October 16, 2026

Found a YouTube channel that works for your brand? Find more channels like it with your AI agent

One channel that worked is the best brief you will ever write for your next YouTube buy. Here is how to turn it into a shortlist of look-alike channels with your AI agent, filter out the ones a sponsor can't use, and vet the rest on YouTube, with a real run.

> find channels whose audience looks like Veritasium's # 20 look-alikes, 6 left after the sponsor filter Joe Scott US 57% The Action Lab track record: strong Steve Mould US 37%

The hardest part of buying YouTube sponsorships is finding the next channel. When one channel has already worked for you, you have something better than any category search: a brief written by your own results. The audience showed up, the read landed, and you would book it again. The job now is to find more channels whose audience looks like that one.

This post shows how to do it with your AI agent, using a real run from October 16, 2026. We used Veritasium as the starting channel, the way a science, education or tech brand might after a good read there.

Step 1: Start from the channel that worked

Pick the channel you would renew without thinking. If you have several, start with the one closest to your product, because look-alikes inherit its audience and its topics. If you have no winner of your own yet, borrow one: the channels a similar brand keeps renewing are a good proxy. We wrote that walkthrough here: Which YouTube creators does your competitor sponsor?

Step 2: Ask your agent for look-alikes

With the ThoughtLeaders CLI connected, ask in plain words:

Find YouTube channels whose audience looks like Veritasium’s.

Your agent resolves the channel, then runs:

tl channels find "Veritasium"
tl channels look-alike 1784 --limit 20

The look-alike search matches on audience and topics, then re-ranks by things a buyer cares about, such as keyword overlap, demographics, posting rhythm and video length. It is a different tool from a generic “similar channels” list, and it is the same step ThoughtLeaders uses when it matches brands to channels.

In this run it returned 20 channels. The top of the list was what you would hope for: Joe Scott, Professor Dave Explains, Dr Ben Miles, The Action Lab and Steve Mould, all science and explainer channels with big audiences.

Step 3: Filter out the channels a sponsor can’t use

A raw look-alike list is a starting point, not a shortlist. Some channels look right and still won’t work for a sponsor. Ask your agent to apply a buyer’s filter:

Keep only channels with at least 100,000 projected views per long-form video, a sponsorship track record of 5 or more, and at least 35% of the audience in the US. Tell me why you dropped the rest.

Six of the 20 passed:

ChannelSubscribersProjected views per long-form videoUS audienceSponsorship track record
Steve Mould3.62M2,445,93137%Middle
Hafu Go19.2M1,955,70041%Middle
Dr Ben Miles2.62M1,327,16537%Middle
The Action Lab5.17M842,53735%Strong
Joe Scott2.56M660,59457%Strong
Sciencephile the AI1.11M229,55242%Middle

The other 14 dropped out, several for more than one reason, and none of it shows up in a list of names:

  • Too few views for the audience size. Professor Dave Explains has 4.47M subscribers but projected about 39,000 views per new long-form video. One channel with 845K subscribers projected 316. Four more had no projection at all because they have too few recent long-form uploads.
  • No audience data on file. Eleven channels had no country breakdown in ThoughtLeaders. That doesn’t make them bad channels: SmarterEveryDay dropped for this reason alone. It means you should ask the creator for a YouTube Studio screenshot before you go further.
  • Audience in the wrong place. Two channels had 32% and 20% of their audience in the US, which matters if that is where you sell.
  • Little sponsor history. Nine channels scored under 5 on sponsorship track record, which means few brands sponsored them or came back in the last two years.

How to read the numbers:

  • Projected views are what a new long-form video is expected to get by day 30, built from the channel’s recent uploads with outliers capped. They are not the subscriber count and not the lifetime average.
  • YouTube changed how it counts public views on August 24, 2026. Long-form videos now show about 1.2 to 1.5 times the views they would have shown before in the first week, and projections are drifting toward the new count as newer videos age in. Compare channels to each other on the same date, and count delivery on engaged views from the creator’s YouTube Studio.
  • Sponsorship track record is the channel’s ThoughtLeaders sponsorship score in bands. It rewards how many different brands sponsored the channel over two years and how many came back after a break of at least two months. Under 5 means little track record, 5 to 7.5 is middle and 7.5 or more is strong.
  • US audience is the US share of the channel’s audience, from its country breakdown where we have one.

All figures are from ThoughtLeaders on October 16, 2026. They move as channels publish, so rerun the search before you buy.

Step 4: Vet the shortlist on YouTube

Open each channel with the ThoughtLeaders Chrome extension. On the channel page you get:

  • Top sponsors. Which brands the channel has mentioned, whether each mention was sponsored or organic, and its renewal rate. Detection isn’t exhaustive, so a short list is a reason to look closer, not a verdict.
  • Views performance. The best and worst months over the last year and the 30-day projection, so you can see whether recent videos are growing or fading.
  • Audience demographics. Where the viewers are and who they are, where we have the data.
  • Similar channels. More look-alikes, right on YouTube, so the shortlist keeps growing while you browse.

This is where you catch what no list can show: a channel whose sponsors all stopped months ago, a run of videos far off the channel’s usual topic, or views that hang on one viral hit.

Step 5: Price it before you negotiate

When the creators send quotes, check each one against the channel’s projected views and a fair-value estimate, not just the asking price. Full walkthrough: How to check a YouTube sponsorship quote with your AI assistant.

A category search finds channels that talk about the same things. A look-alike search starts from an audience you already know buys from you, and finds more of it. That is the step after “find the winners” in the method ThoughtLeaders has used to buy YouTube sponsorships since 2018: find what works, expand to look-alikes, then decide on value, not price. Your AI agent now runs the search and the filter in a few minutes. The judgment calls stay with you.

Try it with your own winner

Install the CLI and ask your agent for look-alikes of the last channel you’d renew. Or add the Chrome extension and open the Similar channels panel on any YouTube channel.