Frase is primarily a SERP research, content brief, optimisation, and writing platform. It can reveal the topics and questions covered by ranking pages and may display keyword metrics through its current features or data partners, but it should not be the only authority for “high-volume, low-difficulty” selection. Search volume is an estimate, difficulty is vendor-specific, and the lowest score does not automatically represent a useful or winnable query.
Define a viable keyword for your site
High volume is relative. A specialist B2B service may profit from a phrase searched only a few hundred times a month, while a broad consumer publisher may need thousands. Low difficulty also depends on the site’s authority, topical coverage, links, brand recognition, and ability to satisfy the intent.
Set thresholds from the business model. Record the target country, language, minimum volume, maximum difficulty, expected conversion value, and content type the team can produce. Exclude topics that require credentials or firsthand testing the site does not possess.
Use the right tools for each decision
Frase is valuable after a candidate query has been identified because it analyses search results and creates a research workspace. Pair it with first-party and dedicated keyword sources.
| Question | Strong source | Why it matters |
|---|---|---|
| Do we already appear? | Google Search Console | Real impressions, queries, pages, and country/device data |
| Is interest seasonal? | Google Trends | Direction and relative timing, not absolute volume |
| What are estimated volume and difficulty? | Ahrefs, Semrush, Moz, or another keyword database | Consistent comparison within one vendor’s model |
| What do ranking pages cover? | Frase SERP Research | Source analysis, headings, topics, and questions |
| Can we produce the best answer? | Human editorial review | Expertise, evidence, format, and business fit |
Check current Frase plans and add-ons before subscribing. Document limits, search-query allowances, AI word or generation credits, team seats, and any separate SEO-data product, because packaging changes over time.
Start with Search Console opportunities
Export the previous three to six months of non-brand queries and pages. Filter for queries with meaningful impressions and average positions roughly outside the strongest click range, such as positions 8–30. These are not guaranteed wins, but Google already associates the site with the topic.
Group queries by likely intent and landing page. If five variations refer to the same task, optimise one strong page rather than producing five thin articles. Compare recent and prior periods to detect declining pages or emerging language.
Search Console does not provide total market volume, and average position can hide variation by location, device, and query. Use it as evidence of existing relevance, not a complete keyword database.
Expand the seed list
Enter proven customer questions, product categories, problems, use cases, and competitor alternatives into a dedicated keyword tool. Export phrase variants, questions, parent topics, estimated volume, difficulty, cost-per-click where relevant, and SERP features.
Keep all metrics from the same provider during scoring. Ahrefs and Semrush calculate difficulty differently, so a score of 25 in one is not interchangeable with 25 in another. Record the data date and target country.
Remove obvious mismatches: navigational searches for another brand, free-download intent when the business sells consulting, locations the company cannot serve, and terms whose results are dominated by shopping or video when the team plans a text article.
Investigate candidates in Frase
Create a Frase document or SERP research query for one candidate. Frase currently aims to return roughly 15–20 useful sources in many searches, though it may exclude PDFs, thin pages, LinkedIn posts, and other results it cannot analyse effectively. Fewer sources may reflect the SERP rather than a technical error.
Inspect the actual ranking URLs, titles, headings, word-count range, topics, and questions. Do not copy the competitors’ outline. Ask what Google is rewarding: a tutorial, list, product page, calculator, category page, forum discussion, or mixed result.
Look for weaknesses you can address:
- Ranking pages are outdated or show discontinued interfaces.
- Advice lacks screenshots, original testing, or a working example.
- The results answer beginners but ignore advanced constraints.
- A forum ranks because conventional articles avoid the real question.
- Results cover a broad topic while the query implies a specific role or use case.
If authoritative brands satisfy the intent comprehensively and you have nothing distinctive, a low numerical difficulty score should not override that evidence.
Evaluate the live SERP manually
Open an incognito or neutral search session with the target location where practical. Note ads, featured snippets, local packs, shopping, videos, discussions, and AI-generated result features. These elements affect potential clicks even when the nominal search volume is large.
Review the ranking domains and page types. Estimate whether your site can plausibly compete based on topical authority and link profile. A query may be easy for a major health publisher but difficult for a new ecommerce blog. Conversely, a focused site with original data may outrank larger generic pages.
Check freshness. Software queries often reward current screenshots and pricing; evergreen definitions may be stable. Also examine whether multiple intents coexist. If the SERP mixes “buy,” “learn,” and “download,” decide which intent your page will serve and accept that rankings may fluctuate.
Score opportunity with business context
Use a spreadsheet with 1–5 scores for relevance, intent match, evidence advantage, attainable authority, traffic potential, and commercial value. Subtract production effort and freshness burden. Keep the raw keyword metrics beside the score rather than hiding them in one opaque total.
A practical formula might weight intent and relevance more heavily than volume. For example, a query with 500 estimated monthly searches, clear buyer intent, and weak current answers can outrank a 10,000-volume informational term unrelated to the offer.
Require an “edge statement” for every approved topic: the original test, dataset, template, expert input, or clearer workflow that will make the page better. No edge means no assignment.
Build the Frase brief
Select only competitors that match the intended format. A product category page and a 4,000-word guide may rank for the same phrase but serve different jobs. Use Frase to gather recurring subtopics and questions, then organise them around the reader’s sequence.
The brief should define audience, intent, promise, primary query, related language, required evidence, real products or examples, internal links, sources to verify, and desired action. Treat Frase’s optimisation score as a coverage diagnostic, not a mandate to insert every term. Forced repetition harms clarity and can make claims inaccurate.
During drafting, cite primary sources, capture original screenshots where permitted, and answer the main question early. Use headings that reflect decisions or steps rather than competitor wording. Update prices and product capabilities immediately before publication.
Validate after publishing
Record the publication date and inspect indexing. Track impressions, clicks, query mix, and position in Search Console over several weeks or months, depending on site size. Measure conversions or assisted pipeline, not only traffic. A low-volume page that generates qualified enquiries can be more valuable than a broad traffic winner.
Refresh the page when the SERP intent, product interface, or competitive standard changes. Avoid rewriting solely because a Frase score is below an arbitrary target. Investigate whether missing coverage is relevant to the reader.
Pros, cons, and verdict
Frase accelerates competitor research, source discovery, brief creation, question collection, and coverage review. It is useful for editors who need a consistent evidence workspace. Its limitations are reliance on available SERP sources, potential pressure toward consensus content, subscription or usage limits, and keyword metrics that still require contextual validation.
Find candidate opportunities in Search Console and a consistent keyword database, then use Frase to inspect intent and build the brief. A high-volume, low-difficulty label is only the start. Publish when the live results reveal a genuine information gap your team can fill with credible evidence.
Our pick: Frase for SERP analysis and briefs, paired with Search Console and one dedicated keyword database for opportunity scoring.