A skincare client arrives with a budget and a platform already chosen. They post on Instagram, so the campaign is Instagram. Nobody in the room actually decided that: it was inherited from whichever channel their social manager was handed two years ago, and it is now quietly setting the reach model, the shelf life and the cost per asset for a campaign it was never meant to shape.
The client's own channel tells you where their existing followers are. It tells you almost nothing about where the people they want to reach can be found, or about what a given budget buys on each platform. Those are different questions, and the brief should answer them separately.
Distribution decides everything downstream
The three platforms hand a post to an audience in different ways, and almost every other difference follows from that.
Instagram still leans substantially on the follower graph: a creator's followers are the base audience, with recommendation layered on top of Reels. TikTok distributes primarily by recommendation, with the follower graph acting more as a signal than as a delivery list. YouTube runs on recommendation as well, and adds a third mechanism the other two barely have: search, with intent behind it.
The consequence lands on the number every brief is written in. On a follower-graph platform, follower count is a reasonable proxy for reach. On a recommendation platform it is a weak one: a creator with 20,000 followers can out-reach one with 200,000 on any given post, and the same creator's own posts can vary by an order of magnitude between one week and the next. Buying a TikTok creator on follower count is buying the wrong variable.
The comparison that matters
| TikTok | YouTube | ||
|---|---|---|---|
| Primary distribution | Follower graph, recommendation on Reels | Recommendation | Recommendation plus search |
| Follower count as a reach proxy | Reasonable | Weak | Weak for reach, useful as a floor |
| Content half-life | A day for Stories, days for feed | Days, occasionally far longer | Months to years |
| Production cost per asset | Low to moderate | Low | High |
| Visible from outside | Likes, comments, Reels views | Views, likes, comments | Views, likes, comments |
| Suits | Brand association, product presence, fast turnaround | Reach, demonstration, trend participation | Consideration, explanation, durable search demand |
Platform mechanics change, sometimes quietly and without an announcement. Re-check the top two rows against how content in your niche is actually behaving before you commit a budget to them.
Half-life is the axis nobody prices
A Story is gone in a day. A feed post does most of its work in the first few days and then flattens. A YouTube video that answers a question a person types into a search box keeps returning results for months, and if it holds its position it keeps working long after the campaign has been reported on and closed.
That changes the arithmetic. Cost per asset is the wrong denominator on YouTube; cost per month of useful life is closer to right. A single video that looks expensive next to three Reels sometimes is not, and a Story package that looks cheap per unit is buying a day of visibility per unit.
The reverse case is just as real. If the campaign is tied to a launch date, a promotion window or a seasonal moment, durability is worth nothing and you are paying a production premium for shelf life the brief cannot use.
The same creator is three different propositions
A cross-platform creator is usually strong on one platform and merely present on the others. The media kit shows all three together, aggregated into one impressive block of numbers, and the platform being sold to you is frequently the weakest of the three: because it is the one with capacity, and the one whose figures benefit most from being read next to the strong one.
Evaluate each platform on its own terms. Engagement against that platform's peer group, posting consistency on that platform, and whether the audience actually followed the creator across or was merely told about it. An engagement rate only means something against the right peer group, and the peer group changes with the platform as much as it changes with the niche.
Audience composition splits too. The same person can have a pan-regional TikTok audience and a domestic Instagram one, because the recommendation system pushed one across borders and the follower graph did not. In a market with a large diaspora, that gap is often the entire decision, and it means the answer to "is this creator right for us" can genuinely be yes on one platform and no on another.
Choosing from the brief
A short sequence, run before anyone opens a discovery tool.
- Name the objective as a verb. Reach people, explain something, be seen alongside a brand, drive a purchase this month. The vague version of this step is what lets the client's existing channel win by default.
- Ask how long the asset needs to keep working. A launch window and an always-on consideration play answer this differently, and the answer largely picks the platform on its own.
- Ask what has to be shown. Anything that needs demonstration, comparison or explanation is fighting the format on Stories. Anything that needs only association is over-produced as a video essay.
- Then check where the audience for this niche actually is, in this market. Not where the platform is popular in general. The wider process this sits inside is here.
If steps 1 to 3 point one way and step 4 points another, step 4 wins. A brilliant format on a platform where your audience is not present is a well-made miss.
What this comparison does not settle
It narrows a decision. It does not make one, and there are limits worth naming before you take it into a client meeting.
- Platform mechanics move. Every row in that table describes behaviour that has changed before and will change again. Treat it as a current reading, not a constant.
- Aggregate platform behaviour is not your niche's behaviour. Some niches sustain long-form demand on YouTube that the platform-level picture would not predict, and some collapse entirely into short form. Check the niche, in the market, before generalising.
- What you can measure from outside is not what matters most. Saves, shares, watch time, click-through and completion rate are usually the better signals, and they stay in the creator's own analytics. You are choosing partly on the metrics that happen to be public, which is a real bias in this comparison and in every tool that runs on public data, ours included.
- Cross-posting is not multi-platform. A creator who exports the same vertical video to three places has one audience and three URLs, and should be priced as one.
- The platform never rescues the wrong creator. Format decisions matter less than fit, and they matter a great deal less than whether the audience is the one the brief asked for.
Lyren covers YouTube, Instagram and TikTok, and takes platforms as an explicit input alongside niche and audience market rather than treating a creator as a single cross-platform entity. Scoring happens within the niche and per platform, so a creator's weakest channel is not flattered by their strongest. Access is invite-only from the waitlist.