By Mehdi HassanFounders building something new often open a keyword tool, type their category name and see zero searches. The usual advice is to pick a different term with more volume. For a product that really is new, that advice pushes you into describing yourself with an old category you're trying to replace.
There's a better route. People don't search for categories that don't exist yet, but they do search for their problems, their goals and the tools they're frustrated with. This guide shows how to research those queries. It's part of our SaaS SEO strategy playbook.
Why doesn't standard keyword research work for a new product category?
Standard keyword research assumes buyers already know what to call the solution, and for a new category they don't. Keyword tools report searches people make today. If your category is new, the searches that matter describe symptoms, workarounds and goals, and they rarely contain your product's vocabulary.
Keyword tools also underreport niche B2B queries. A search with "0" or "10" monthly volume may still be typed by exactly the buyers you want, and long, specific queries are often grouped or hidden. Treat low reported volume on a highly specific query as a signal to investigate, not a reason to drop it.
AI search adds another layer. Buyers increasingly describe their situation to a chatbot in full sentences rather than typing a short keyword. Those conversations never show up in a keyword tool, but the answers draw on pages that explain the problem clearly. Problem-first pages serve both.
What three query types should you research instead?
Research problem queries, job-to-be-done queries and incumbent-alternative queries. Together they capture the people who feel the pain your product solves, even though they don't know your category exists.
Query type | What the buyer types | Intent | Page to build |
|---|---|---|---|
Problem | "why are our invoices always late" | Early, aware of pain | Answer-first problem guide |
Job to be done | "automate supplier payment reminders" | Mid, looking for a method | Use-case or job page |
Incumbent alternative | "[spreadsheet or old tool] alternative" | Late, ready to switch | Alternatives page |
Workaround | "[tool] template for payment tracking" | Mid, using a hack today | Template or free tool |
Category (emerging) | "[your category name]" | Late, educated buyer | Category-defining page |
The jobs-to-be-done lens is especially useful because it describes progress the buyer wants, independent of any product. If your team already uses it for product work, the same interviews feed keyword research; see jobs to be done for product design and how it compares with personas.
How do you do keyword research for a new product category, step by step?
Start from real buyer language, expand it with search tools, then group the results by intent and page. The order matters: tools are good at expanding a list, but poor at telling you which words your buyers actually use.
Collect buyer language
Pull phrases from sales call notes, demo recordings, support tickets, onboarding surveys and lost-deal reasons. Note how people describe the problem before they knew about you.
Search the communities
Read Reddit threads, Slack groups, review sites and Q&A forums where your buyers complain. Copy the exact wording of questions and complaints.
List the incumbents
Write down every tool, spreadsheet, agency or manual process buyers use today. Each one is a potential "[incumbent] alternative" or "[incumbent] vs" query.
Expand with tools
Run your seed phrases through a keyword tool, Google's autocomplete and "People also ask" boxes to find variations and related questions.
Check the results pages
Search each promising query. If the results are forums, thin posts or off-topic pages, that's an opening even when reported volume is tiny.
Group by intent and page
Cluster queries that one page can answer, assign each cluster a page type, and rank clusters by how close the searcher is to buying.
Which queries should a category creator target first?
Target incumbent-alternative queries first, then job-to-be-done queries, then problem queries. Alternative queries come from people who already feel the limits of the current solution and are actively looking, which makes them the closest to a purchase and the easiest to convert into a new category.
An alternatives page lets you meet that buyer and explain why the old approach falls short, then introduce your category as the answer. Our guide to building competitor alternatives pages covers the copy, and SaaS comparison pages cover direct head-to-heads once buyers start naming you alongside others.
Job queries come next because they show a buyer looking for a method. Problem queries are worth covering but convert slowest; build those as answer-first guides that link to your job and alternatives pages.
For example, suppose a startup creates an "AI vendor risk monitoring" category
Hypothetical example: nobody searches for "AI vendor risk monitoring" yet. The team's sales calls show buyers say "we track vendors in a spreadsheet" and "our security questionnaires take weeks".
Incumbent-alternative queries: "vendor risk spreadsheet alternative" and alternatives to two established questionnaire tools. Job queries: "automate vendor security questionnaires", "monitor vendor SOC 2 expiry". Problem queries: "how to keep track of vendor security reviews".
The first pages built are two alternatives pages and two job pages. The homepage still names the new category in its headline, with a plain-language subheadline that uses the buyer's words: "Stop tracking vendor risk in spreadsheets."
How do you build demand for the category name itself?
Use the category name consistently on your own pages and earn mentions of it elsewhere, so search engines and AI models learn to connect the term with you. Category searches grow when analysts, customers and the press start using your term. You can't force that, but you can make sure you're the clearest source when they do.
- Define the category in one sentence on the homepage and on a dedicated "what is [category]" page.
- Use the same name everywhere: site, profiles, decks, partner listings and press materials.
- Publish something others will cite, such as a dataset or framework that uses the term.
- Pursue mentions in podcasts, newsletters and communities, which Ahrefs' study of 75,000 brands found correlated 0.664 with AI Overview visibility.
Expect much of the early influence to happen without clicks, as buyers ask AI tools about the problem. Our guides to zero-click search and B2B attribution when buyers research in ChatGPT cover how to measure it.
How do you prioritize the keyword list?
Prioritize by closeness to purchase, then by how clearly you can win, and only then by reported volume. For a new category, volume is the least reliable signal. A query with tiny reported volume that only your buyers type can be worth more than a broad term with thousands of searches from people who'll never buy.
- Alternatives to the incumbents buyers mention most in sales calls.
- Jobs your product does better than anything else, phrased in buyer language.
- Workaround queries where a template or free tool can capture the searcher.
- Problem queries with weak current results.
- The category name and its variations, as a long-term page you keep improving.
