What budget do I need to validate a niche this way?

Zero, if all you want is to validate that a market exists. Meta Ad Library is free, TikTok Creative Center is free, and the longevity reading takes about 20 minutes per niche. The budget question only matters when you decide to actually test the product yourself — that is a separate 500 to 1500 euro decision, not a niche-selection decision.

What is a saturated niche, and how do I recognise one?

Saturation looks like 8 or more advertisers all running near-identical hook structures, identical offers, and identical CTAs in the same week. The pattern overlap is the signal — when every winning ad reads like a copy of the previous winning ad, the niche has commoditised and entry costs have spiked. You can still enter; you just need a structurally different angle (price, product variant, distribution channel).

Why focus on advertising volume instead of search demand?

Search demand is what consumers say they want. Advertising volume is what operators are spending money to validate right now. The gap between the two is where most picking errors happen. Search-volume-led picks often land in saturated paid-acquisition markets (every keyword has a real bidder). Advertising-volume-led picks tell you where money is currently being made — which is where you want to be, even if the language used in the ads does not match the search terms.

Can a niche have low advertising volume and still be a good pick?

Sometimes. The two cases where low ad volume is positive are: (1) the niche is new enough that operators have not yet discovered it through paid acquisition, and (2) the dominant channel for the niche is organic — content, SEO, or word-of-mouth — and ads are a small slice of the customer acquisition story. Most low-volume niches do not fit either case and are simply markets that do not work.

How long does this niche-validation read take to be predictive?

The descriptive read takes 20 minutes per niche, weekly. The predictive value compounds over 6 to 8 weeks — by then you have seen which categories are still being advertised, which have died, and which have grown. You should not act on a single week's read. You should pick a niche after watching it for at least one month.

ecommerce · 7 min read

What to sell online — how competitive intelligence replaces gut feeling

Last updated: May 2026

Fast answer

Look at what the top 10 advertisers in a niche have been running ads for in the past 14 days. If 5 or more of them are advertising the same product subcategory with active ads of 30 or more days, the subcategory is validated. If only 1 or 2 are advertising it, you are either early or in a dead niche — the difference is whether the leaders run ads on adjacent products at all. The 14 day window plus the 30+ day longevity filter eliminates 80 percent of the noise.

The problem with gut-pick

Most operators picking a niche do one of three things. They pick something they personally use (skincare for a self-care enthusiast, supplements for a gym-goer). They pick something a YouTube video told them was hot. Or they pick something a TikTok dropshipping account told them was hot. All three approaches fail for the same reason: they replace evidence with affinity.

The skincare enthusiast picks skincare because they feel they understand the category. They do not. Understanding a category as a CONSUMER is structurally different from understanding it as an OPERATOR — the consumer sees the products, the marketing, the unboxing experience. The operator sees the margins, the return rates, the ad-spend efficiency, the regulatory frictions. These are not correlated.

The YouTube video has a publishing bias — it surfaces the niches that produce content-worthy stories, which is not the same as the niches that produce profit. The dropshipping account has a deeper bias — it surfaces the niches whose creative templates can be made to look successful in a 30-second video, which is even further from the niches that produce profit.

The competitive-intelligence framing replaces all three with one question: what is the SECOND-WEEK retention of an advertising spend in this category? If multiple sophisticated advertisers are spending in week 4 of the same campaign on the same product type, the category has been validated by the people whose job is to validate categories. That is a higher-fidelity signal than your own gut, the YouTube channel, or the dropshipping account.

The three signals — and how to read them in 20 minutes

Signal 1 — Advertiser concentration

Open the Meta Ad Library (set country to European Union). Search for the niche keyword (e.g., "skincare", "fitness supplements", "home decor"). Look at the top 20 results sorted by recency.

Count how many DISTINCT advertisers appear in the top 20. The pattern:

15 or more distinct advertisers = healthy concentration. Multiple operators are competing on the same surface. Market is alive.
8 to 14 distinct advertisers = consolidated. A few brands dominate. Entry is harder but the value pool is larger per winner.
Fewer than 8 distinct advertisers = either dead or dominated by one or two players. Dangerous to enter unless you have a structural angle the dominant players cannot match.

Signal 2 — Advertising longevity

For each of the top 5 advertisers, open their library page directly. Read the "Active period" line below each ad — Meta shows the date the ad first ran and (if active) the implication that it is still running. Any ad with 30 or more days of active period is high-fidelity validation: Meta does not let underperforming ads run that long.

Tally: how many of the top 5 advertisers have at least one ad that has been active 30+ days?

4 or 5 of 5 = strongly validated niche. Operators are spending sustainably.
2 or 3 of 5 = mixed. Some operators are succeeding; others are still testing.
0 or 1 of 5 = the niche is in turnover. Either operators are testing many short-lived creatives (which can be a healthy launch phase) or operators are unable to sustain spend (a dying niche). Tell the difference by looking at how new each advertiser's library is — new libraries means new operators entering; old libraries with short-lived ads means established operators struggling.

Signal 3 — Product subcategory clustering

This is the one most beginners miss. Look at the product CATEGORY each advertiser is selling within the niche. Skincare is a niche; "vitamin C serums" is a subcategory; "hydroquinone-free brightening serums" is a sub-subcategory. The narrower the subcategory cluster across multiple advertisers, the more validated that specific subcategory is.

If you see 4 of 5 top skincare advertisers all running vitamin C serum ads in the same week, vitamin C serums are validated as a subcategory. If you see 5 of 5 running different subcategories, the broader niche is healthy but no specific subcategory is dominantly validated — you are picking among healthy options rather than between a winner and a loser.

Where to look — the platforms ranked by signal quality

Meta Ad Library (free, EU-complete). Highest-fidelity surface. Captions, creative, active period, advertiser identity, all public. The TIME COST is the only friction.
TikTok Creative Center (free, partial coverage). Useful for video format trends and engagement-proxy data. Lower fidelity for hook-level intelligence because TikTok hides ad captions for most regions.
Google Ads Transparency Center (free, search-ad-focused). Useful if you are picking a niche where search-intent ads dominate the customer journey. Lower value for e-commerce-led niches.
Paid ad-intelligence tools like Pipiads or Adspy. Useful when you are picking outside the European Union and the free Meta library is incomplete. Cost 99 to 299 euros monthly.
CommonWealth Ops weekly digest. What we built. Reads the whole niche across all the surfaces above and synthesises the patterns weekly. EUR 49/month plus 20% of net profit.

What the KobiiSpy data tells us about niche selection right now

Our ingestion pipeline currently captures the highest volume of signals in two niches: fitness and skincare. Both niches show what we would call HEALTHY concentration — multiple distinct advertisers, multiple distinct product subcategories within each, and persistent ad longevity.

We do NOT currently have meaningful data on supplements, home goods, or apparel. That does not mean these niches are bad picks — it means our crawl coverage has gaps and we cannot honestly tell you. We are adding coverage; you should not assume "low captured volume" maps to "low niche viability" until our coverage matures.

If you are looking at a niche we do not yet cover, the manual workflow described above is the workflow. The CW Ops weekly digest is the automated version of it, for the niches we DO cover.

What this method does NOT solve

It does not pick a product within the subcategory for you. Even if you decide vitamin C serums are validated, you still need to choose the specific formulation, price point, and positioning. The competitive-intelligence read tells you a subcategory works; the product decision is yours.

It does not tell you the MARGIN structure. Two operators can both be selling vitamin C serum profitably with very different cost structures — one importing at 4 euros and selling at 30, another manufacturing at 12 and selling at 60. The library shows you their advertising; it does not show you their P&L.

It does not replace a small-spend test. Even the most-validated niche will still produce some losing product picks. The method gets you to the right SUBCATEGORY; you still need to spend 200 to 500 euros testing your specific product variant before scaling.

How CW Ops compounds this

The 20-minute weekly read is the manual baseline. Once you have a niche picked, the CW Ops weekly digest tells you which HOOK structures, EMOTIONAL TRIGGERS, and CTA patterns are dominating that niche each week. You do the niche selection once. You watch the weekly read every week to keep your creative aligned with what is working RIGHT NOW.

Both views compound. The niche-selection read picks where to play. The weekly digest tells you how to play this week. Skip either and you are operating on a stale model.

Pricing: EUR 49/month plus 20% of net profit (a single rate for every operator, no threshold, and EUR 0 share in any month you do not profit). That is the alignment by design.

Frequently asked questions

What budget do I need to validate a niche this way?: Zero, if all you want is to validate that a market exists. Meta Ad Library is free, TikTok Creative Center is free, and the longevity reading takes about 20 minutes per niche. The budget question only matters when you decide to actually test the product yourself — that is a separate 500 to 1500 euro decision, not a niche-selection decision.
What is a saturated niche, and how do I recognise one?: Saturation looks like 8 or more advertisers all running near-identical hook structures, identical offers, and identical CTAs in the same week. The pattern overlap is the signal — when every winning ad reads like a copy of the previous winning ad, the niche has commoditised and entry costs have spiked. You can still enter; you just need a structurally different angle (price, product variant, distribution channel).
Why focus on advertising volume instead of search demand?: Search demand is what consumers say they want. Advertising volume is what operators are spending money to validate right now. The gap between the two is where most picking errors happen. Search-volume-led picks often land in saturated paid-acquisition markets (every keyword has a real bidder). Advertising-volume-led picks tell you where money is currently being made — which is where you want to be, even if the language used in the ads does not match the search terms.
Can a niche have low advertising volume and still be a good pick?: Sometimes. The two cases where low ad volume is positive are: (1) the niche is new enough that operators have not yet discovered it through paid acquisition, and (2) the dominant channel for the niche is organic — content, SEO, or word-of-mouth — and ads are a small slice of the customer acquisition story. Most low-volume niches do not fit either case and are simply markets that do not work.
How long does this niche-validation read take to be predictive?: The descriptive read takes 20 minutes per niche, weekly. The predictive value compounds over 6 to 8 weeks — by then you have seen which categories are still being advertised, which have died, and which have grown. You should not act on a single week's read. You should pick a niche after watching it for at least one month.

Become an operator

Stop guessing what to sell.

CommonWealth Ops turns your market's competitor activity into ranked, data-backed intelligence — and protects your capital before you spend a euro on ads. EUR 49/mo + 20% of net profit. No free trial: skin in the game both ways.

Join the waitlist

See pricing Real and aspirational stories

← All posts