Vetting a YouTube channel before you pay for a sponsorship
Most vetting is too short and too manual. In the EMARKETER and Viral Nation survey of 117 US marketers in September 2025, over half spend 30 minutes or less checking a creator, 81.2 percent rely on manual review and only 9.1 percent call their process scalable; fake or bot followers were 56.5 percent of the fraud issues brands reported to Influencer Marketing Hub's 2026 benchmark. The public data can answer more of this than most teams use.
Discover puts the public checks on one row: uploads in the last 30 days and the last upload date (is the channel alive), the sponsor list with dates (is it experienced, and with whom), the Commercial score (is it set up for brand work) and the Audience score (does it answer its viewers), with the full report behind each. What it cannot do is audit the audience itself: demographics, watch time and retention are in the creator's YouTube Studio only, and Discover has no fake-follower detector. The checklist below says which is which.
How do I vet a YouTube creator before a sponsorship?
Seven public checks, in order: is the channel active (uploads in 30 days, last upload); do views on the last 10 long videos make sense against subscribers; does it already run sponsorships and with whom, and when; does it disclose them (paid-promotion flag plus description text); what are the recent video titles and topics (brand safety); does it answer its viewers (Audience score); is a business contact listed. Discover's row and report cover all seven. Then ask the creator for demographics and watch time from Studio. Our 15-minute vetting checklist post walks through three real channels.
How do I tell if a YouTuber has fake subscribers?
From public data, by ratios and shapes, not by a detector: median views on the last 10 long videos divided by subscribers (a 200,000-subscriber channel averaging 1,500 views is a warning), a subscriber jump with no video that explains it, comment sections full of generic praise, and likes out of proportion to views. Discover shows the subscribers, pace and recent videos with view counts in the report; it does not score authenticity. Paid audit tools estimate it from engagement patterns, with reviewers reporting both misses and false alarms.
What is a good views-to-subscribers ratio?
Guides quote 10 to 20 percent of subscribers as a healthy median view count for an established channel, lower for very large channels and higher for small, fast-growing ones. It is a rough gauge; what matters more is consistency across the last 10 videos. Take the median from the report's video list rather than the average, so one viral video does not hide nine quiet ones.
How do I check if a YouTube channel is still active?
The last-upload column and the uploads-in-30-days count on Discover, counted on long-form videos. Filter by at least 1, 2, 4 or 8 uploads in 30 days to exclude dormant channels. Rows marked not analyzed yet have no pace until someone opens the report; the report also lists the newest videos with dates.
How often should a YouTuber post to be worth sponsoring?
Often enough that your video is not the only one that month. For an integration, a channel posting at least two long videos a month keeps its audience warm; weekly is the usual bar for mid-size channels. vidIQ's upload-frequency study found more frequent uploaders grow faster on average, but a channel posting daily may also bury your integration faster. Use the pace filter and read the view consistency.
What is a good engagement rate on YouTube?
The tools disagree on the formula (likes plus comments over views, or over subscribers), so a number without its formula is not comparable. Likes at 3 to 5 percent of views and comments at 0.2 to 0.5 percent of views are common ranges quoted for mid-size channels. Discover's Audience score measures something narrower and verifiable: whether the channel replies to its viewers and how fast, from a sample of recent comment threads.
What does the Audience score tell me as a brand?
Whether the creator treats comments as a conversation. It is built from the reply rate on the newest threads of the newest long videos, the median response time, and whether buying questions from viewers got an answer. A creator who answers viewers will answer the questions your integration raises; measured on eight tech channels on 31 October 2026, reply rates ran from 0 to 100 percent with a median of 15.
What does the Commercial score tell me?
How set up for brand work the channel already is: disclosed sponsors and their recency, affiliate links, the creator's own products, and a listed business contact, on a 0 to 100 scale with the rules on the report. High means experienced and reachable; low means a first deal with explaining to do. It is not a quality score and says nothing about the audience.
How do I check brand safety on a YouTube channel?
Read the last 10 to 20 video titles and topics in the report, skim the comment tone on two recent videos, check the sponsor list for categories you cannot sit next to, and look for made-for-kids or age-restricted flags, which the report shows from YouTube's own data. Then a 10-minute watch of one recent video. Viral Nation estimates that 30 minutes of vetting covers about 0.01 percent of a channel's history, so prioritise the newest.
Can Discover detect bought views?
No. It shows the public view counts of recent videos so you can see the shape: a sudden spike on one video with no comment increase, or a flat 10,000 views on every upload regardless of topic, is what bought views look like. There is no reliable public-data detector; treat vendors' fraud scores as estimates.
What should I ask the creator for that public data cannot show?
Dated Studio screenshots of the audience tab (age, gender, top countries), average view duration on the last 10 videos, and traffic sources. Label them creator-reported in your file. Ask for the last three sponsors and compare with the report's sponsor list; a mismatch is a conversation, not necessarily a problem.
How do I check a creator's past sponsorships for competitor conflicts?
The report lists every sponsored video found with its brand and date. A direct competitor in the last 90 days means exclusivity is unavailable or expensive and the creator may be bound by a clause. Only 22.2 percent of marketers check this (EMARKETER and Viral Nation, 2025); it takes 20 seconds on Discover.
Does a channel's country mean its audience is there?
No. The country column is the creator's declared channel country as YouTube reports it. A US-registered channel can have a mostly Indian audience and vice versa. Audience geography is in the creator's Studio; ask for the screenshot, and look at the language of comments on recent videos as a cheap proxy.
How do I spot a channel that buys engagement with giveaways?
Titles and descriptions with recurring giveaways, comment sections dominated by "done" and "entered", and subscriber growth that outruns view growth. The report's recent video list and the comment sample used for the Audience score make this visible; the score itself is not fooled by giveaway comments because it measures the creator's replies, not the viewers' volume.
Should I vet a creator with 5,000 subscribers the same way?
Yes, faster. Small channels fail vetting on activity (last upload months ago) and on views (200 views on 5,000 subscribers) more often than on fraud. The same seven checks apply; the sponsor list is usually empty, which is normal, and a listed business contact is a good sign that the creator is open to deals.
Can I vet a whole shortlist at once?
Partly. The search table already shows pace, last upload, sponsors, scores and the contact flag for every analyzed row, so a 25-row shortlist is one screen. Bulk paste (Agency plans and trial) loads a list of handles; sort by Commercial readiness or recency; export the CSV and keep the rows that pass. The per-channel report is the second pass on the survivors.
How current are the numbers I am vetting on?
Each row carries its refresh date; a report is reused for a day and refreshable after that; rows without a report for 30 days are deleted. Vet on a report you refreshed that week, and take a dated screenshot of it for the client file.
What does "not analyzed yet" mean for vetting?
That the channel came from a keyword search and only its name, size, country and topics are known. Click Analyze on the row; about 18 API units and a minute later the row has pace, sponsors, scores and the contact flag. Never vet on a stub.
Does Discover flag made-for-kids or age-restricted channels?
The report shows YouTube's own flags across the scanned videos (made-for-kids count, age-restricted count). Made-for-kids videos have comments disabled and limited ad formats, which changes what an integration can do; check before you brief.
What does skipping vetting cost?
A placement nobody sees, a disclosure problem that lands on the advertiser, or a creator who posts once a quarter after you paid. Influencer Marketing Hub's 2025 survey found 36 percent of brands had a campaign dispute in the prior year. Twenty minutes with the row, the report and a Studio screenshot prevents most of them.
Can I get a vetting checklist I can reuse?
Yes: the vetting checklist post has the seven public checks with what to look for and three worked examples, and the analyzer report is the tool that answers them. Keep the completed checklist with the date in the client file; 96.6 percent of marketers want vetting documentation and 25.6 percent always get it.
Before you pay a creator, what do you actually check?
Four things that are not the follower count: the median views on the last 10 to 15 uploads, the comment quality on two recent videos, where the audience is (a Studio screenshot, since the channel's country is not the audience's), and whether the creator already works with brands and with whom. Discover's row and report answer the first and the fourth from public data; the second takes five minutes on YouTube; the third you must ask for.
What engagement rate is a red flag for bought followers on a 20,000 to 50,000 account?
Brand-side rules of thumb put under 2 percent engagement (likes plus comments over views) on an account that size as a flag, and the agreed method is to pull the last 10 to 15 uploads, not the best one. A YouTube-specific tell is a flat view count on every video regardless of topic. The report's video list makes both checks a glance.
Is it worth sponsoring a small YouTuber with 1,000 to 5,000 subscribers?
Yes when the views are real and the niche is specific: a 4,000-subscriber channel on a tool your customers use, averaging 1,500 views, posts weekly and lists a contact, is cheap, reachable and untested by competitors. The checks are the same as for a big channel, with activity and view consistency doing most of the work.
How do I know if a YouTuber is open to sponsorships?
Two public signals: a listed business contact (the "Business contact listed" column on Discover) and disclosed sponsorships on recent videos (the sponsor chips). Both present means experienced and reachable; contact listed with "none found" means open but not yet pitched; neither means you are guessing. Creators also say so in their About text, which the report shows as published.
How can I tell if a YouTube channel is part of an MCN or managed by an agency?
The About page and video descriptions usually name the management or give a business address at an agency's domain; the report shows those lines as the creator wrote them. A managed channel answers from the manager's inbox and quotes from a rate card; budget for that in the first email.
Before an AI-generated follower count goes into your shortlist, where did that number come from?
If you cannot answer, it does not go in. Every number on Discover is the YouTube Data API's public figure with the date it was fetched, shown on the row and the report, and the scores carry their rules. A shortlist from a chat assistant should be re-run against a dated source before a client sees it.
Who actually has influence over our target buyers, versus who just has a big following?
In B2B the answer is usually the smaller channel whose comments are from practitioners. Read the comment sample behind the Audience score on the report: a channel where viewers ask implementation questions and get answers is influencing buyers; a channel with a large count and generic praise is not. Sponsor chips from vendors in your category confirm that someone already tested it.
Now run the search
Discover shows the first 10 channels of any search with size, posting pace, sponsors, scores and whether a business contact is listed. When the inventory has nothing for your keyword, it searches YouTube for you.
Other jobs
- Finding YouTube channels to sponsor: by niche, size, country and posting pace42 questions
- Seeing which brands sponsor a YouTube channel, and tracking competitors24 questions
- Contacting YouTube creators: business emails, response rates and what works30 questions
- What a YouTube sponsorship costs, and how to price one without a rate database32 questions
- Disclosure rules, briefs and contracts for a sponsored YouTube video31 questions
- What a client report can show from public YouTube data, and what it cannot24 questions
- How Discover compares with the influencer tools agencies pay for34 questions
- Where the data comes from, how fresh it is, and why searches are rationed30 questions
- For creators: which brands sponsor channels like yours, and how to reach them27 questions