An SEO campaign is a time-bound project: a stated goal, a set of pages and queries to work on, a start date and a date on which you add up the results. It is a different thing from ongoing SEO maintenance, which has no end.
The distinction looks like a formality and it isn't: it decides what you can ask of the people working on the project, and when you can ask it.
One number to set the scene. Analysing around 14 billion pages in its index, Ahrefs found that 96.55% get no clicks at all from Google (study published in December 2023, global index). An SEO campaign exists to move a few dozen pages to the right side of that percentage. It is not work on the website: it is work on a list of specific pages.
A campaign and ongoing SEO are not the same thing
The most useful definition contains one word that most guides leave out: time-bound. SEOmonitor spells it out: unlike continuous SEO work, campaigns are narrower in scope and have a deadline, with stated goals and metrics.
On Google UK the AI Overview for "seo campaign" picks up that word from SEOmonitor, but the pages ranking below it (Digitas, Loop Digital, Semrush) define a campaign as "a strategic plan", "a strategic effort" or "a carefully organized plan" to improve a website's visibility (DataForSEO, Google UK, September 2026). Each of those descriptions fits a monthly retainer with no end date just as well. The reader comes away without the tool they needed.
In practice the difference shows in four points. A campaign has a scope (these pages, these queries), a numerical goal, a deadline and a failure criterion stated before it starts. Ongoing SEO has a monthly budget and a list of recurring tasks. Both are legitimate. What doesn't work is selling the second under the name of the first, because then there is never a moment when someone says whether it went well.
If you are still deciding which of the two you need, the difference between their goals is explained in more detail in the guide on how to set measurable SEO goals, and the relationship with the long-term plan in the one on SEO strategy.
How much people search for it in the UK, and what the price of each query says
Before explaining how a campaign is built, it is worth looking at how many people are searching for one, and how much those people are worth to advertisers. The data comes from Google Ads Keyword Planner for the United Kingdom, average of the twelve months to August 2026.
| Query | Monthly searches in the UK | Average CPC |
|---|---|---|
| seo | 18,100 | £9.63 |
| seo agency | 6,600 | £21.38 |
| seo consultant | 2,900 | £14.36 |
| website ranking | 1,900 | £3.73 |
| seo audit | 1,300 | £8.80 |
| seo strategy | 720 | £4.55 |
| how much does seo cost | 140 | £11.36 |
| seo campaign | 90 | £14.17 |
Two things stand out. "seo campaign" gets 90 searches a month with a cost per click of £14.17, the third highest in the table and almost level with "seo consultant": whoever types those words into Google is weighing up a spend, not revising for an exam. And in the UK "seo agency" is searched more than twice as often as "seo consultant", 6,600 against 2,900, with the highest CPC in the table at £21.38. In Italy the ratio is the other way round, with the consultant query ahead.
The low volume of the main query says something else too: people with a budget don't search for "seo campaign", they search for the result. Capturing those searches is the job of keyword research, and it starts from recognising search intent before looking at volumes.
The phases, and the output that must exist at the end of each one
Lists of phases all look alike: analysis, keywords, on-site, off-site, monitoring. The problem is not the list, it is that nobody writes down which document or file must exist at the end of each phase. Without that you have no way of knowing whether the phase was done.
| Phase | What must exist at the end | Typical duration |
|---|---|---|
| Initial analysis | A list of errors ranked by impact, plus the day-zero snapshot: rankings, clicks and impressions on the start date | 2-4 weeks |
| Keywords and intent | A query-to-landing-page table, with the intent stated for each row | 1-2 weeks |
| Technical fixes | A dated change log, and a before/after comparison on indexing and speed | 1-3 months, depending on who touches the code |
| Pages | URLs published and live, not "content delivered" | ongoing |
| External authority | A list of linking domains, with the date each was acquired | ongoing |
| Measurement | A report comparing every metric with day zero, not with the previous month | monthly |
The most neglected row is the first. If the day-zero snapshot isn't taken before anyone touches the site, at month six nobody can prove anything, and the discussion becomes one opinion against another. How to build that snapshot is covered in the SEO audit guide; which numbers to keep in it is in the guide to SEO KPIs.
What you can ask for at month 3, month 6 and month 12
This is the table missing from the guides that rank for this query, and in practice it prevents half the arguments. It tells you in advance which question is reasonable at each checkpoint, and which isn't yet.
| Checkpoint | Must already exist | What you look at | Too early to ask for |
|---|---|---|---|
| Month 1 | Audit delivered, tracking verified, day zero recorded | No results yet. Only that the tools actually measure | Rankings |
| Month 3 | Blocking errors fixed, first key pages live | Impressions in Search Console and number of indexed pages | Enquiries, revenue |
| Month 6 | The core pages published, first external links acquired | Rankings on secondary queries, first clicks, first enquiries | Position one on the head query |
| Month 12 | The whole scope covered, updates started on older pages | Enquiries and sales attributed to organic, share of voice on the cluster | That results hold with no further work |
The right-hand column is the one that matters. Asking for organic revenue at month three is like weighing a tree planted in March: there is an answer, but it tells you nothing useful. The full reasoning on timelines, with the variables that stretch them, is in how long SEO takes to deliver results.
Seven years on the same project: the Macropix numbers
Macropix makes custom LED screens in Milan and has passed five thousand installations worldwide. When the project started, in 2020, the site did not appear for any of the category keywords in its sector. It is the case we use to explain what happens when a campaign doesn't stop at month twelve.
| Indicator | Value |
|---|---|
| Years of continuous work | 7, since 2020 |
| Total budget across SEO and advertising | €420,000 |
| Enquiries generated by the project | around 7,000 |
| Keywords in the top three on Google Italy | over 100 |
| Position for "monitor pubblicitario" (advertising monitor, 390 searches/month) | from 88 to 2 |
| Position for "ledwall" (4,400 searches/month) | from absent to 1 |
| Share of voice on the ledwall cluster, July 2026 | 25.55%, ahead of amazon.it at 18.08% |
The row we talk about most is not the top spot for "ledwall". It is "monitor pubblicitario": position 88 means page nine, in other words zero traffic, and you don't get from there to second place with a technical fix. You get there with a page written by the people who build that product. The full method, with the table for all nine keywords, is in the piece on SEO for manufacturing, and the work done year by year is on the Macropix project page.
On the same account, for comparison, Google Ads between July 2025 and June 2026 produced 374 conversions from €16,805 spent: €44.93 per conversion, average CTR 3.59%, average CPC €0.81. The organic rankings in the table above have no cost per click, but they took years to form. Ads brings enquiries from day one and stops the day you switch off the budget. How the two channels support each other is covered in the piece on SEO and Google Ads working together.
What an SEO campaign cannot promise
In its official documentation, on the page about whether you need an SEO, Google writes one short, blunt sentence: "No one can guarantee a #1 ranking on Google." On the same page it warns against SEOs who guarantee rankings or claim a special relationship with Google.
There is a second limit, less often quoted and harder. Across 1.3 million US keywords, Ahrefs measured that 72.9% of pages in the top 10 are more than three years old, that the page in position one is five years old on average and that only 13.7% of first-page results are less than twelve months old. The data was collected in September 2023 and published in May 2025, on a US sample: the order of magnitude matches what we see on industrial projects, where the top positions formed between the third and fifth year.
In contractual terms: a six-month campaign can deliver indexed pages, fixed errors, secondary queries won and first enquiries. The head query of the sector, almost never.
When an SEO campaign should not be started
This is the section most guides on the topic skip, and it shouldn't be. There are five situations in which it makes sense to move the budget elsewhere and wait.
- The site will be rebuilt within six months. Optimising pages that will change URL and structure is work thrown away twice. Rebuild the site first, with SEO inside the rebuild project.
- The demand doesn't exist. If the product is new and nobody searches for the category, there is nothing to capture. What you need there is to create demand, and that is done with advertising, not SEO.
- It can't be tracked. Forms without tracking, phone calls not attributed, quotes handled over WhatsApp by the sales rep. Without attribution, at month twelve you'll be arguing about gut feelings.
- The budget covers less than six months. Three months of SEO produce work done and no readable result: it is the most expensive way of concluding that SEO doesn't work.
- The decision-maker changes their mind every month. The enemy of a campaign isn't Google, it is priorities reshuffled at every meeting. One week blog posts, the next links, the week after the technical side.
The reverse also applies: when a campaign under way should be closed. The reliable signal is not falling rankings, which always fluctuate. It is this: impressions stable or growing and clicks falling for three months in a row, with the same scope. It means you appear but people don't choose you, and the problem is in the page or the SERP, not the budget. The other possible diagnoses are collected in why your website traffic is dropping, and the criteria for judging whoever is working on it in how to tell whether an SEO agency is working.
What changes now that some answers are written by AI
The classic criterion for closing a campaign was ranking position. Today it holds up less well, and there is a measurement that shows it. The Pew Research Center tracked the real browsing of 900 US adults across 68,879 searches made in March 2025: when an AI-generated summary appears on the results page, clicks on a traditional result stop at 8% of visits, against 15% when there is no summary. US sample and stated period: the figure should not be read as if it applied identically to the UK.
The practical consequence is concrete. A campaign that closes by declaring "we're number one" on queries that show an AI Overview is declaring a result worth less than it was in 2022. A second item has to be added to the scope, citation in generated answers, which is measured and earned differently: it is the work we have put into our answer engine optimisation service. On Google UK, the query "seo campaign" itself opens with an AI Overview.
How to measure it without fooling yourself
Four rules we apply on every project, in order of importance.
- The comparison is always with day zero, never with the previous month. The previous month contains seasonality and makes you celebrate or despair over nothing.
- Impressions move before clicks, and clicks before enquiries. Looking at the third indicator when the first hasn't moved yet tells you nothing.
- Share of voice across a group of queries beats a single ranking. Macropix on the ledwall cluster is at 25.55% against amazon.it's 18.08%: information no single ranking line would have given.
- Enquiries must be attributed, not estimated. "We got more calls" is not data.
There is a context problem here worth naming. According to the ONS article on artificial intelligence in UK businesses published on 20 July 2026, the share of businesses with 10 or more employees using at least one AI technology rose from around 12% in September 2023 to around 35%. The gap by size is wide: 28% of businesses with 0 to 9 employees against 49% of those with 250 or more, and lack of expertise is one of the most cited barriers. Put the other way round: most small firms still don't have in-house the skills needed to judge a campaign on data.
If the financial calculation is the part you care about, the method is in the piece on SEO ROI, and the orders of magnitude of spend in how much SEO costs. If instead you are choosing who to hand the project to, the questions to ask before signing are listed in questions to ask an SEO agency. How we work on organic projects is described on the SEO service page.
Frequently asked questions
In the contracts we sign, the term is six or twelve months. Six months is enough to deliver indexed pages, fixed technical errors, rankings on secondary queries and the first enquiries. The head queries of a competitive sector take years: Ahrefs measured that 72.9% of pages in the top 10 are more than three years old, on a sample of 1.3 million US keywords.
The strategy decides which markets, queries and pages are worth being present on, and has no deadline. The campaign is the time-bound project that carries out part of that strategy: defined scope, numerical goal, closing date. A strategy without campaigns stays a document; a campaign without a strategy wins queries that don't bring customers.
The analysis and writing, yes, if someone in the company knows the product and has time for it every week. The two parts that almost always get dropped when you go it alone are the technical fixes, which depend on whoever touches the code, and acquiring external links. The real constraint is not skill, it is continuity: a campaign stopped at month three produces nothing readable.
Look at impressions in Search Console and the number of indexed pages, compared with day zero. If at month three impressions are flat and the new pages aren't indexed, something is wrong. If impressions are rising and clicks aren't, it is early: that is the normal stage before rankings become useful.
No, and Google says so in its own documentation: no one can guarantee a number one ranking. Anyone who promises it in a contract is selling something they don't control. A sensible commitment covers the work delivered and the intermediate indicators, not the final position.
In the first months, yes, for a practical reason: Ads brings enquiries while SEO doesn't yet, and paid campaign data shows which queries really convert. On an industrial project we manage, Ads produced 374 conversions from 16,805 euros over twelve months, 44.93 euros per conversion, while the top organic rankings took years to form.
One thing the industry rarely says. The biggest cost item in an SEO campaign is not the monthly fee, it is the restart. Every time a project stops at month nine and starts again two years later with another supplier, the company pays twice for the same climb, and in the meantime the pages that took its place have built up the age it lacks. The Macropix rankings formed between the third and fifth year: a project stopped at month nine and restarted from scratch pays for that time twice, and the second time costs more because someone ahead of you has kept going.
