Tool card · one word, two meanings
A search engine optimizer is either a piece of software that crawls, audits and reports on a website, or the human specialist who reads those reports and decides what to change first. English uses the same phrase for both, and mixing them up is expensive: a tool can surface four hundred issues in an hour, but it cannot tell you which five of them are costing you revenue this quarter.
A search engine optimizer is a person who improves a site's organic search performance, or the software that scans a site and reports what needs fixing. The person sets priorities, writes and ships changes, and owns the outcome. The tool crawls, audits and measures. One decides, one reports, and only the person is accountable for results.
Replies usually within a few working hours. No obligation to buy.
Where the confusion started
For most of the 2000s the phrase was a job title. Agencies sold the work of a person, and that person read crawl data, rewrote title tags, debugged redirect chains and pitched link partnerships by hand. Then the software industry borrowed the word. Tools started shipping products called optimizers, page optimizers, image optimizers, and the term stopped pointing at one thing.
Both meanings survived because both are useful. A buyer searching for the phrase may want a subscription that audits a fifty page site every Monday, or may want a specialist who can sit in a meeting and explain why organic traffic fell after a replatform. If you answer the wrong question, you either sign up for a dashboard you will never open, or you hire a full retainer when all you needed was a better sitemap.
If the word itself is still fuzzy, start one step back with Whats Is SEO? and return here once the practice makes sense. This page is about the noun, not the discipline.
The two definitions
A search optimizer in this sense is a SaaS product or a desktop app. You point it at a domain, it pulls a crawl, and it produces a list: broken links, thin pages, missing canonicals, slow templates, duplicate titles.
An optimizer in this sense is the specialist on the other side of the report. They read the list, throw most of it away, and sequence the rest against what the business is trying to sell this year.
Capability check
Most disappointment comes from expecting one column to do the other column's job. The table below is the one we use in first calls, because it makes the boundary obvious without arguing about budget.
| Job to be done | Software | Specialist |
|---|---|---|
| Crawl 10,000 URLs overnight | Yes, easily | Too slow by hand |
| Spot a redirect chain 6 hops deep | Yes | Yes, if they look |
| Decide which 5 issues matter this quarter | No | Core skill |
| Read buyer intent behind a query | No | Core skill |
| Write copy a procurement manager trusts | No | Core skill |
| Ship the change in code and verify it | No | Yes, with your dev |
| Explain a traffic drop to your board | No | Yes |
| Monitor 50 sites for regressions weekly | Yes, cheaply | Too expensive |
| Say no to a tactic that risks the domain | No | Judgement call |
The short test
You do not need to be technical to run this test. Five questions, asked in order, separate a subscription from a practitioner in about ten minutes.
A tool answers with a count. A person answers with one page, one reason and one number it should move. If the reason is "it is flagged as an error", you are still talking to the tool.
This is the fastest tell. Anyone can list problems. Deleting two thirds of a crawl report takes judgement and a working theory of how your buyers search.
Software never implements. If nobody on the call can describe how a change reaches production, you are buying a report, not an outcome.
Practitioners revise. They test a title rewrite, watch it for six weeks, and put it back. A product roadmap cannot tell you about a reversal, because it never made a judgement to begin with.
Real specialists name a stopping condition: a query set that turned out to have no commercial value, a category that belongs on paid search, a client whose margins cannot absorb the work.
If four of five answers come back as "the dashboard shows", you are looking at tooling and should buy tooling. If four come back as "because that page earns", you are looking at a practitioner and should scope a retainer. Most real engagements are a mix, and that is fine as long as you know which half you are paying for.
Send us the domain. We will tell you which three findings matter and which forty you can safely ignore.
Costly assumptions
It is the same as buying a stethoscope. The subscription gives you findings; nothing in it prioritises, drafts, or ships. Teams that buy software and no operator usually end up with a growing issue count and flat traffic.
A crawl that returns 40,000 warnings has not done 40,000 units of work. It has run a rule set. Value starts when someone converts that list into a sequence with owners and dates.
Crawlers cannot see demand. A site can score 98 out of 100 on every technical metric and still rank for nothing, because the pages answer questions nobody in the buying market asks.
This is the most expensive version of the mistake. Every finding still needs a decision about scope, sequence and trade-off against engineering time, and that decision is the job.
Neither definition is a single channel. The work usually touches templates, copy, internal linking, page speed and digital PR, and the sequence depends on what is already working.
A retainer that starts before anyone agrees on the metric becomes a monthly report nobody reads. Agree on the number first, then decide whether you need software, a person, or both.
Our side of the table
We work as the person in the second definition, and we bring our own tooling to it. In practice that means we take responsibility for the sequence, not just the scan. A typical month starts with the pages that already earn, because improvements there compound while new pages are still being indexed.
Technical findings get translated into tickets a developer can pick up without a meeting. That often means Technical SEO Services on templates, canonicals, crawl budget and Core Web Vitals, plus a written spec for anything that needs code. On the content side we brief writers with the actual queries, the objections buyers raise, and the proof a procurement team will ask for.
We do not promise positions. We do commit to a documented sequence, measurable movement on agreed queries, and a straight answer when a tactic is not worth your engineering hours.
First month
Crawl, index check, query export, conversion data. We record the starting numbers before touching anything.
Two thirds of the findings are dropped. The rest are scored by pages that already convert and queries that already appear.
The first batch goes live: titles, internal links, one template fix. Small enough to attribute, big enough to matter.
Four to six weeks later we compare against the baseline, keep what worked and revert what did not.
A small example
Take an industrial supplier with 900 product pages. A crawl returns 1,100 warnings: 600 missing meta descriptions, 300 images without alt text, 200 internal links pointing to a deprecated category. A tool ranks these by count and shows a red score.
A practitioner reads the same export and notices something else. Four pages in that crawl bring 70 percent of the organic enquiries, and all four sit three clicks from the homepage while the deprecated category sits in the main navigation, leaking authority to a dead end. The missing descriptions are real but worth almost nothing: those pages get no impressions.
So the first month is one navigation change and a short internal linking pass, not 600 rewritten descriptions. That is the whole difference between the two meanings of the word, and it is why the person is still on the payroll after the software is installed.
If you want the wider picture of how this fits with content, links and local pages, the footer links to our full service menu. The point here is narrower: the noun has two meanings, and only one of them can be held responsible for a result.
Questions we get
Both, in current usage. Vendors use the phrase for auditing software, and agencies use it for the specialist who runs the programme. Only context separates them, so ask any supplier which one they are selling.
For monitoring, largely yes. For prioritisation, no. A crawler can tell you that a page is slow; it cannot tell you that the page is slow and also the only page ranking for a phrase your buyers actually use.
If you need weekly regression checks across many templates, buy tooling. If traffic is flat while the crawl is clean, or the report is long and nothing ships, you need a person. Most teams above a few hundred pages end up with both.
Ask for the first three changes, the metric they expect to move, and the date they expect to see it. A credible answer names pages and queries. A vague answer talks about best practice and long-term strategy.
Largely overlapping. Consultant usually signals advice and strategy; optimizer signals hands-on execution. The label matters far less than whether the person can ship changes and measure them.
After a short call and read-only access, the baseline usually takes a week and the first batch ships in the second or third. Message us and we will tell you what we would look at first.
