Check 01
Do the artefacts exist?
Ask for the issue log, the spec sheets, the briefs, the outreach log. If the month's work produced nothing you can open, the invoice is paying for meetings.
Service menu
SEO services are the outsourced work of improving how a site appears in search: a technical audit, page-level optimisation, content production, link acquisition, local listings and reporting. You are not buying a position in the results. You are buying a defined set of tasks, artefacts and decisions, delivered on a schedule you can inspect. That distinction is the whole game, because it turns a vague retainer into a checklist. Below are the six lines of work, what each one hands over, how long it takes, and how to judge the value before you sign.
In practice it means a monthly package of work across four layers of a site: technical health, page-level targeting, content production and off-site authority. A provider researches, fixes, writes, links and reports. The invoice buys time, judgement and artefacts, never positions. That is why every line of work should name what you receive and when.
The six lines
A credible package covers all six lines, even if the depth varies by month. A proposal that lists only two is usually a content retainer wearing a broader name.
Crawling, indexing, canonicals, redirect chains, page speed, mobile layout, structured data, and the log of pages that should never be indexed. This is the plumbing. Nothing else works if a crawler cannot reach the page or the page competes with a duplicate of itself.
One keyword to one page, then everything that follows from it: title, meta, H1, subheadings, internal links, answer block, image labels. This is where a page stops describing everything and starts answering one question completely.
Briefs built from real questions, drafts written with the people who know the product, and pages published with a clear next step. Depth beats length here, and the brief does more for quality than any word-count target.
Earning mentions from sites that a buyer would recognise: trade associations, supplier directories, industry publications, resource lists. Volume without relevance is the most common way to waste a link budget.
For businesses with a physical address or a service area: business profile, categories, hours, photos, name-address-phone consistency across directories, and a simple review request flow.
The read-out matters more than the dashboard. What moved, what did not, what was shipped, what is next. Reporting that only lists positions tells you nothing about whether the work is reaching buyers.
At a glance
Use this table as a contract checklist. If a proposal cannot fill a row, ask why before you sign rather than after month three.
| Line of work | Deliverable you receive | Cycle | How to verify it |
|---|---|---|---|
| Technical audit | Prioritised issue log with severity and fix notes | 2–3 weeks, then quarterly | Re-crawl shows the issue closed |
| On-page | Page spec sheet plus change log | 2–4 days per page | View source on the live URL |
| Content | Brief, draft, revision, published page | 4–8 pages monthly | Published URL, not a Google Doc |
| Link building | Outreach log with live placements | 3–6 weeks per batch | Open each link, check it is indexed |
| Local | Profile refresh, citation fixes, review flow | Setup 3 weeks, then monthly | Search your brand plus city |
| Reporting | Dashboard and written read-out | Monthly | Numbers reconcile with your analytics |
Acceptance test
Value in search work is not a feeling. It is three checks you can run on any provider, at the end of every month, without asking for a special report.
Check 01
Ask for the issue log, the spec sheets, the briefs, the outreach log. If the month's work produced nothing you can open, the invoice is paying for meetings.
Check 02
Open three URLs from the change log and inspect the page yourself. Published work is verifiable work. Screenshots of a draft are not.
Check 03
Look at impressions on target queries over ninety days, and at enquiries from organic sessions. One bad week means nothing. A flat quarter means the pages are not matching demand.
If a provider answers those three clearly, the rest is usually process detail. If they dodge any of them, the retainer will be hard to audit and harder to end.
Before you sign
One more, less obvious: a proposal with no stated cadence for technical work. If the audit is not in the scope, the content will fight a site that cannot be crawled efficiently.
First ninety days
Months one to three are the least glamorous and the most decisive. This is the shape we work to, and the shape you should expect from anyone credible.
Crawl, index review, keyword map, competitor gap notes, analytics check. Deliverable: an issue log and a build order you agree with.
Template fixes, four to six pages rewritten or built, internal links repointed, schema added where it earns rich results.
Re-crawl, impression movement on target queries, first outreach placements live, and a written decision on what to repeat.
If you want the plain-language foundation first, read Whats Is SEO?. For the page-level line of work described above in detail, see On-Page SEO Services.
FAQ
It is a set. Technical, on-page, content, links, local and reporting are separate lines of work with separate artefacts. Good providers sell them as a package because the lines depend on each other.
Yes, and sometimes that is the right call. A site with strong content but broken indexing should buy the audit first. A site with clean plumbing should buy pages.
You do. The pages sit on your domain, the analysis files should be handed over, and the accounts should be yours from day one. Make that explicit in the contract.
A project delivers a fixed artefact, such as an audit or a cluster of pages. A retainer delivers a cadence of work across the lines. Most sites need a project first, then a lighter retainer.
Not necessarily. Content usually moves first because it is fully in your control. Links take longer to land and can start in month two without delaying anything.
Start with technical setup and four to six pages aimed at specific questions. Buying a large link package before the pages exist is the most common early mistake.