Keyword Researcher: The Direct Answer and What to Do Next (2026)

The short version: a keyword researcher is either a person you pay to find and organise the search terms your customers use, or the job you do yourself with a free tool and an hour on a Sunday. Most small businesses don't need to hire one. They need to stop treating keyword research as a document and start treating it as a decision tool that leads straight into content that gets written and published that week.

What "keyword researcher" means right now

The term covers three different things and people mix them up constantly.

• A freelancer or agency employee whose job title is literally Keyword Researcher, usually working under an SEO manager, paid to build spreadsheets of search terms with volume, difficulty and intent tags.
• A tool or software function, like the "keyword research" tab inside Ahrefs, SEMrush, Ubersuggest, or Google's own Keyword Planner, that spits out a list when you type in a seed word.
• A task, not a role, that any marketer, founder, or content writer does as one step in a bigger content plan.

Nine times out of ten, when someone types "keyword researcher" into Google in 2026, they want one of two things: they want to hire a human to do this for them, or they want to understand what the job involves so they can do it themselves without wasting a weekend on a tool they don't know how to read. I'm going to answer both.

Do you need to hire a keyword researcher, or do it yourself

Here's the honest split, based on what I've watched play out with clients over about fifteen years of doing this for a living.

Hire someone if you're publishing more than four pieces of content a week across multiple sites or clients, if you're running an agency and need a repeatable process someone junior can execute, or if you don't have an hour to spare and the cost of a freelancer is small next to the cost of your own time. A decent freelance keyword researcher on Upwork or through a specialist SEO agency will charge somewhere between £15 and £45 an hour, or a flat £150 to £400 for a full research pass on a small site with 20 to 40 target pages.

Do it yourself if you're a solo business owner, a small team, or you're writing fewer than ten pieces a month. The tools have got good enough, and the process has got simple enough, that paying someone else to do this specific task rarely earns its cost back unless you're also paying them to write the content and build the strategy around it.

If you're at the stage where you're weighing up hiring outside help for the wider marketing picture, not just this one task, it's worth reading about what an AI consultant costs before you commit, because keyword research on its own is rarely the bottleneck. The bottleneck is almost always what happens after the list is built.

The actual process, step by step

This is the process I use with clients, stripped down to what matters.

• Step 1: Start with questions, not words. Go to your own inbox, your sales calls, your customer service tickets, and write down the exact phrases people used when they asked about your product or service. This beats any tool because it's real language, not guessed language.
• Step 2: Check what people are asking in public. Search groups and forums where your customers already hang out. I use Facebook groups for this constantly because the questions there are unfiltered and current, and I also run searches through Twitter's advanced search to catch complaints and comparisons people make in real time, which almost never show up in a keyword tool.
• Step 3: Put your seed terms into a proper tool. Ahrefs, SEMrush, or even the free Google Keyword Planner will give you volume and a difficulty score. Don't chase volume alone. A term with 40 searches a month that matches exactly what a buyer says two weeks before purchase beats a term with 4,000 searches a month that's mostly people doing research for a school project.
• Step 4: Group by intent, not by topic. Split your list into "just looking," "comparing options," and "ready to buy." Write for the ready-to-buy group first. This is the single biggest thing people get backwards.
• Step 5: Pick fewer keywords than feels comfortable. I tell clients to commit to 15 to 20 core terms for the next quarter, not 200. You cannot write well for 200 things at once, and nobody ever has.

The tools worth paying for, and where each one falls short

Ahrefs gives you the most complete backlink and ranking data, but the keyword volume numbers are estimates, not gospel, and they can be wildly off for smaller or newer search terms. SEMrush is strong for competitor gap analysis, showing you what your competitor ranks for that you don't, but its interface takes a proper afternoon to learn. Google's own Keyword Planner is free and its volume data comes straight from the source, but it was built for ad buyers, not content writers, so it lumps related terms together in unhelpful ways. AnswerThePublic and Also Asked are good for finding question-shaped keywords fast, but they thin out quickly once you've mined the obvious ones.

None of them tell you which keyword will make you money. That's a human judgement call, and it's the part people skip. I've written a longer breakdown of where each tool works and where it doesn't in this comparison of AI tools for marketing research, worth reading before you subscribe to anything.

A story that changed how I do this

In 2019 I had a client, a small B2B software company, who paid an agency £3,200 for a keyword research report before I came on board. It was 380 keywords, colour coded, tagged by difficulty, beautifully laid out in a spreadsheet nobody ever opened again. When I asked the marketing manager which of those 380 keywords they'd built content around, she said none. She'd printed it once, felt overwhelmed, and gone back to writing whatever felt urgent that week.

We threw the spreadsheet out and instead pulled 40 real questions straight from their sales team's call notes and their support tickets. We picked 12 of those and wrote one solid page for each over the next three months. Six of those twelve pages started ranking on page one within four months, and two of them were directly credited by the sales team with closing deals, because the exact phrase a prospect had typed into Google matched the exact phrase on the page. The £3,200 report contributed nothing. Forty minutes with a support inbox contributed everything.

I tell that story because it's the uncomfortable bit most keyword research advice skips over: the research isn't usually the hard part. Publishing consistently against a short, well chosen list is the hard part, and a big scary spreadsheet often becomes an excuse to delay writing rather than a tool that speeds it up.

The part nobody wants to say out loud

Search volume as a metric is getting less trustworthy every year, not more. With AI overviews answering more questions directly on the results page, and with zero-click search now accounting for a huge share of Google searches, ranking first for a keyword with 5,000 monthly searches can bring you far fewer visitors than ranking third for a keyword with 200 searches where people are still clicking through because the answer needs more than a summary box. I've watched pages hold position one and lose 40 percent of their traffic year on year purely because Google now answers the question itself before the click happens.

What this means practically: stop optimising purely for volume. Optimise for the searches where a click is still worth something to the searcher, which usually means anything comparative, anything requiring a decision, anything where the searcher needs proof rather than a definition. A keyword researcher who hands you a list sorted only by volume in 2026 is handing you a list built for 2018.

If you're hiring, what to check before you pay anyone

If you decide a freelancer is the right call, ask to see a sample of their past keyword research turned into published content, not just the spreadsheet on its own. A list with no outcome attached to it tells you nothing about whether the person understands your business. If you're sourcing candidates through LinkedIn, the process I use is laid out in detail in how I find freelance writers through LinkedIn, and the same search logic works almost exactly the same way for finding keyword researchers, because you're looking for the same signals: recent relevant work, specific numbers in their posts, and clients willing to be named.

Also worth knowing, if you're hiring at any real scale, that a growing number of agencies and in-house teams now run first-pass candidate screening through automated tools rather than a human reading every CV. It's worth understanding how that works from the other side, covered well in this piece on how recruitment agencies use AI automation to screen candidates, because the same filtering logic increasingly applies to freelance platforms too, not just permanent hires.

What to do next, this week

• Pull ten real questions from your last month of customer emails or calls, today, before you open any tool.
• Run those ten through one free tool, either Google Keyword Planner or AnswerThePublic, and note volume without letting it override the ten real phrases you already have.
• Pick your top five by how close they sit to an actual buying decision, not by which number is biggest.
• Write one page for the highest priority term by the end of the week. Don't wait for the full list. The full list is a trap.
• Revisit the list every quarter, not every week. Search behaviour doesn't move fast enough to justify weekly research, and constant re-researching is often procrastination wearing a productivity costume.

Frequently asked questions

What does a keyword researcher do day to day

They pull seed terms from a business's products, services, and customer questions, run them through tools like Ahrefs or SEMrush to check volume and competition, group the results by search intent, and hand over a prioritised list, ideally alongside recommendations for which pages or blog posts should target which terms.

Is keyword research still worth doing in 2026 with AI overviews taking over search results

Yes, but the goal has shifted. You're now researching to find questions worth a full click and a real answer, not just questions with high volume, because Google is increasingly answering the simple, high volume questions on the results page itself before anyone reaches your site.

How much does it cost to hire a freelance keyword researcher

Expect £15 to £45 an hour on freelance platforms, or a flat fee of roughly £150 to £400 for a full research pass on a small site with 20 to 40 pages. Agencies charge considerably more, often bundling keyword research into a wider SEO retainer priced from £500 a month upward.

What's the biggest mistake people make with keyword research

Building a huge list and never publishing against it. A smaller list of 15 to 20 terms tied to real customer questions, turned into published content within weeks, outperforms a 300-term spreadsheet that sits unopened in a shared drive, which is exactly what happens to most keyword research reports once the invoice is paid.