How to Vet a YouTube Creator: 7 Metrics That Predict Campaign ROI

You found three promising channels. The budget's approved. Now comes the part that decides whether the campaign returns 3x or zero: figuring out whether those subscriber counts and view numbers mean anything at all. This is how to vet YouTube influencers before you wire the money, a pre-spend, due-diligence checklist of the seven metrics that genuinely predict campaign ROI, each with a clear "healthy vs. red flag" line.

The twist that makes this guide different: for the two numbers that matter most (the views-to-subscriber ratio and the engagement rate) there is no single "good" threshold. Both swing wildly by niche. A 2% engagement rate is healthy for a comedy channel and a red flag for a finance one. So instead of generic rules of thumb, we benchmark every threshold against real per-niche medians computed from the top US channels in each category. Score a candidate against the right peer set and the bought-subs and dying-channel tells jump right out.

Already past vetting and just need a price? Jump to the sponsorship calculator. Want the full deal-structure picture first? Start with our complete guide to YouTube sponsorships in 2026. Otherwise, let's vet.

Why brands get burned (the cost of skipping vetting)

Most disappointing creator campaigns don't fail because the creative was bad. They fail because the channel never had the audience the deal assumed. A 2M-subscriber badge can sit on top of a channel that now averages 80k views; a "growing" creator can be coasting on a back-catalog while recent uploads quietly flatline; a respectable subscriber count can be padded with subscribers who never watch.

Every one of those problems is detectable before you pay, but only if you know which number to read and what counts as normal for that kind of channel. The seven metrics below are ordered the way a careful brand actually works a shortlist: start with the number that maps to delivery, then pressure-test whether that number is real, then check it's safe and durable. Treat the whole thing as a pre-spend checklist, not a rate card: pricing the creator comes after they pass.

How to vet YouTube influencers: the 7 metrics that predict campaign ROI

Here's the checklist. For each metric: what it is, the healthy-vs-red-flag line, and where to find it.

1. Real average views (not subscriber count)

What it is: the median view count of a creator's recent uploads, say the last 10–15 long-form videos. This is the single number that maps to delivery, because a brand buys views, not the follower badge.

Healthy vs. red flag: healthy = recent average views in line with (or above) the channel's historical norm. Red flag = a big subscriber count attached to recent videos pulling a small fraction of that, the classic "2M subs, 80k views" gap. Subscriber count is a lifetime accumulation; views are what's happening now.

Where to find it: scroll the Videos tab, sort by newest, and eyeball the median of the recent batch. Ignore the one viral outlier and the pinned trailer. This recent-median number is also the input you'll later drop into the sponsorship calculator to price the deal.

2. Views-to-subscriber health ratio

This is the heart of the checklist, and where the per-niche framing earns its keep. The views-to-subscriber ratio is recent average views ÷ subscriber count. It answers: of everyone who subscribed, how many actually show up for a new video?

The old rule of thumb, "about 10% of subscribers should watch each video", is worse than useless, because the healthy ratio depends enormously on the niche. We pulled the median across the top 100 US channels in each category. It ranges from over 50% in some niches to under 5% in others:

Healthy vs. red flag: healthy = at or near the niche median above. Red flag = a ratio sitting well below the niche's median: that's the gap a wave of bought or stale subscribers leaves behind.

3. Engagement rate, read per-niche

What it is: engagement rate (ER) = likes + comments ÷ views. It's how hard the audience reacts to what they watch, a proxy for whether they're paying attention or just background noise.

Healthy vs. red flag: same rule as the ratio: compare to the niche median in the table above, not a blanket "3% is good" line. That generic threshold wrongly flunks healthy big-format niches and wrongly passes weak ones. Median ER runs from roughly 5% in Spirituality & Religion and ~4.6% in Literature & Languages down to under 1% in DIY & Home and Education & Science. A finance channel at ~3.9% is right on the money; an education channel at 1% is completely normal.

There's a second axis, too: ER falls as channels get bigger. Across the top US channels, the median by subscriber tier looks like this:

4. Audience authenticity

What it is: whether the subscribers and engagement are real people or purchased padding: bought subs, engagement pods, comment bots.

Healthy vs. red flag: the most reliable tell is a combination, not a single cutoff. A views-to-subs ratio far below the niche median together with an ER far below the niche median is the signature of a bought or hollow audience, the badge says one thing, the behavior says another. Real subscribers watch and react; purchased ones inflate the count and nothing else. A single low number can have innocent explanations; the double deviation is the red flag.

Where to find it: start with the two tables above, then confirm with a third-party authenticity check (the kind covered in our guide to social media influencers). Generic comments ("Nice video!", "First!"), a follower count that jumped in steps rather than a curve, and likes-to-comments ratios wildly off the niche norm all corroborate.

5. Growth trajectory

What it is: the direction recent uploads are trending versus the channel's catalog: is the audience growing, flat, or quietly shrinking?

Healthy vs. red flag: healthy = recent videos pulling views at or above the channel's running average; subscriber growth tracking upload cadence. Red flag = recent uploads consistently under-performing the back-catalog, a coasting or dying channel living off old viral hits. A channel can show a huge lifetime subscriber count and still be in decline; you're buying its next video, not its greatest hits.

Where to find it: compare the median views of the last ~10 uploads against videos from 6–12 months ago. A clear downward step is the tell. Third-party analytics tools chart this trend line directly.

6. Brand safety & content fit

What it is: whether the creator's content, tone, and history are safe to stand next to your brand, and whether their audience actually overlaps with your customer.

Healthy vs. red flag: healthy = on-topic content, constructive comment sentiment, a clean controversy history, and past sponsors in adjacent (non-competing) categories. Red flag = recurring controversy, hostile or bot-filled comments, or a values mismatch you'd have to explain to your boss. Fit matters as much as safety: a technically clean channel whose audience never buys your category is still a miss.

Where to find it: read the comments on the last few videos, search the creator's name plus "controversy" or "apology," and scan which brands have already integrated with them. One bad video ages into a screenshot that outlives the campaign.

7. Upload consistency & audience geography

What it is: how reliably the creator publishes, and where their viewers actually live.

Healthy vs. red flag: healthy = a steady cadence (so your integration lands in an active feed) and an audience concentrated in your target market. Red flag = sporadic uploads after a long gap, or, for a US campaign, an audience that's mostly outside tier-1 markets, which quietly tanks effective reach even when raw views look healthy. Geography also moves the price: a mostly-US audience commands a premium, a mostly tier-2/3 audience a discount.

Where to find it: the upload pattern is visible on the Videos tab; audience geography comes from the creator's own analytics (ask in the media kit) or a third-party estimate.

How to spot bought subscribers and dying channels in 5 minutes

Two of the seven metrics do most of the fraud-catching work, and together they form a fast screen. Run this before anything else:

1. Pull recent average views (metric 1) and the subscriber count, and compute the views-to-subs ratio (metric 2).
2. Compute the engagement rate (metric 3), or use the engagement calculator.
3. Look up the niche row in the table above and compare both numbers to the median.
4. Flag the double deviation. Both far below niche median = likely bought/hollow audience. Recent views far below the channel's own catalog = likely dying channel.

A worked example. Say you're handed a Finance & Investing channel: 600k subscribers, recent videos averaging 18k views, ER around 1.2%. Against the finance medians (views-to-subs ~7%, ER ~3.9%), this channel's ratio is ~3% (less than half its niche norm) and its ER is roughly a third of the norm. Both numbers sit far below peers at once. That double deviation is the classic bought-subscriber pattern: a large badge, a small real audience. Pass.

Flip it: a Comedy & Entertainment channel with 5M subscribers and 1.7M average views looks like it has a "low" 1.7% engagement rate, but comedy's median ER is ~1.3% and its views-to-subs median is ~35% (this channel: ~34%). Both numbers are right on the niche line. Healthy. A sub-2% ER that would ring alarm bells in finance is perfectly normal here. Read the same number against a different peer set and the verdict flips.

From "passes the checklist" to a fair offer

Once a channel clears all seven, you've answered who's worth paying, but not what to pay. Those are two different jobs, and this guide deliberately stops at the first. For the second, take the one number you trusted in metric 1, the creator's honest recent average views, and turn it into a price.

The mechanics: a sponsorship base price is roughly niche CPM × average views ÷ 1,000, then adjusted for format, usage rights, exclusivity and audience geography. Plug your vetted creator's real views into the sponsorship calculator for a tailored number, or check the raw niche rate on the CPM calculator. For the full breakdown of what each niche pays and how deals are structured, see the complete YouTube sponsorships guide, and for the size ladder, how much sponsors pay YouTubers.

Vet at scale, not one tab at a time

Working a shortlist of three channels by hand is fine. Working fifty isn't, and that's where the checklist needs a faster surface, like the Channel Crawler. To find candidate channels by niche and country in the first place, start at find YouTube channels; it's the discovery step that feeds this whole vetting process. The per-niche peer sets you're benchmarking against live on top YouTube channels, broken out by country and category, so you can eyeball where a candidate sits among real leaders.

And if you've been quoted a five-figure price for an enterprise vetting suite, know that the seven metrics here cover the genuinely predictive signals: you don't need a heavyweight platform to apply them. See how the lighter-weight approach compares on our alternatives page.

FAQ