Keyword Research for Product Ideas: Find Demand Before You Build

A product idea can sound useful and still have no clear market. Before writing code, you need to know whether people are already searching for help with the problem, what words they use to describe it, and whether existing products are solving it well.

That is where keyword research for product ideas helps.

The goal is not to find a popular keyword and build a product around it. The goal is to discover the searches behind a problem, understand what the searcher is trying to accomplish, and collect enough evidence to decide whether the opportunity deserves further validation.

This guide walks through a practical process using Google Search, Google Keyword Planner, Google Trends and other research sources. It focuses on finding product opportunities before a product exists, not choosing SEO keywords for an existing SaaS website.

What keyword research can tell you about a product idea

Keyword research can help answer four questions:

1. Does the problem appear in people’s searches? Look for queries describing the task, frustration or outcome.
2. How do people describe the problem? Search language can reveal the words buyers actually use instead of the terminology a founder invents.
3. Are people looking for solutions? Queries about software, alternatives, automation, pricing and comparisons can reveal solution awareness.
4. What is the existing market offering? Search results show the products, content and workarounds already competing for attention.

But search data has limits. A person searching for a free template may not want software. Someone searching for a problem may be researching it for school rather than trying to solve it at work. And a high-volume keyword does not automatically indicate willingness to pay.

Treat keyword research as one source of demand evidence, not proof that a business will work.

Step 1: Describe the product idea as a problem, not a feature

Do not begin by entering your proposed product name into a keyword tool. A product that does not exist cannot have an established category name that people already search for.

Start with the problem.

Write down:

• Target user: Who experiences this problem?
• Task: What are they trying to accomplish?
• Pain: What makes the task difficult, slow or expensive?
• Current workaround: How do they handle it today?
• Desired outcome: What would a better solution help them achieve?

For example, consider this illustrative idea:

An automated client-reporting tool for small marketing agencies.

The product description is not a useful starting keyword by itself. Break the idea into the user’s actual work:

Research angle	Illustrative starting phrase
Target user	Marketing agency owner
Task	Prepare client performance reports
Problem	Reporting takes too much manual work
Current workaround	Spreadsheets and copied analytics data
Desired outcome	Prepare client reports automatically

Now you have several directions to investigate:

• How do agencies prepare client reports?
• How to automate client reporting
• Marketing agency reporting template
• Client reporting software
• Automated marketing reports
• Agency reporting dashboard

These are seed keywords. They are starting points for research, not proof of demand.

The important change is that you are researching a real job and its alternatives rather than trying to confirm a product idea you already like.

Step 2: Collect keywords from several sources

A keyword tool can expand your list, but it should not be the only place you look. Different sources reveal different parts of the problem.

1. Google Search and Autocomplete

Start typing a seed phrase into Google and note the predictions that appear.

Try different beginnings:

• how to [task]
• [task] software
• [problem] solution
• [task] automation
• [competitor] alternative
• [task] template

Autocomplete predictions reflect real searches, but they are influenced by factors such as location, language, trends and search history. They are not a complete list of popular searches or a source of search-volume figures. See Google’s explanation of Autocomplete.

Next, inspect the actual search results. Look at the wording in page titles, product descriptions, comparison pages and discussion results. Record the phrases that repeatedly describe the same underlying task.

Do not assume every suggested phrase deserves its own product or page. Several phrases may describe the same need.

2. Customer complaints and community discussions

Look for how people describe the problem when they are not using your terminology.

Useful places to investigate include:

• Relevant subreddits and specialist forums
• Product reviews and comparison discussions
• Public support forums
• Competitor review pages
• Questions under tutorials and product demonstrations

Search for phrases such as:

• How do you handle [task]?
• Is there a tool for [problem]?
• [competitor] alternative
• I hate doing [task] manually
• How to automate [workflow]

Record the exact wording and the context of each complaint. A phrase appearing in a discussion is not automatically a Google search keyword, but it can give you a better seed phrase to investigate.

For a more detailed method of discovering opportunities through complaints, see How to Find SaaS Ideas From Customer Complaints.

3. Competitor websites

Find products that solve the same or a related problem.

Look at their:

• Product and feature pages
• Pricing pages
• Integrations
• Comparison and alternatives pages
• Documentation and FAQs

Use their terminology to expand your own research. A product might call a feature something different from the phrase its customers use.

Do not copy a competitor’s keyword list without checking it. A competitor may attract traffic from a broad audience that does not match your intended buyer.

4. Existing customer or business evidence

When available, use first-party evidence before relying on keyword suggestions.

Examples include:

• Customer interviews
• Support tickets
• Sales conversations
• Search terms entered into an existing product
• Lost-deal notes
• Repeated manual work reported by users

These sources are especially useful because they can connect a search phrase to a real person, role or workflow.

When you do not have customers yet, use public discussions and reviews as research leads. Keep those observations separate from verified customer evidence.

Step 3: Expand your seed list with keyword tools

Once you have a list of problem-focused phrases, use keyword tools to find related queries and estimate search demand.

Google Keyword Planner

Google Keyword Planner is useful for discovering related keyword ideas and viewing estimated monthly searches.

To use it:

1. Open Google Ads and go to Keyword Planner.
2. Choose Discover new keywords.
3. Enter your seed phrases, such as client reporting, agency reporting and automated reports.
4. Select the relevant target location and language.
5. Review the keyword ideas.
6. Export or record the relevant phrases for further analysis.

Google currently requires account setup, including billing information, to access basic Keyword Planner features such as discovering new keywords. You do not have to start an advertising campaign just to plan your research, but check the account requirements before relying on it.

A few things to understand about its data:

• Average monthly searches are estimates for the selected location, Search Network and date range.
• The metric includes close variants.
• Its default monthly average uses a 12-month period.
• Its competition metric measures advertiser competition, not organic SEO ranking difficulty.

Read Google’s documentation on Keyword Planner and historical keyword metrics.

Important: Do not interpret a high advertiser-competition value as proof that organic search is difficult. They measure different things.

Google Trends

Google Trends helps you compare search-interest patterns over time and across locations.

Use it to investigate questions such as:

• Is interest growing, declining or staying relatively steady?
• Does interest appear seasonal?
• Are people searching for one phrase more often relative to another?
• Does the topic show a short-lived spike or a longer pattern?

Google Trends does not provide absolute monthly search counts. It normalizes sampled search data to a scale from 0 to 100 for the selected location and time range. A value of 100 represents the highest relative interest in that comparison, not 100 searches. Very low-volume terms may appear as zero.

Google also notes that small fluctuations can be statistical noise, particularly for terms with little search activity. Avoid treating one isolated spike as proof of a growing market. See Google’s FAQ about Trends data.

Third-party keyword tools

Tools such as Ahrefs Keywords Explorer can help expand a seed list, examine related searches, group keywords and inspect ranking pages.

Their metrics are useful for comparison, but each provider has its own database and calculation methods. A keyword difficulty score from one provider is not a universal measurement.

Use third-party estimates as supporting evidence. Check important keywords manually in Google and keep the source and date of each estimate in your research record.

Step 4: Group keywords by the problem and intent behind them

A large keyword export is not a product opportunity list. Before interpreting the numbers, group related searches by what the searcher wants to do.

For product research, these groups are a useful starting point:

Search type	Illustrative query	What it may reveal
Problem-aware	How to reduce manual client reporting	A task or frustration
Workflow-aware	How to prepare monthly agency reports	How people currently work
Solution-aware	Automated client reporting software	Awareness of a software category
Comparison	Best client reporting tools for agencies	Active solution evaluation
Alternatives	Alternatives to a reporting tool	Interest in replacing an existing solution
Cost-aware	Client reporting software pricing	Interest in the cost of a solution
Workaround-aware	Client reporting spreadsheet template	A manual method people may be using

These labels are a research framework, not a rigid buying funnel. A person may move between them or use several types of searches at once.

Look for groups of related queries that point to the same problem. For example, searches for reporting templates, reporting automation and reporting software may all relate to the same workflow, but they do not necessarily express the same intent.

Do not add the monthly search estimates for every variation and call the result total market demand. Related phrases and close variants can overlap. Adding them without checking can double-count demand.

Step 5: Inspect the search results manually

Search volume tells you that a phrase has estimated search activity. The results page helps you understand what people expect to find.

Search your most promising phrases one at a time. Review the results for:

1. Result type: Are the top results software products, tutorials, templates, discussions or comparison pages?
2. Audience: Do the pages target the same type of user you want to serve?
3. Problem match: Do the results address the problem your product would solve?
4. Existing solutions: Are there established products already serving this need?
5. Gaps: Are users still looking for a particular workflow, integration or capability that existing solutions handle poorly?

For example, a search for a reporting template may return spreadsheets and downloadable documents. That shows interest in completing the task, but it does not establish that searchers want to buy reporting software.

A search for reporting software may reveal a more solution-aware audience. You still need to inspect the products, their positioning and the buyer they serve.

Do not confuse the number of results with competition. A search engine can return millions of pages without those pages being meaningful competitors for the exact intent.

Also separate two different decisions:

• Business opportunity: Does a valuable problem exist for a reachable buyer?
• Search acquisition opportunity: Could your product realistically attract that buyer through organic search?

A useful product can have weak SEO potential, and an easy keyword to rank for can attract the wrong audience.

Step 6: Evaluate the quality of the demand

Do not rank ideas by search volume alone. Consider the following evidence dimensions.

A. Audience relevance

Can you identify the person behind the search?

A phrase used by a business owner, operations manager or specialist buyer may be more useful than a broad phrase that attracts unrelated visitors.

B. Problem relevance

Does the query describe a problem your proposed product can genuinely solve?

A phrase can have plenty of searches while being only loosely connected to the actual product.

C. Search consistency

Does the phrase appear across more than one source or research session? Are related queries pointing to the same workflow?

A single autocomplete prediction or Trends spike is weak evidence on its own.

D. Solution awareness

Are people looking for a category of software, alternatives, comparisons or pricing? These phrases can indicate that people are already evaluating possible solutions.

E. Existing competition

Are current solutions serving the same buyer and workflow? Competition can indicate that a market exists, but you need to understand whether the available products leave a meaningful problem unresolved.

F. Acquisition fit

Could the intended buyer realistically discover and evaluate your product through search? Consider whether the relevant queries lead to product discovery, or mostly to unrelated educational content.

Use these dimensions to record evidence, not to manufacture certainty.

Step 7: Build a keyword evidence ledger

A useful research output is a small evidence ledger that lets you compare ideas without losing the reason behind each score.

Create a spreadsheet with these columns:

Column	What to record
Product idea	The opportunity being investigated
Target user	The intended buyer or user
Problem	The task or pain being addressed
Seed keyword	The original research phrase
Related queries	Relevant variations and questions
Search location	Country or market selected in the tool
Search volume	Estimate, source and date, if available
Trends pattern	Stable, growing, declining or unclear
Search intent	Problem, workflow, solution, comparison or cost
Existing alternatives	Products, templates or manual workarounds
Evidence source	Link to the source or results page
Uncertainty	What you still do not know
Next action	Interview, competitor review, experiment or reject

Keep the raw observations separate from your interpretation. For example, record that a keyword tool returned an estimate rather than writing that the market is proven.

A simple opportunity score

You can use a small scoring model to compare candidates after collecting evidence.

Score each dimension from 1 to 5:

Dimension	Score 1	Score 5
Audience relevance	Unclear audience	Specific, relevant buyer
Problem fit	Weak connection	Directly describes the problem
Demand evidence	Little or conflicting evidence	Multiple consistent signals
Solution awareness	No clear solution interest	Clear searches for software or alternatives
Market gap	No meaningful gap identified	Specific unmet need worth investigating
Acquisition fit	Search unlikely to reach the buyer	Search appears to be a plausible discovery channel

Add the six scores for a total out of 30.

This is an IndieOS-style prioritization aid, not a validated market-scoring formula. A high score does not prove willingness to pay, and a low score may simply mean that the problem is described using different language.

Do not fill an unknown field with a neutral score to make an idea look complete. Mark it Unknown and collect more evidence.

Step 8: Work through one example

The following is an illustrative example, not a real keyword export or a measured market result.

Suppose you are considering software that automates client reporting for small marketing agencies.

Start with the problem

Agency owners need to prepare reports for clients. The work may involve collecting data from different platforms, formatting it and explaining performance.

Create seed phrases

• Marketing agency client reporting
• Automated client reporting
• Monthly marketing reports
• Client reporting template
• Agency reporting software
• Marketing report dashboard

Expand the phrases

Use Autocomplete and Keyword Planner to discover variations. Record the location, search estimates and date rather than mixing figures from different markets.

Group the intent

Query	Initial interpretation
Client reporting template	A possible spreadsheet or document workaround
Automated client reporting	Interest in reducing manual work
Agency reporting software	Interest in a software category
Client reporting tool alternatives	Possible replacement intent
Marketing report dashboard	Interest in viewing or presenting data

Inspect the results

Check whether the pages serve small agencies, large agencies, freelancers or in-house marketing teams. Those audiences may have different requirements.

Look at the existing products. Do they already automate the task? Are there repeated complaints about integrations, branding, setup or report customization? Record specific evidence rather than assuming a missing feature is a market gap.

Make a decision

At this point, the correct conclusion is not that the product is validated. The research has produced a more precise hypothesis:

Small marketing agencies may want an easier way to create client reports from multiple marketing data sources.

The next step is to investigate whether agency owners experience this problem frequently, how they currently solve it, what existing products fail to provide and whether they would pay for a better solution.

That is the difference between collecting keywords and using them to research an actual product opportunity.

Common mistakes to avoid

1. Choosing an idea because one keyword has high volume.
High search volume can come from a broad or unrelated audience. Check the audience and intent.

2. Treating Google Trends as a monthly search-volume tool.
Its 0–100 scale is relative, not an absolute count.

3. Confusing advertiser competition with SEO difficulty.
Keyword Planner’s competition metric describes advertisers, not organic ranking difficulty.

4. Using the product name as the only seed keyword.
People cannot search for a category name they do not know. Start with their task, problem and existing workaround.

5. Treating every keyword variation as a separate opportunity.
Group searches by intent before counting or scoring them.

6. Ignoring existing competitors.
A competitor is evidence that someone is trying to solve the problem. It is not automatically evidence that your version has a market gap.

7. Treating search demand as willingness to pay.
Searches show interest in information or solutions. They do not prove purchasing intent.

8. Skipping direct validation.
Keyword research should help you decide what to investigate next. Speak to potential users and test the problem before committing to a build.

Frequently asked questions

Final checklist

Before treating keyword research as useful evidence for a product idea, confirm that you have:

• Defined a specific user and problem.
• Collected seed phrases from more than one source.
• Expanded the list with keyword tools.
• Recorded the target location and date for search estimates.
• Grouped queries by intent.
• Inspected the actual search results.
• Identified existing products and workarounds.
• Separated observed evidence from assumptions.
• Recorded unknowns and conflicting signals.
• Chosen the next validation step.

Keyword research is a way to make product discovery more evidence-led. It helps you understand what people search for, how they describe their problems and what solutions they already consider.

Use it to narrow your ideas, not to declare a product successful before it exists.

The next step is to validate the strongest opportunity with potential users. Start with How to Validate a Micro-SaaS Idea Without Building It.

---

Sources and further reading

1. Google Keyword Planner: Use Keyword Planner
2. Google Ads: About Keyword Planner forecasts and historical metrics
3. Google Trends: FAQ about Trends data
4. Google Search Help: How Autocomplete predictions work
5. Ahrefs Help Center: Best practices for keyword research
6. Ahrefs Help Center: Keyword ideas reports
7. Google Search Console: Performance report overview