The finding
Net Influencer, working with the TikTok-first listening firm dig, indexed 58,037 US TikTok posts from the first half of 2026 across twenty editing formats. Published 12 September 2026.
The headline is about the formats built to sell.
Post count up 55%. Engagement against the average post down 41%. Typical engagement rate 1.4%, against a platform average of 6.4%.
Live shopping went the same way: engagement down 73% while its share of posts rose 27%.
Read those two lines together, because separately they mislead. Volume is climbing and attention is falling. More people are making the thing, and fewer people are watching it.
What is actually gaining
Six formats grew. Not one of them is a selling format.
- Photo carousels, which nearly doubled in post count, from 717 to 1,317, at roughly 7.4% engagement, about 15% above the platform average.
- Talking head with B-roll
- Lo-fi native UGC
- Raw phone footage
- Fast-cut edit
- Montage set to music
Three of those six are defined by looking unproduced. That is the pattern worth sitting with, and it is not new: an audience learns what an advert looks like, and then learns to scroll it.
Not sure which of your formats is actually carrying the channel?
The Quick Scan reads your public uploads as a set and shows which ones hold attention and which ones only look busy. No login, no account access.
Run your free Quick ScanWhat it does not say
It does not say tagging products loses you money. Engagement is attention, not revenue, and a video with a 1.4% engagement rate can still outsell one at 7.4% if the people it does reach are ready to buy. Nobody has published that comparison and this index does not claim to.
It also does not say photo carousels will work for you. The index measures what is happening across 58,037 posts, not what happens on your channel.
Their own limits are worth repeating: formats with under 300 posts are marked directional rather than firm, and the top 1% most-viewed posts were removed so one viral video cannot carry a format. That second decision is the reason the numbers are worth reading at all.
What to do with it
The useful move is not to stop selling. It is to stop letting the selling format do the reaching.
Separate the two jobs. The formats that hold attention should do the reaching. Selling belongs where someone has already chosen to be: the profile, the link, the page. That is the structure of any well-built funnel, and TikTok has just put a number on the cost of skipping it.
And if your view counts fell this quarter while you were doing more shoppable posts, this is a likelier explanation than anything about the algorithm.
Common questions
Does this mean I should stop using TikTok Shop?
No. The index measures engagement, not sales. What it shows is that the selling format is a weaker way to REACH people than it used to be, so leaning on it for discovery costs you. Selling to people who already found you is a different question and this data does not answer it.
Why would tagging a product lower engagement?
Because an audience learns what an advert looks like. Once a format reads as a sales post, people scroll it before watching, and the engagement rate reflects the scroll rather than the product.
What is the most useful format on that list?
Photo carousels showed the clearest gain, nearly doubling in post count at roughly 7.4% engagement against a 6.4% average. That said, three of the six gaining formats are defined by looking unproduced, which is the cheaper lesson: polish is not what is winning.
Where do these numbers come from?
The Q2 Editing Format Index by Net Influencer with dig, published 12 September 2026: 58,037 US TikTok posts from the first half of 2026 across 20 editing formats, with the top 1% most-viewed posts removed and formats under 300 posts marked directional.