Search for “best perfume” and a small beauty blog is up against retailers, magazines and established review sites. Search for “does iris perfume smell powdery on skin?” and that blog has a job it can do well: compare scents and report what happened when they were worn. A low-competition keyword isn’t just a long phrase. It’s a question a focused, credible post could answer more usefully than the pages already in the results.
Start with a topic you can actually cover
Choose a product category, ingredient, technique or shopping decision that fits your blog. A fragrance note, a foundation finish, a seasonal colour trend or the difference between two cleanser formats could all work. Before collecting hundreds of phrases, ask whether you have the products, experience or sources to answer the questions they raise.
Write down a seed topic, then add distinctions a buyer might care about:
- Product properties: powdery iris, sheer blush, fragrance-free moisturiser.
- Use conditions: under makeup, on oily skin, in humid weather.
- Comparisons: cream versus powder blush, eau de toilette versus eau de parfum.
- Buying constraints: refill availability, travel size, shade range or UK stockist.
These are research prompts, not phrases to cram into a headline. “Best powdery iris perfume for everyone” is still broad, and “for everyone” is a claim you can’t sensibly test. “Iris perfumes that smell more like makeup powder than violet” gives a fragrance writer something specific to investigate.

Collect real phrasing before judging difficulty
Enter your seed topic into a search engine. Note autocomplete suggestions, related searches and the wording of questions in the results. Check your comments and emails too. Product reviews can be useful for vocabulary: a reader might ask for “lipstick that survives coffee” rather than “long-wear lip colour.” Record the phrases without assuming every suggestion brings meaningful traffic.
A spreadsheet needs only a few columns: phrase, likely reader need, evidence you can provide, current results and decision. Group near-duplicates. “Does powder blush work over cream foundation?” and “Can I put powder blush on cream foundation?” usually deserve one good answer, not two thin posts.
Keyword tools offer estimated search volume and difficulty scores, but both are approximations. A precise question may show little reported volume and still be worth answering. A low difficulty score, meanwhile, says nothing about whether you can support a claim about wear time or how a scent develops.
Check the results, not just the score
Search each promising phrase and read the first page as a reader would. Use the location relevant to your audience where possible: UK shoppers may see different retailers, product names and availability from US shoppers. Results also change by device and over time. This is a snapshot, not a ranking forecast.
Look for room to answer better
- Match: Do the leading pages answer the precise question, or do they just mention its words in a broad buying guide?
- Format: Are the results mostly product listings, videos, forum discussions or explanatory articles? A written review may not help someone who needs a current price or a visual demonstration.
- Evidence: Do top posts repeat brand descriptions where you could add swatches, application notes or a clear fragrance comparison?
- Authority: Are major retailers and specialist publications taking every spot, or can you see smaller, focused sites?
- Currency: Does the answer depend on this season’s stock, a reformulation or current shade availability?
Thin search results are an opening only if you can fill the gap responsibly. If posts about an active ingredient are vague, don’t respond with bigger skin-health promises. Explain what the ingredient is, separate evidence from marketing, and advise readers with persistent skin problems to consult a qualified specialist.

Separate low competition from low usefulness
Sometimes a phrase is easy to rank for because hardly anyone searches it. Sometimes it’s so oddly specific that no coherent article could satisfy it. Keep queries that suggest a real editorial task: explain why a perfume reads powdery, test whether a blush lifts a particular base, or compare what two finish descriptions mean in practice. Drop phrases assembled solely to sound unique.
Check the buying intent, too. “Where to buy” searches often favour live stock and location data; an editorial post can date quickly. “How to tell whether a refill fits an existing perfume bottle” has more scope for a lasting explanation, as long as you verify the packaging before naming brands.
Choose one query and define the article’s contribution
Shortlist queries with three questions: Can I answer this with evidence? Are the existing results missing something important? Does the answer belong on this blog? A phrase that passes all three is a better bet than one picked because a tool coloured its difficulty score green.
Take “why does iris perfume smell like lipstick?” The reader probably wants an explanation of that association, not a catalogue of every iris fragrance on sale. You could distinguish the powdery impression associated with iris materials from the waxy, floral and vanilla facets that can add to a lipstick-like effect. Then compare a few perfumes you’ve smelled, noting when that impression appears and whether it lasts. Scent perception varies; an evocative description isn’t a laboratory analysis of a formula.
Before drafting, check whether an existing post already answers the question. If your fragrance review covers it thoroughly, improve that review rather than publish a near-copy. A [review of how to audit a small beauty website’s content](/content-audit-checklist-small-beauty-websites/) can help identify pages that need that attention.
Measure the outcome without pretending rankings are fixed
When you publish, record the date, target query and what the page adds to the earlier results. Later, check which searches actually bring impressions and whether the page delivers what its title promises. Search engines may connect it with wording you hadn’t considered. If that wording reveals a reader question you’ve missed, clarify the relevant section rather than bolt on repetitive sentences.
For the iris query, the draft brief could be: “Explain the lipstick association, compare three tested scents and state when the powdery note appears.” If you haven’t smelled those three scents yet, keep the query on your shortlist. The smelling notes come first.