Key findings: new Stillfeed users hit their first blocked Reels or Shorts page a median 4 minutes 40 seconds after first launching the app (253 installs), and 76% within 10 minutes. On the days Stillfeed blocked anything, the median person had 4 block events that day, while the top tenth of days had 80 or more (308 people, 656 person-days). The median person who used a platform inside the app spent 7 minutes there that day (208 people, 672 person-days). Of 6,714 platform opens from 1 to 10 November, 71% were Instagram and 18% YouTube. Read the methodology first: users of a feed-hiding app are not a random sample.
Stillfeed is an iPhone app that opens Instagram, YouTube and six other platforms in its own browser and hides the parts built to keep you scrolling, such as Reels, Shorts and recommendation feeds. This report uses the app’s anonymous usage statistics to describe what happens in the first minutes and days of use. It is a first edition from a young dataset: the app’s analytics began on 14 October 2026, so the window is four weeks long and the groups are small. Where a group had fewer than 100 people, we did not publish it.
Time to first block
For each install, the app records the seconds between its first launch and the first time it blocked something. The table covers installs that first launched on or after 27 October 2026, so every row measures a fresh install. Installs that never recorded a block are not in it.
| Measure | Figure | Sample | Period |
|---|---|---|---|
| Median time from first launch to first block | 4 min 40 s | 253 installs | 27 Oct to 10 Nov 2026 |
| Middle half of installs (25th to 75th percentile) | 2 min 31 s to 9 min 43 s | 253 installs | 27 Oct to 10 Nov 2026 |
| Installs with a first block within 10 minutes | 76% (192 of 253) | 253 installs | 27 Oct to 10 Nov 2026 |
| Installs with a first block within 24 hours | 96% (243 of 253) | 253 installs | 27 Oct to 10 Nov 2026 |
Most people hit a block almost immediately, usually the first time they open a platform and tap toward a Reels, Explore or Shorts page. The 90th percentile is about 3 hours 40 minutes, so a small group takes much longer.
Block events per day
A block event is one thing Stillfeed stopped: a Reels, Explore, Shorts or recommendation-feed page it kept from opening, or a card it removed from a feed, such as YouTube’s Shorts shelf. These are counts of what the app did, not counts of videos, and different kinds of block are counted at different scales, so we publish the overall figure and not a per-kind comparison. The table covers person-days with at least one block event.
| Measure | Figure | Sample | Period |
|---|---|---|---|
| Median block events per person per day | 4 | 308 people, 656 person-days | 22 Oct to 10 Nov 2026 |
| 90th percentile (the heaviest tenth of days) | 80 or more | 308 people, 656 person-days | 22 Oct to 10 Nov 2026 |
| Mean block events per person per day | 54 | 308 people, 656 person-days | 22 Oct to 10 Nov 2026 |
The mean is more than ten times the median because a few heavy days dominate it, mostly from the Shorts shelf on YouTube, where one scroll can remove many cards. The median is the better description of a typical day.
Minutes per day inside the app
The app counts the seconds a platform is open in the foreground inside Stillfeed and sends one total per completed day. A day with no platform time is not sent, so the figures describe days on which someone used at least one platform in the app. This is time inside Stillfeed only, so it says nothing about time in other apps or in the platforms’ own apps.
| Measure | Median | 25th to 75th percentile | 90th percentile | Sample |
|---|---|---|---|---|
| All days | 7 min | 2 to 25 min | 85 min | 208 people, 672 person-days |
| Monday to Friday | 7 min | 2 to 25 min | 77 min | 181 people, 467 person-days |
| Saturday and Sunday | 6 min | 2 to 25 min | 100 min | 117 people, 205 person-days |
All three rows cover 26 October to 10 November 2026, and the days are the device’s own calendar days. The weekday and weekend medians differ by one minute, which is within the noise of a sample this size. The weekend’s longer heavy tail (100 minutes at the 90th percentile, against 77) is worth watching as the dataset grows, but we would not call it a finding yet. Two people with impossible dates in their records, from device clock errors, were excluded.
Which platforms people open
Each time someone opens a platform inside Stillfeed, the app records which one. Platforms joined the app at different times in October, so this table uses 1 to 10 November 2026, when all eight were available. An open is a tap to start a platform, not a measure of time spent. One person can open several platforms, so the people columns add up to more than the total.
| Platform | Share of opens | Opens | People who opened it |
|---|---|---|---|
| 71.2% | 4,781 | 604 (90% of 673) | |
| YouTube | 17.7% | 1,188 | 195 (29% of 673) |
| All six others combined (X, Snapchat, TikTok, Facebook, Reddit, LinkedIn) | 11.1% | 745 | 186 (28% of 673) |
| All platforms | 100% | 6,714 | 673 |
The six smaller platforms are combined because most had fewer than 100 people in the window. Instagram leads on both measures: nine in ten people who opened any platform opened Instagram. For the published, sourced numbers on time spent per platform across the wider population, see our screen time and social media statistics.
Methodology and limits
Source. Product analytics events sent by the Stillfeed iPhone app, stored by our analytics provider in the EU. The events used are the app’s first-block event, its block events, its daily usage total and its platform-open event. Each event carries a random identifier generated on the device, not a name, email or account; see the privacy policy. Analytics are on by default and can be turned off in the app, and people who turned them off are not in this data.
Sample and period. Each table states its own people, person-days and dates. The overall window runs from 14 October 2026, when analytics began, to 10 November 2026. Counts of people are random identifiers, so one person with two devices counts twice and a reinstall may count as a new person.
Statistics. We report medians and percentiles because the distributions are heavily skewed. Block events and minutes are summed per person per day first, and the median is taken across those person-days, so people who used the app more often contribute more rows. Days follow the UTC calendar for block events and the device’s own calendar for minutes.
Privacy. Only aggregates are published. No individual’s data is shown, and no group of fewer than 100 people is reported. Internal test traffic (developer devices, simulators and beta builds) was excluded.
Limits you should keep in mind.
- Not a random sample. These are people who installed an app that hides feeds. They may be more concerned about scrolling than the average user, and people who dislike analytics are less likely to appear.
- Inside the app only. Minutes and opens describe use of platforms through Stillfeed, not total phone or social media time.
- Short and early. Four weeks, small groups, and an app still adding platforms and features. Figures can move a lot between editions.
- Counts depend on how the app counts. A block event is whatever the app records as one block, and the definition differs by kind.
- Correlation only. The report describes behavior. It does not show that blocking changes how long anyone scrolls.
What we left out
We published only measures whose meaning we could verify and whose groups had at least 100 people. We left out the following, and will add them when the data supports them:
- Blocks per kind (Reels, Shorts, Explore) and per platform, because most groups were under 100 people and the counting scale differs by kind.
- Time-of-day patterns, because the events carry no reliable local time.
- How session length changes over the first 30 days, because the analytics began four weeks ago and no install has a full 30 days yet.
- Quiet hours, time limits and session timers, because only a handful of people had used them in the window.
- Day-by-day weekday figures, because single weekdays fell under 100 people.
Cite this report
Stillfeed Team, “Stillfeed Usage Report 2026: Reels, Shorts and Feed Time”, 11 November 2026. https://stillfeed.pro/research/reels-shorts-report-2026/
Each row has its own link: add #finding- and the row id, for example https://stillfeed.pro/research/reels-shorts-report-2026/#finding-first-block-median. When you quote a figure, please include its sample size and period, which sit beside it in every table.
License. The tables and figures in this report are free to reuse, including in press and teaching, under CC BY 4.0, with credit to Stillfeed and a link to this page. Questions or corrections: see the press page.
Common questions
How long does it take a new Stillfeed user to hit their first blocked Reels or Shorts page?
Among 253 installs that first launched the app between 27 October and 10 November 2026 and later recorded a block, the median time from first launch to the first block was 4 minutes 40 seconds. 76% of them had one within 10 minutes and 96% within a day.
How much time do people spend inside Stillfeed each day?
On days when they used at least one platform inside the app, the median person spent 7 minutes there (208 people, 672 person-days, 26 October to 10 November 2026). The top 10% of days ran to 85 minutes or more. This counts time inside Stillfeed only, not time in other apps.
Is this a representative sample of phone users?
No. These are people who chose to install an app that hides feeds, and only those who left anonymous usage statistics on. The figures describe this group during a four-week window and should not be read as averages for all social media users.
