How to tell if a tech channel is worth sponsoring
Subscriber count will not tell you whether a sponsorship works. Here is what to check on a tech or AI channel before you pay for a placement.
Subscriber count is the first number everyone looks at and close to the least useful one. A channel with 400,000 subscribers from a viral video three years ago can pull fewer views today than one with 60,000 that posts every week to an audience that waits for it. You are not buying the subscriber number. You are buying the views a video will get and the chance that the people watching want your product.
Here is the order we go through before we put a channel in front of a client.
Average views, not peak views
Open the channel’s video list, sort by newest, and look at the last ten to fifteen uploads. Ignore the biggest one and ignore the smallest one. What the rest cluster around is roughly what your video will do.
Two things to watch for. First, a single outlier that drags the average up: one video at two million and nine at forty thousand means you are buying forty thousand. Second, recent uploads that are clearly below the older ones. A channel on the way down is cheaper for a reason, and the rate card usually has not caught up yet.
Whether the numbers hold over months
One good month is not a pattern. Scroll back half a year. You want views that stay in the same band, or climb. What you do not want is a channel that spiked, got its rate card rewritten at the top, and settled back down.
This is also where you catch a channel that changed topic. A hardware reviewer who pivoted to commentary still has the old subscribers, but the audience that shows up is a different audience, and it may not be yours.
Engagement in proportion to the views
Comments and likes matter less as an absolute number than as a ratio to views. A video with 80,000 views and 40 comments is being watched by people who are not especially invested. The same views with 600 comments is a channel whose audience treats it as something to participate in, and those are the viewers who act on a recommendation.
Read the comments themselves, not just the count. On a channel worth sponsoring the comments are about the subject, people argue with the host, and they ask follow-up questions. If the comments are all emoji and one-word praise, that is worth knowing before you commit.
Whether the audience is your audience
This is the part that is easy to skip and it decides most of the outcome. A large AI channel is not automatically right for a developer tool. “AI” covers people building with APIs, people following the industry as news, and people watching image generators for entertainment. Those three groups overlap less than the label suggests.
The useful question is not whether the channel is in your category. It is whether the people watching are the people who would actually buy. Look at which of the channel’s videos do best. That tells you what the audience comes for. If the best performing videos are tutorials and your product is a tool that needs setting up, that is a fit. If the best performing videos are reaction content, the same subscriber count buys you much less.
Geography matters too, for anything sold in one market. YouTube Studio shows the creator where their viewers are, and a creator who is used to working with brands will share that without being asked.
How past sponsorships were handled
Watch two or three previous sponsor segments in full. You learn more from this than from any metric.
Does the creator sound like they used the thing, or like they are reading a script you could have sent anyone? Is the segment at the start, where it is seen, or at minute fourteen? Did the comments push back on the sponsor, and did the creator engage or ignore it? A creator who takes everything, in any category, has an audience that has learned to tune the segments out.
Numbers that do not mean much on their own
- Subscriber count. Covered above. Useful only next to average views.
- Total channel views. Accumulated over years, including from an audience that has moved on.
- A media kit screenshot. Not wrong, usually, but selected. The channel’s own video list is the primary source and it is public.
- Follower counts on other platforms. A large X following does not move YouTube views.
What this is for
None of this requires a tool. It requires half an hour per channel and knowing which half hour to spend. The reason agencies exist is not that the information is hidden. It is that doing this across sixty channels to find the four worth approaching, then negotiating with those four, is a week of work that most in-house teams do not have spare.
If you would rather hand that part over, that is what we do.