I still remember the smug little rush I got the first time I nailed a keyword brief.
I’d found a term with decent volume and low competition. I’d packed it into the title, headings, meta description, and even tweaked a few internal links to support it. The content itself was polished. Clean. Technically sound.
There was just one problem: nobody who could actually buy from us would ever search for that keyword.
We’d built the entire article around how a marketer imagines people search — not around how actual customers do. And because of that, the piece behaved exactly how it deserved to: a few impressions, a trickle of clicks, zero meaningful leads.
That’s the trap a lot of SEO strategies fall into. We obsess over tools, export thousands of ideas, and forget the most obvious source of keyword insight sitting right in front of us: our customers.
This article is about fixing that. You’ll learn how to run simple, practical customer interviews that turn into sharper keyword research, better briefs, and content that feels like it was written with the reader — not just the algorithm — in mind.
Why most keyword research ignores customers
Most keyword research happens in a vacuum.
A marketer opens their tool of choice, types in a seed phrase, clicks “export,” and ends up with a spreadsheet full of ideas. Then they cluster those ideas, rank them by search volume, filter by keyword difficulty, and hand the “best” ones to a writer.
On paper, it looks airtight. In reality, it’s missing one thing: the people who are actually typing those searches.
The three big gaps in tool-only keyword research
1. You don’t know who is searching. A keyword like how to do financial projections might be searched by a startup founder, a finance manager, a student, or someone studying for an exam. Different people. Different needs. Different levels of urgency. Same keyword.
2. You don’t know why they’re searching. A query like best way to manage multiple projects could come from a solo freelancer, an agency owner, or an internal project lead who’s drowning in tickets. Tools tell you volume, not context.
3. You don’t know what language buyers actually use. Tools show you cleaned-up versions of searches. Customers don’t always talk that way. They use jargon, half-finished phrases, and specific references to their workflows that never show up in keyword suggestions.
Customer interviews plug these gaps. They don’t replace tools — they make them more accurate. They give you the language, the pain points, and the lived reality behind the numbers.
Step 1: Find the right customers to talk to
You don’t need a giant research project to improve your keyword research. You just need a handful of conversations with the right people.
Who to interview
Start with three simple buckets:
- New customers. They still remember the problems, frustrations, and searches that led them to you.
- Long-term customers. They can tell you what actually matters most now that they’ve lived with your product or service.
- Lost opportunities or churned customers (if you can get them). They show you what didn’t resonate, what was confusing, and where they went instead.
You don’t need 50 interviews. Start with 5–10 conversations from a mix of these groups. You’re not chasing statistical significance. You’re after patterns and real language.
How to get them to say yes
Keep it light and honest. Something like:
“We’re trying to make our content and resources more useful for people like you. Could we steal 20–30 minutes to ask how you actually searched for and evaluated solutions like ours? No sales pitch — just research.”
People are surprisingly happy to help when they know it’s not a disguised sales call.
Step 2: Prepare like a researcher, not a salesperson
Most bad interviews feel like surveys. A list of stiff questions. No real curiosity. No follow-ups. You want the opposite of that.
Your job on these calls isn’t to pitch, correct, or defend your product. It’s to sit quietly, ask good questions, and let them talk.
Know enough to ask better questions
Do your homework first. For each customer you’re speaking to, ask yourself:
- What did they buy from us?
- Which plan, tier, product line, or service?
- What do we know about their role and industry?
- Do we have any notes from sales about their original pain points?
You’re not trying to script the conversation. You just want enough context so your questions don’t sound generic.
Set the frame before you begin
At the start of the call, say something like:
“I’m not here to sell you anything. I’m trying to understand what your world looked like before you found us — what you were struggling with, how you searched for solutions, and what made you trust certain websites over others. There are no wrong answers. The more honest you are, the better we can make our content.”
That gives them permission to be blunt. You want that.
Step 3: Ask questions that reveal search behaviour
You’re not just doing generic customer research. You’re specifically trying to improve your keyword research and content.
So you need questions that uncover how they searched, how they thought about their problem, and why certain pages resonated more than others.
Questions that get them talking about their “before” state
Start with context:
- “What was going on in your business when you first realised you needed something like this?”
- “What were you using before you came to us?”
- “What finally pushed you to go looking for a better option?”
You’re trying to surface the real trigger. Was it a missed deadline? A stressed-out employee? A scary invoice? A failed audit? That trigger is where most meaningful searches begin.
Questions that uncover actual queries and language
Then move into search behaviour:
- “Do you remember what you typed into Google the first few times you went looking for help?”
- “If you had to guess, what kind of phrases were you using?”
- “Did you search more for ‘how to’ content, or for specific solutions and tools?”
Most people won’t remember their searches word-for-word — and that’s fine. They’ll give you fragments. Things like:
- “Probably something like ‘easy way to clean oven’…”
- “I think I typed ‘best way to manage multiple projects’ because I was drowning.”
- “I started with ‘ways to conserve energy’ then got more specific once I saw what was out there.”
That’s exactly what you want: real, messy, human language. Not what a tool thinks they searched — what they remember searching in the moment.
Questions that show you which pages actually helped
Finally, ask about the content they trusted along the way:
- “Can you remember any article, guide, or page that really helped you understand the problem?”
- “What made you trust that page?”
- “Was there anything you read that made you think ‘these people actually get it’?”
- “And on the flip side — was there anything that made you click away instantly?”
Now you’re not just getting keyword ideas. You’re learning what good content looked like from the buyer’s side of the screen.
Step 4: Turn interview notes into keyword ideas
Once you’ve spoken to a few customers, you’ll have a messy notebook or a stack of call transcripts. Your next job is to turn those raw notes into usable keyword ideas.
Pull out exact phrases and patterns
Go through your notes with a highlighter (digital or literal) and look for:
- repeat phrases — things multiple customers say in similar ways
- complaints and frustrations (“I was sick of…”, “what annoyed me was…”)
- jobs they were trying to complete (“I just needed a simple way to…”)
- any time they say “I probably searched for…”
Turn those into draft keyword ideas by wrapping them in patterns you know people use:
Example: A customer tells you, “I was just looking for an easy way to keep track of all my client projects in one place.”
That can turn into queries like:
- best way to manage multiple projects
- how to track multiple client projects in one place
- tools to manage multiple projects
You’re still going to run these through a keyword tool, but now you’re starting from reality, not guesses.
Bucket them by job, not just topic
Group your ideas by what the person is trying to do. For example:
- Learn something from scratch: how to start a vending machine business, how to set up a trust
- Do a recurring task more easily: easy way to clean oven, best way to clean grout
- Make a decision or choose a method: best way to file taxes online, safest way to buy bitcoin
Those “jobs” are far more useful to content planning than vague categories like “awareness” or “consideration.” They tell you exactly what the content needs to help the reader do.
Step 5: Validate with tools without losing the nuance
Now you bring your tools back in — but on your terms.
Take the phrases and patterns from your interviews and plug them into your keyword tool. You’re looking for three things:
1. Confirm there’s real search demand
Some phrases will show up almost exactly as customers said them. Others won’t. That’s fine. Look for close matches, long-tail variations, and related phrases. For example, “a simple way to keep track of client work” might surface best way to manage multiple projects or something similar.
2. Check what kind of pages are ranking
Pull up the SERP. What do you see for a query like how to do financial projections?
Is it mostly beginner guides for students? Accounting firm articles for small businesses? Templates and calculators?
This tells you who’s being served well — and who isn’t. If your interviews show that CFOs of established businesses are struggling with scenario planning, but the SERP is full of “finance 101” posts, that’s a content gap you can step into.
3. Decide if this keyword realistically fits your offer
Just because people search it doesn’t mean you should chase it.
Ask:
- “Does this describe a problem our product or service genuinely solves?”
- “Would we be happy to get leads from people who search this?”
- “Could this content logically move a reader closer to working with us?”
If the answer is no, park it. Customer language without product fit is just another way to waste effort.
Step 6: Feed interview insights into your briefs and content
Customer interviews shouldn’t just change your keyword list. They should change how you write about those keywords.
Rewrite your briefs with customer language baked in
Instead of a sterile brief like:
Target keyword: best way to manage multiple projects
Audience: project managers
Goal: rank and drive traffic
You can now create a brief that sounds more like:
Target keyword: best way to manage multiple projects
Audience: agency owners and in-house project leads who are overloaded and feel like things are constantly slipping through the cracks
Context: Interviews show they’re juggling client work in spreadsheets, email, and chat. They’ve tried basic task tools but nothing sticks. They’re not just looking for “features” — they want a way of working that doesn’t burn them out.
Now your writer isn’t just stuffing in a keyword. They’re writing for someone specific, in a specific situation, with a specific problem.
Answer the questions customers actually asked you
Look back at your interview notes. Turn their real questions into subtopics and sections. If three customers asked, “How do I know if I’m choosing the wrong tool?” — that becomes a section in your article.
If someone said, “I didn’t trust half the sites I saw because they just repeated generic advice,” that becomes a line in your copy — and a standard you hold your own content against.
Bringing it all together
Running customer interviews to improve your keyword research isn’t about building a perfect research programme. It’s about grounding your SEO strategy in the same reality your buyers live in.
When you do this well, a few things change:
- Your keyword ideas stop feeling random and start feeling obvious.
- Your content briefs stop sounding like generic templates and start sounding like battle plans.
- Your articles stop attracting people who will never buy from you and start pulling in readers who recognise themselves in your words.
Tools are still useful. Search volume still matters. Competition still matters. But they sit in the passenger seat. Your customers drive.
So before you export another thousand keyword ideas, try something different.
Pick five customers. Book five short calls. Ask them what was happening the day they finally opened Google. Ask them what they typed. Ask them which pages helped and which ones felt like a waste of time.
Then look at your current content and ask the only question that really matters:
“Would this have genuinely helped them in that moment?”
If the answer is no, you don’t need more keywords. You need better conversations.