Industry
Part of Benefits of social media, in six rules: how to test any claim
How three kinds of adoption reshape a social media format
Social media updates from outside a format: three kinds of adoption, what a built-in button does to variance, and how to test the burnout claim.
A brand account uses a joke format, and the format changes for everyone who sees it. That includes people who were using it an hour earlier and changed nothing about how they use it.
The change is not in the joke. It is in what using the joke says about you.
What to take away
- Adoption often produces a visible reactionthe format becomes embarrassing, and using it marks you as behind.
- Featurization trades variation for participationa template or a sound page widens who can join and narrows what they can make.
Three kinds of adoption, with different effects
The three run in the order that Everett Rogers set out for products in Diffusion of Innovations (1962), and diffusion of innovations works no differently for a format.
Commercial. A brand account posts in the format. The immediate effect is not that the joke is ruined. It is that the format has been publicly certified as understood by marketing departments, so it is no longer evidence that you are close to where things start.
Product. The platform itself turns the format into a feature: a template, a one-tap effect, a reusable audio clip, a button. This is the most consequential of the three and the least discussed.
Institutional. Organizations that need to appear current adopt the format, usually late and usually slightly wrong. The wrongness is informative: it marks the point at which the format has become legible to people who learned it secondhand.
What happens when a platform builds the format into a button
Two things at once, pulling in opposite directions.
Featurization: two opposite effects
Widening
- Who can make it
- Anyone, no skill
- Skill needed
- None
- Variance
- Reaches new audiences
- Authorship
- New contributors
- Surprise
- Format spreads widely
- Portability
- Wider participation
Collapse
- Who can make it
- Only tool users
- Skill needed
- Selection only
- Variance
- Converges, same start
- Authorship
- Act of selection
- Surprise
- Harder to produce
- Portability
- Bound to one service
Participation widens enormously. Making a version stops requiring editing skill, software or time. People who could never have contributed now can, and the format reaches audiences that would never have made anything by hand.
The built-in tools show how little is left to decide. Instagram's Add Yours sticker gives a prompt a slot. TikTok's Stitch button turns any clip into a response. CapCut templates fix the cut, the text placement and the timing, and a TikTok sound page collects every video that used the same audio.
This binds the format to one service's tool, and a format that needs that tool cannot leave it, as memes and virality platforms sets out. Variance collapses, and when the tool makes most choices, the rest are small.
Contributions converge because the same instrument starts from the same point. An act of authorship becomes an act of selection, and the surprise that made the format work gets harder to produce.
Both effects are real, and which one dominates is a question about a particular case rather than a general law. A page that claims featurization always kills formats is overclaiming.
The claim that adoption shortens lifespans
You will read that formats burn out faster now because the pipeline from origin to saturation has become more efficient. That argument is plausible and may well be right.
It is also a claim about measured durations, and measuring a format's duration means deciding when it started, when it ended, on which platforms, and by what count. All of that is contested.
No public dataset tracks format lifespans across eras on one set of rules, so nobody can put a defensible duration on the burnout claim today.
Google Trends is the tool people reach for, and it reports relative interest over a window you choose rather than a start or an end. The same format can look like it rose in one window and fell in another. That is a fact about the window, not the format.
If you want to make the claim, here is what would have to hold and what to check:
| The claim needs | How you would check it | Where it usually falls down |
|---|---|---|
| A start date: first post, first use by someone outside the origin group, or first use that spread | Pick one rule, apply it to every era, and write down the date you applied it | Early posts are often deleted, private or backdated, so the choice of rule can move a start date a long way |
| An end date: last new version, last use by the original group, or last use by a brand account | Date the same event in every case, from an archived copy rather than memory | Formats fade, return as mockery or continue in small groups, so the end depends on which of those you count |
| A platform scope: one named service, or a fixed list of them | Hold the list fixed across the periods you compare | A format can be spent on one service and alive on another, and widening the list can reverse the result |
| A counting method: posts, views, sound page entries or hashtag uses | Count the same unit in every period | Views, hashtags and sound pages measure different things, and platforms change what they report |
| A sample that includes formats that never spread | Draw formats from a dated list instead of from memory | The formats you can name are the ones that lasted, which is the thing being measured |
Testing the burnout claim
- Consistent start rulesame evidence standard
- Consistent end rulestated threshold for over
- Comparable measurementone instrument both eras
- Representative samplechosen without knowing outcomes
- Survivorship biasold formats that lasted
That last row is the killer. The old formats you can name are the ones that lasted, which is exactly the property being measured, and reasoning from them is textbook survivorship bias.
The backlash turn
Adoption often causes a visible reaction: the format becomes embarrassing, and using it marks you as behind. Separate that from simple exhaustion, where the slot has run out of positions.
The backlash turn is about group membership: the format now signals the wrong affiliation, and it can happen while the slot still has plenty of room. Vocabulary follows the same turn, and its signals are listed under social media slang, which sets out the terms and what each one signals.
You can tell them apart by looking at what people do next. After exhaustion, versions get weaker and then stop. After a backlash, versions stop abruptly and the format reappears shortly afterwards as a way of mocking the people who still use it.
What to record when you see adoption happening
- The date you saw it, and on which surface.
- Whether it was commercial, product or institutional adoptionthey are not interchangeable.
- Whether the format was still in active use elsewhere at the time, since the "brands killed it" story usually assumes rather than checks this.
- What the reaction was in the communities that used it, in their own words rather than as summarized elsewhere.
- Whether attribution survived the adoption, since commercial reuse rarely carries it and this is a large part of why people object.
Separate what you observed from what you concluded, and keep the observation retrievable. Explanations of why a format died are cheap; dated observations of it being used are not.
For how a format is built in the first place, see the benefits of social media trends guide, which covers how a trend starts and how it reaches a wider audience. For the pillar overview, see benefits of social media.
Common questions
Do brands actually kill formats?
They accelerate a signal that was already forming. A format only becomes usable by an outside account once it is widely legible, and wide legibility is itself the end of its use as an in-group marker.
Is featurization bad for internet culture?
It is a trade: far more participants, far less variation. Whether that is a loss depends on whether you value the output or the participation, and reasonable people land in different places.
Why does institutional use feel worse than commercial use?
Usually because it comes latest and with the least understanding, so it reads as an attempt to borrow familiarity rather than as an attempt to be funny.
