Most "niche ideas" lists are someone's opinion. Google already publishes the raw signals you need to rank niches yourself: what people type (Autocomplete), how interest changes over time, and which related searches are taking off (Trends). The hard part is combining them into one number you can sort by.
Here's a method that scores any niche keyword from 0 to 100, with a real run on three seed topics and the pitfalls I hit along the way.
What makes a niche worth building around
Four things matter, and you can measure each one from free Google data:
- Growth. Is interest higher than a year ago? I compare the latest 4 weeks with the same 4 weeks a year earlier, so seasonality doesn't distort it.
- Demand. Is anyone actually searching? Google Trends only gives relative numbers (0-100), so every candidate has to go on one shared scale. More on that below.
- Buyer intent. "best standing desk" and "standing desk price" come from people with a wallet out. "what is a standing desk" doesn't.
- Steadiness. A niche that's only searched in December is harder to build a business on.
The score weights them like this:
| Part | Weight |
|---|---|
| Growth (year over year) | 30% |
| Demand (log scale) | 20% |
| Buyer intent | 20% |
| Momentum (last quarter vs the one before) | 15% |
| Steadiness (low seasonality) | 10% |
| Breadth (how often it shows up in Autocomplete) | 5% |
Niches with breakout related searches (Google's label for growth above 5000%) get +8 points, and falling trends get -8.
Where the candidates come from
For each seed topic I collect candidates from two places:
- Google Trends related searches for the seed: top, rising and breakout. These are measured demand, not guesses.
- Google Autocomplete variations like "best …", "… vs", "… near me", "… price", and optionally a-z for the long tail.
Then every candidate gets its own 5-year Google Trends series.
Putting everything on one scale
Google Trends compares at most 5 terms at once and rescales each comparison to its own 0-100. To compare 15 niches, I chain comparisons through the biggest keyword so far: compare it with 4 new ones, record the ratios, repeat. (I wrote up the chaining method separately.)
The new step here is comparing the seeds with each other too. That puts a niche from "home gym" and a niche from "air fryer" on the same demand scale.
Real run: air fryer, home gym, standing desk (US, past 5 years)
I ran the Niche Value Finder on Apify with those three seeds. Data was pulled on 29 October 2026, and the run took 2 minutes 13 seconds and returned 48 rows. The top 10:
| Niche | Seed | Score | YoY % | Buyer intent | Demand (all seeds, top = 100) |
|---|---|---|---|---|---|
| best standing desk | standing desk | 86 | +126.7 | 0.8 | 0.47 |
| standing desk | standing desk | 76 | +60.3 | 0.3 | 5.45 |
| standing desk near me | standing desk | 70 | +369.4 | 0.8 | 0.03 |
| home gym | home gym | 66 | +12.9 | 0.3 | 4.65 |
| flexispot standing desk | standing desk | 64 | +77.8 | 0.3 | 0.44 |
| home gym setup | home gym | 63 | +83.0 | 0.3 | 0.02 |
| smart home gym | home gym | 60 | +83.3 | 0.3 | 0.10 |
| gym set for home | home gym | 59 | +110.0 | 0.4 | 0.14 |
| standing desk with drawers | standing desk | 58 | +77.8 | 0.3 | 0.04 |
| best air fryer | air fryer | 58 | -3.9 | 0.8 | 3.91 |
"air fryer" itself is the 100 on the demand scale. It's by far the biggest topic of the three, but it scored only 48 because interest is roughly flat year over year (+3.2%).
What the table says
- Standing desks are growing fast in this data, and "best standing desk" combines growth with buyer intent.
- Read the score and the demand together. "standing desk near me" grew +369%, but its demand is 0.03 on a scale where "air fryer" is 100. Fast growth from a tiny base is a local-SEO idea, not a product line.
- "best air fryer" is the opposite case: flat growth but large demand and strong buyer intent. It's a crowded, mature market rather than a rising one.
One level deeper: sub-niches, seasons, regions
Each niche also gets its own Google Trends rising related searches, which are usually the most specific ideas in the dataset. For the seeds:
- home gym: "smart home gym" (+650%), "adjustable bench for home gym" (+300%), and "home gym deduction act" (Breakout)
- standing desk: "walking pad" (+4,750%), "deskhaus" (+1,900%), and "flexispot e7" (Breakout)
- air fryer: "non toxic air fryer" (+2,800%), "cosori air fryer recall" (Breakout)
Seasonality, as average interest per month over 5 years with the strongest month = 100:
- air fryer: January 100, September 70.
- home gym: April 100, September and October 63.
If you sell to either, the low months are when to build and the peak is when to be ready.
Pitfalls I hit
- Breakout doesn't mean big. Several breakout searches ("shop air fryer deals", "best smart home gym") had too little data for Google to draw a reliable line. I return those free and unscored, flagged "too little search data", instead of computing huge growth figures from mostly-zero series.
- Region data can carry Google-side artifacts. In this run, Wyoming came out as the #1 US state for almost every keyword, from air fryers to standing desks. Google's own map on trends.google.com shows the same thing, so it's in Google's data, not the scraper. Compare the other states with each other and don't read much into the top spot.
- Old-year searches show up as rising. "best air fryer 2022" is listed as a breakout. Filter out year-stamped queries before you treat them as ideas.
- Trends is relative. A score ranks ideas against each other. It's not a revenue forecast, and it says nothing about competition or margins.
Try it
The Niche Value Finder takes a list of seed topics and returns one row per niche. Each row has the score and its parts (scoreParts), growth, demand on a per-seed and an all-seeds scale, buyer intent, a month-by-month seasonal profile, rising and breakout searches, top regions and a Google Trends link. You can export it to CSV, Excel or JSON. It charges per niche scored, and keywords without enough data are free.
Data source: Google Trends and Google Autocomplete. Not affiliated with Google.
Top comments (1)
The score tells you where demand is heading, but nothing in it says whether you can actually rank. "best standing desk" tops your table at 86, and that is the kind of query where page one is usually held by big review publishers and the brands themselves. The buyer intent is real; it just belongs to someone else.
I'd add one manual pass before trusting the top 10. Open the results page for each candidate and count how many of the ten results come from sites a new domain could plausibly outrank: forums, thin affiliate pages, small blogs. If the count is zero, skip it no matter what the score says.
"standing desk near me" needs its own filter too. Its results are map listings and stores, so that 0.8 intent only helps a business with a showroom. I'd tag local intent separately rather than scoring it alongside "best" queries.
That pass tends to push long tail rows like "standing desk with drawers" up the list, and that is usually where a new site wins first.