reach into Google's systems and change the rankings [2].

The skeptics deserve a straight answer about first page promises too. Ahrefs sums up its view as "what cannot be controlled, cannot be guaranteed", and calls anyone who guarantees rankings over-confident at best [3]. About rankings we agree; a first page share is a different promise. The same article says a guarantee that isn't a scam "clearly and transparently communicates what is and what isn't covered" [3]. We cover the warning signs in our guide to SEO scams, and the wider question in can SEO be guaranteed.

The three places a phrase can count, one at a time

The map results

Google says local results are "mainly based on relevance, distance, and popularity", and explains the third factor under the heading "Prominence": how well known a business is, including how many websites link to it and how many reviews it has [4]. Distance depends on where the searcher stands, so it is measured, not worked; relevance and prominence are the levers. Our local SEO guide covers the profile in depth.

The organic results

A phrase counts here when your page is among the first ten organic results. Google says its ranking systems "are designed to work on the page level" [5], so the organic work is page by page: one page for each service, built to beat the pages already there.

Google's AI answer

The third place is, in the words of our written terms, "the answer written by AI that Google shows on the results page, when it names or links your business". Google says its generative AI features are "rooted in our core Search ranking and quality systems", and may run many related searches at once, which it calls "query fan-out", to build a response [6] [7].

Two limits matter. First, AI answers appear only on some searches: Google says AI Overviews show only when its systems judge them additive, "and as such, often don't trigger" [7]. In our own check of 12 commercial searches made from Las Vegas on September 27, 2026, Google showed map results for 11, organic results for all 12 and an AI Overview for only 2 [8]. So the written terms work out the AI share only over the phrases for which Google shows an AI answer when we check; if it shows none at a check, that milestone is judged in the map and organic results. We never promise an AI answer for every phrase.

Second, only Google's answer counts. ChatGPT and Perplexity are real channels, covered in our AI search optimization guide, but results on search engines other than Google are not part of the first page promise.

Page one in each place

This is the rule we hold ourselves to, and it is the harder one. At each milestone the share is counted three times, once per place, and every count has to reach the milestone. Take a list of 200 phrases at day 90. If 60 phrases show your business in the map results but only 20 show it in the organic results, the milestone is not met: the map share clears 20% and the organic share does not. The plan has to carry all three places at once.

Step 1: Build the phrase universe (service × modifier × place)

In practice it is a multiplication. This is the method, not a measured fact:

Ten services, six modifiers and five places already make 300 phrases, most of them long, specific searches.

• Autocomplete. Google says its predictions "reflect real searches that have been done on Google" and vary with the location a search comes from [10].
• Keyword Planner. Its volumes are 12-month averages by default, rounded and shared across close variants [11].
• Your own data. Your Business Profile's Performance report lists the search terms that found you, and Search Console lists the queries that already show your pages [12] [13].

Do not drop a phrase because a tool shows no volume. When we sent 606 phrases to Google Ads on September 27, 2026, 158 of them (26%) came back with no monthly figure, and six close variants of one phrase each returned the same 1,300, because Google reports variants as one bucket, not a sum [14].

Then the written terms' rules apply. Searches for your own business name do not count, because you should show up for those already. Each phrase counts once in each place, even if two of your pages show for it there. The list stays fixed unless both sides agree in writing to change it.

Each city page needs its own content. Google's spam policies say "Doorway abuse is when sites or pages are created to rank for specific, similar search queries", and give pages targeted at specific cities that funnel users to one page as an example [16]. Every variation is won with a useful page or a real profile. Our SEO competitor analysis guide walks through this step in detail, and our keyword difficulty guide explains how tools score each phrase.

Step 2: Capture every page-one competitor, per phrase and per place

Location changes the answer. Google says its systems weigh many signals, including "your location and settings", and use your location to show nearby results even when you don't type a place [17] [18].

A national check cannot stand in for a local one. When we searched "near me seo" US-wide on September 27, 2026, the map results included an agency in France [19]. So every capture is made from the agreed market, on the same settings, every time.

The places do not always share one competitor set. In our Las Vegas check, neither of the two AI Overviews named a business from the map results, and the "dental implants" answer linked dental practice websites outside Las Vegas [8]. So each place gets its own list of rivals, per phrase.

Step 3: Measure what the page-one businesses have

Google is plain that outside tools "don't have access to Google's internal ranking data" [1]. Everything below measures the competitors in order to estimate how Google treats them. None of it is Google's own data.

In the map results

For its 2026 Local Search Ranking Factors report, Whitespark asked 47 local search experts to score 187 factors. They put the primary Business Profile category first for the local pack, then the searcher's distance from the business and keywords in the business name, followed by an address in the city searched, being open at the time of the search and high Google ratings [20]. None of the experts has special access to Google's algorithm, so treat the list as informed opinion. The same experts weight Business Profile signals at 32% and reviews at 20% of what decides the local pack [20].

For reviews, we record the count, the rating and the recent pace, not just the total. In Sterling Sky's correlational study of 8,186 businesses in 200 US cities, the reviews a business gained in the current month mattered more to "near me" rankings than its total count [22].

In the organic results

Google says the most basic relevance signal is a page containing the words of the search in its headings or body, and that its systems then look for "other relevant content beyond the keyword" [17]. Its Search Essentials say to put those words in prominent places, "such as the title and main heading of a page" [23].

We count referring domains to each ranking URL, page by page, because averages hide weak spots. Ahrefs builds its Keyword Difficulty score from the links pointing at the current top ten, and says the score "can sometimes hide specific weaknesses in the search results that you can exploit" [24].

In the AI answer

We record what the AI answer cites and names, per phrase, from the agreed market. Its overlap with the organic top ten is real but uneven: studies put the share of AI Overview citations that also rank in the top ten anywhere from 16.7% (BrightEdge, September 2025) to 76% (Ahrefs, July 2025), and Ahrefs' own re-run found about 38% (March 2026) [25] [26] [27].

Mentions appear to matter too. In Ahrefs' study of 75K brands (May 2025), brand web mentions correlated with AI Overview visibility at 0.664, against 0.218 for backlinks; that is correlation, not proof of cause [28]. Whitespark's report adds that "3 of the top 5 AI Search visibility factors are citation factors" [20].

The bar: the weakest business still on page one

The number that matters for each phrase and place is not the leader's. It is the bar: the weakest business or page still holding a page-one slot. To appear in the map results you have to pass the last business shown, not the first; to reach the first page of organic results you have to pass the tenth result. This is the center of the method, and it is why a share of page one can be planned when a position cannot.

Step 4: Match every signal, then exceed it

Ahrefs states the link side of this rule almost word for word: to reach the top ten, acquire "as many backlinks as the current top-ranking pages have (if not more)" [24].

On content, "more" means more value, not a longer copy. Google asks: "If the content draws on other sources, does it avoid simply copying or rewriting those sources, and instead provide substantial additional value and originality?" [29]

• Map results. The right primary category, chosen with Google's test that a category completes the statement "This business IS a" [21]. Complete, accurate hours and services, photos that "represent reality" [30], and a steady flow of genuine reviews, requested with a link or QR code [31]. Google's content policy bans writing, buying or rewarding reviews [32].
• Organic results. A dedicated page for each service, written for the intent page one shows, covering what those pages cover and what they miss, with links from relevant sites. Whitespark's experts rank a dedicated page for each service as the top local organic factor [20].
• AI answer. Pages that are indexed and able to show a snippet, which Google says is the eligibility test, plus an up-to-date Business Profile and genuine mentions [7] [6]. Google warns that seeking inauthentic "mentions" across the web "isn't as helpful as it might seem" [6].

The method keeps moving. We keep up with every change to Google's rules and update how we work as they change.

The page-one math: place by place, smallest gaps first

A simple way to think about it (this is the method, not a measured result):

1. N is the number of phrases on your agreed list, and the bar for a phrase in a place is the weakest competitor holding page one there today.
2. The gap is what you lack against the bar on each measured signal: reviews, a category, referring domains, a page, a subtopic.
3. The planned share in a place is the number of phrases whose gap in that place we expect the work to close by a milestone, divided by N. That expectation is a judgment, and we treat it as one.
4. The promise is made only where the planned share clears every milestone with room to spare in every place.

Here is the worked arithmetic, using only the schedule's percentages and a list of 200 phrases:

The share is a plain count: phrases that count in a place, divided by phrases on the list (in the AI answer, by the phrases that show one). It is not "share of voice", which tools weight by clicks or volume: Ahrefs calculates share of voice from the clicks your site receives against all clicks on the tracked keywords, and Semrush's version also weighs each keyword's search volume and estimated traffic [33] [34].

Because the count is unweighted, a small long-tail phrase counts the same as a busy one. That is why the list has to come from what your customers really type, agreed in the written terms before the first check. Ahrefs warns that weak guarantees lean on irrelevant keywords or "keywords with very little to no search volume" [3].

Sorting by gap turns the list into waves: the smallest gaps in each place first, bigger ones as pages, links and reviews build up. One change often moves many phrases, because a single service page serves every modifier and place built on that service.

The schedule: 20% by day 90, 40% by day 180, 50%+ by year one

Google's current guidance is that "Some changes might take effect in a few hours, others could take several months", and that you "likely want to wait a few weeks" before judging a change [35]. Until June 2026, Google's "Do you need an SEO?" page said results typically take four months to a year; the June 2026 revision removed that sentence [36] [1].

In Ahrefs' 2025 study of new pages, 40.82% of the pages that reached the top ten did so within a month [37]. Ahrefs also suggests updating a page still outside the top ten after about six months [37].

Read that number carefully: it describes random pages with no targeted work. It shows page one is not automatic, which is why measurement comes first, and says nothing about a planned campaign either way.

At each date the milestone has to be met three times: in the map results, in the organic results and in Google's AI answer, as the terms measure it. The written guarantee terms give the exact counting rules, and our guide to how long SEO takes to work covers timing in more depth.

Why the schedule leaves margin

1. Only a share is promised. At least 20%, then 40%, then 50%+ in each place, never every phrase, so in each place the plan starts with the phrases whose gaps are smallest.
2. Most of the list is long tail. Most keywords get very few searches, and Ahrefs found that high-volume keywords take longer to rank for [15] [37]. Our reading, not a measured result for any client, is that the page-one incumbents on specific phrases are often weaker than those on the busiest terms.
3. Qualification comes first. The first page guarantee is offered after an assessment, and only if you qualify. The assessment is the measurement in this guide, done before any promise is made.
4. Much of the work counts twice. Google says links to your business and your reviews feed local prominence, and that its AI features rest on its core ranking systems [4] [6]. Whitespark's experts rank a dedicated page for each service first for local organic results and second for AI search visibility [20]. Strong pages and profiles tend to lift more than one place, though never one-for-one.
5. Checks are pinned to your market. Every milestone check is made from the market we agreed, which removes most of the location noise Google describes [17].

What the margin is not: a statistic. No public study measures how often a schedule like this is met, and we will not quote odds.

There is one honest counterweight. The AI place is the least predictable of the three: AI answers show on only some searches, and Google's two AI surfaces often cite different pages, with Ahrefs finding only 13.7% of citations shared between AI Overviews and AI Mode (December 2025) [38]. That is why the written terms define how the AI share is measured, and why we measure it on the day, from your market.

How the share is checked

If your own search shows something different, we check it again with you, because Google's results can differ a little from one searcher to the next. Between checks, Google's own tools let you watch progress:

• Your Business Profile's Performance report lists the search terms people used to find you, and its searches figure updates at the start of each month [12].
• Search Console's Generative AI performance report, launched on June 3, 2026 and on every site by August 31, 2026, shows how often your pages appeared in Google's AI features [39].

What happens after a missed milestone is set out in the first page guarantee terms, along with the exclusions and how either side can end the agreement. Read them before you sign.

What the first page guarantee does not promise

Visibility is the part SEO controls best; clicks depend on everything else on the results page. Advanced Web Ranking's July 2026 data put the average click-through rate on US desktop at 20.02% for the first organic result, 10.36% for the second and 3.89% for the third, with the first three collecting roughly 92% of the clicks that reach the top ten [40].

AI answers change that math again. Pew found users clicked a traditional result on 8% of visits when an AI summary appeared, against 15% without one [9]. In Ahrefs' December 2025 data, an AI Overview correlated with a 58% lower click-through rate for the top-ranking page [41]. Google disputes the size of such declines, saying total organic clicks have been "relatively stable" year over year [42].

Backlinko goes further and questions whether "ranking on the first page" is a worthy goal on its own [43]. We agree that page one is a floor, not a finish line. It is still the floor that matters.

See the math for your own market

To see where your own business stands, request a free SEO audit. Our page on how the SEO guarantee works sets out the first page guarantee in full, with its written terms. You can see the businesses we have worked for on our client results page, and read about James Compton, who wrote this guide.