How to Find SEO Keywords for a New Website With No Traffic

Your website is live, but Google Search Console shows few impressions and almost no useful query history. That makes advice such as “find keywords you already rank on page two for” impossible to follow. Most keyword research playbooks assume an established site has traffic to analyze.

A new site needs a different starting point: translate buyer problems into plausible searches, investigate the results and competing sites, then use keyword data to validate and organize the opportunities. Keep three stages separate: generating ideas, validating demand and competition, and making an editorial decision about what deserves a page.

Throughout this guide, we will use Draftly, a hypothetical SaaS that helps freelance designers create proposals, obtain approval, and collect deposits. Every assessment below is illustrative, not a claim about measured performance.

Why Keyword Research Is Different for a New Website

An established website can inspect Search Console queries and pages that already receive impressions. A new website has none of that feedback, so its first keyword list must come from market knowledge and external evidence.

That does not mean choosing blindly. It means distinguishing two jobs:

• Keyword discovery produces plausible phrases and questions people may search.

• Keyword selection determines whether a phrase has relevant intent, credible demand, and a search results page the site can realistically compete on.

Skipping selection is expensive. A broad, high-volume phrase may attract the wrong audience or be dominated by established sites, while a narrower query may describe exactly the problem your product solves.

Keyword research without Search Console is entirely possible. Search Console becomes useful later, when impressions and clicks reveal the language Google associates with your pages; it is not a prerequisite.

Step 1: Identify the Problems Your Website Solves

Do not open a keyword tool first. It can expand a phrase, but it cannot decide which customer problems matter. Map four things:

1. Core capabilities: what the product enables someone to do.

2. Customer problems: the friction, risk, or delay the buyer experiences.

3. Search triggers: the event that makes the problem urgent enough to research.

4. Buying intent: whether the person wants advice, a template, a comparison, or software now.

For Draftly, the problems include writing proposals from scratch, waiting for clear approval, and beginning work without payment. A search might be triggered by losing a project after an unclear proposal or by a client who approved informally but never paid.

Translate those situations into ordinary language. “How to write a freelance design proposal” sounds like a search; “integrated document lifecycle” sounds like product language.

Hypothetical example:

These are seed searches, not validated keywords. Business knowledge makes them relevant hypotheses; research must establish whether and how people search for them.

Step 2: Find Seed Keywords Without Existing Traffic

Expand the initial list with sources you can inspect manually and reproduce later.

• Google autocomplete: type a seed and note relevant completions. “Freelance design proposal” might reveal modifiers about templates, examples, or project types.

• Related searches: inspect the related queries shown for a relevant result page. They can expose adjacent wording or a more specific problem.

• Visible questions in results: question modules can suggest concerns such as what a proposal should contain or whether a deposit is refundable.

• Public customer language: read relevant communities and professional forums. Record how people describe the problem, not merely which posts are popular.

• Competitor sites: inspect navigation labels, product pages, help centers, templates, and articles. A section about “proposal acceptance” might suggest an approval-focused query you missed.

For Draftly, “freelance proposal template,” “graphic design proposal example,” “client deposit invoice,” and “proposal approval process” might enter the working list. Each still needs validation.

Autocomplete and visible questions are idea sources, not measurements of monthly demand. They vary by location, language, device, and context. Do not call a suggestion “high volume” because Google displayed it, and do not scrape Google to manufacture a dataset.

Step 3: Identify Your Actual Search Competitors

Business competitors sell an alternative to your product. Search competitors occupy results for your queries. The groups may overlap for “proposal software for freelancers,” while an informational query may rank design publications or template libraries that never compete with Draftly for a subscription.

Use this procedure for representative seeds:

1. Search Google in the target market for each query.

2. Record domains that recur across several relevant searches.

3. Save the specific articles and landing pages that match your intended audience.

4. Extract additional topic and keyword hypotheses from their titles, headings, navigation, and related resources.

5. Reject topics that do not connect to your buyers or product, even when a prominent competitor publishes them.

Classify the pages you see. Educational pages teach a process; commercial pages present software, plans, or alternatives to someone evaluating a solution. If Google overwhelmingly ranks templates, a conventional essay may miss the intent even if the query sounds informational.

Competitor research supplies hypotheses, not your strategy. Copying a large competitor's article list can pull a new site into unrelated subjects and ignore what uniquely connects its product to buyers.

Step 4: Validate Search Demand and Competition

Investigate candidates in Google Keyword Planner or an established keyword tool. Set the correct country, language, and market.

For each candidate, evaluate:

• Estimated monthly demand: evidence that the query or close variations are searched.

• Relevant variations: alternate phrasing that reveals the topic without unrelated terms.

• Search intent: the task the searcher expects the result to help complete.

• Competing results: page type, specificity, quality, and recurring sites.

• Business relevance: how naturally the need connects to the product and a plausible buyer.

Tools use different data sources and models, so estimates can disagree. Keyword Planner may present ranges, while dedicated tools report modeled volume and proprietary difficulty. Treat all figures as directional, not exact.

Do not reject a query merely because its volume is low. A search about collecting a design-project deposit could have strong commercial relevance and realistic competition. A broad term such as “proposal” may have more demand but mix many intents Draftly cannot serve.

Competition requires human review. A low difficulty score does not guarantee easy rankings. Ask whether a focused new site can create the right page type with credible expertise and enough depth to deserve a place.

Step 5: Group Keywords Into Topics Instead of Creating One Article per Keyword

One page can satisfy several phrasings of the same task. Search intent and ranking pages are better grouping criteria than superficial wording.

Illustrative grouping for Draftly:

Compare search results before grouping terms. Similar page types and URLs usually support one page; templates and software product pages indicate distinct intents despite shared wording.

Creating an article for every variation causes unnecessary overlap. Several thin pages can compete for the same intent and leave none as the definitive resource.

Step 6: Choose Your First Five SEO Content Opportunities

Selection is editorial judgment, not a mechanical tool export. This is an illustrative scenario only; none of its decisions represents actual measurements.

The first five cover distinct, product-connected needs: creating a template, assembling it, collecting a deposit, evaluating software, and obtaining approval. Publish only after validating demand and live results.

“Graphic design proposal examples” belongs with the template page if results confirm shared intent. Accounting software is too far from the product. Logo pricing has a weaker product tie. Proposal-versus-contract needs careful sourcing and a clear boundary between operational and legal advice.

For the broader process of turning a shortlist into a roadmap, read How to Find Blog Topics With Search Traffic Potential. This guide focuses on discovery without search history.

Free vs. Paid Keyword Research Tools for a New Website

Each tool has a specific role.

• Google Search reveals intent, page formats, recurring competitors, suggestions, and questions. Its results do not provide reliable monthly volume.

• Google Keyword Planner supplies ideas and market-specific demand estimates. Built for advertising, its ranges are not exact organic forecasts.

• Google Search Console reports your own queries, impressions, clicks, and positions. It is valuable once data accumulates, but it cannot reveal history a brand-new site never had.

• Established dedicated keyword tools accelerate expansion and competitor research. Their volume and difficulty metrics are modeled estimates that differ across providers.

A new site can begin with Google Search and customer-problem mapping, then use Keyword Planner or a paid tool to validate the shortlist. More data does not make an irrelevant keyword useful.

Common Keyword Research Mistakes to Avoid

• Starting exclusively with high-volume keywords. Broad demand often brings mixed intent and formidable competition.

• Confusing search volume with business value. Visits matter less when the searcher's problem has no meaningful connection to the product.

• Targeting audience-adjacent but product-irrelevant terms. Shared demographics alone do not create a sound content strategy.

• Giving every variation its own article. Group phrases that share intent and results to avoid unnecessary overlap.

• Copying a competitor's content strategy. You do not know its economics, authority, conversion path, or reason for publishing each page.

• Assuming low difficulty means easy rankings. Tool scores compress a complicated results page into one estimate; inspect the pages themselves.

• Inventing certainty when data is weak. Mark demand as unvalidated instead of turning autocomplete, a forum mention, or a hunch into a fake fact.

Your First Keyword Research Checklist

• List product capabilities, customer problems, search triggers, and buying stages.

• Turn each relevant problem into natural-language seed searches.

• Expand seeds through search results, public customer language, and relevant competitor pages.

• Label every new phrase as an idea until demand is validated.

• Identify recurring search competitors and the page types Google rewards.

• Check estimates in the correct location and language with Keyword Planner or a dedicated tool.

• Judge intent, competing results, and business relevance alongside volume.

• Group variations when they serve the same intent and return substantially similar results.

• Check whether an existing page already covers the topic.

• Select five distinct, defensible opportunities and record why each earned priority.

Conclusion

A new website does not need existing traffic to discover potential keywords. It needs a disciplined chain of evidence: translate buyer problems into natural search hypotheses; expand them through search results, customer language, and actual search competitors; validate demand without pretending estimates are exact; inspect intent and competing pages; group variations into coherent topics; and choose the opportunities that combine business relevance with a realistic path to ranking. Search Console improves the next cycle once impressions and clicks arrive, but it is not permission to begin the first one.