Before writing any page, run three free checks: does the phrase appear in Google autocomplete, does the live SERP show dedicated pages or just forum scraps, and does the Google Trends line hold steady. Ten minutes, no tools, no guessing.
Check autocomplete first
Type the phrase into Google slowly and watch what the box suggests. Autocomplete is built from real query behavior, so if your phrase or a close variant appears, people type it. If nothing related shows up no matter how you word it, be suspicious — you may be about to write for a phrase that exists only in your head. Try the natural rewordings too. Sometimes the demand is real but the market phrases it differently than you do, and autocomplete hands you the correction for free.
Read the SERP that already exists
Search the term and look at what ranks. Three outcomes. Dedicated, current pages from real sites: demand confirmed, competition visible, judge whether you can do better. Reddit threads and stale forum posts: demand confirmed and the field is open, which is the best result you can get. Results that ignore your phrasing entirely and answer some other question: warning sign — either the demand is not there or Google reads the intent differently than you intended. The SERP is Google showing you its homework. Read it before writing yours.
Pull up the Trends line
Google Trends will not give you a volume number, but the shape of the line answers a different question: is this term alive. Steady or climbing across five years is what you want. A single sharp spike that decays to nothing is a news event, not a keyword. And a flat zero on a niche term is not always fatal, because Trends hides low-volume queries — check a broader parent term for the shape and lean harder on the other two checks.
Worked example: one term in ten minutes
Suppose a dispensary blog wants to target solventless gummies. Minute one: autocomplete shows solventless gummies and solventless edibles as suggestions. Good. Minutes two through five: the SERP has two product pages, one Reddit thread, and no dedicated explainer comparing solventless to distillate edibles. Gap identified. Minutes six through ten: Trends shows solventless climbing steadily over five years as a parent topic. Verdict: write the explainer, link it to the gummy products, and target the comparison phrasing the SERP is missing. That is the entire validation, and it cost nothing.
Key takeaways
- Autocomplete presence proves real people type the phrase.
- A SERP of forum threads means demand with no real competition — the best case.
- Use the Trends line for shape, not volume; steady beats spike.
- Ten free minutes of checking protects hours of writing.
Frequently asked questions
Why validate a keyword before writing the page?
A page costs hours to produce and the check costs minutes. Skipping it is how sites end up with dozens of posts targeting phrases nobody has ever typed.
What does it mean if a phrase appears in autocomplete?
Google suggests completions from real query behavior, so an autocomplete appearance means actual people type that phrase. It is the fastest positive signal available and it costs nothing.
What if only forums rank for the term?
That is usually good news. Reddit threads and old forum posts on page one mean the demand exists but nobody has written a proper page. Confirm the phrasing in autocomplete, then treat it as an opportunity.
What if Google Trends shows nothing for my term?
Trends hides low-volume terms, so a flat line does not always mean zero demand. Try a broader parent term for the shape, then lean on autocomplete and the SERP. If all three come back empty, walk away.
Do I need a paid volume tool to validate demand?
No. Autocomplete, the live SERP, and Google Trends answer the core question, which is whether the term is worth a page at all. Paid tools refine priorities later, and their estimates for cannabis terms are often shaky anyway.
Want this done for your brand?
We build cannabis & CBD search visibility with full transparency, for dispensaries, brands, and MSOs.
Get a free cannabis SEO audit ›