How we ran this study
To answer this, a sample of videos wasn't enough. We needed every upload a company had posted, month by month, so we could see who kept going, for how long, and what happened to their views. Nobody publishes that, so we built it.
- 4,697
- company homepages checked for a channel
- 1,772
- company YouTube channels
- 241,060
- videos dated, from full upload histories
- 100,572
- videos with their views and length
- 1,042,740
- minutes of video
- 4.8 billion
- views counted
We drew 4,697 US companies with 51 to 500 employees at random from lists of private software and industrial firms, and checked every homepage for a link to a YouTube channel. Then we pulled each channel's full upload history from the YouTube Data API: the date of every video since 2019, and for every channel that kept a streak, each video's views and length. Figure 1 shows how the sample narrowed.
From 4,697 companies to 211 that kept it up
Each step of the sample, and what was kept
- 4,697US companies with 51 to 500 employees, drawn at random from lists of software and industrial firms
- 2,181of their homepages link to a YouTube channel
- 1,772channels we read, with the date of every upload: 241,060 videos
- 211posted at least twice a month for 12 months in a row
- 24,902videos in those 211 streaks, each with its views and length
The full method, and the three results that looked real but weren't, are at the end of the page.
Key findings
- No: posting consistently doesn't make a channel grow. It stops it from shrinking. Two years in, new videos on channels that kept posting got 6% fewer views than their first ones. Measured the same way, everyone else's got 43% fewer.
- Almost nobody keeps it up. Only 211 of 1,772 company channels, about 1 in 8, ever posted twice a month for 12 months in a row.
- Most of the ones that did still didn't grow. 45% of them got more views per video at the end of the year than at the start. Among channels without a streak, 32% did.
- It takes a year to show. The gap opens slowly through the first year and keeps widening in the second. No month brings a jump.
- One viral video doesn't carry a channel. The videos after a hit did no better than the videos after an ordinary upload.
- It isn't harder now than in 2019. Streaks that started in 2023 and 2024 did as well as streaks from 2019 and 2020.
- What they posted mattered less than you'd think. Product videos pointed up and videos about the company pointed down, as in Study 01, but neither is clear of chance yet.
- The biggest growers had no shared formula. Against everyone else, they leaned toward product videos, questions in titles and less about the company, but nothing we measured set them apart clearly. What they did share was the result: four times the subscribers.
Consistency stops the fade
Measured today, a channel's newer videos almost always have fewer views than its older ones. Part of that is real: audiences drift away. Part of it is time: a video posted two years earlier has had two more years to collect views. Both kinds of channel carry that same handicap, so the gap between the two lines in Figure 2 is what posting consistently changed.
Channels that posted at least twice a month for a year held close to what their first videos got, month after month. Channels without a streak slid: 0.82x by month 12, 0.57x by month 24.
Channels that kept posting held their views. The rest faded.
Median views of each month's new videos, against the first 3 months (1.0x)
Show the numbers
| Month | Kept posting | Without a streak | Channels (kept / without) |
|---|---|---|---|
| 4 | 0.97x | 0.95x | 211 / 326 |
| 5 | 0.99x | 0.92x | 211 / 316 |
| 6 | 0.95x | 0.94x | 211 / 312 |
| 7 | 0.95x | 0.91x | 211 / 317 |
| 8 | 0.80x | 0.89x | 211 / 313 |
| 9 | 0.92x | 0.87x | 211 / 309 |
| 10 | 0.91x | 0.82x | 211 / 297 |
| 11 | 0.94x | 0.83x | 211 / 298 |
| 12 | 0.93x | 0.82x | 211 / 300 |
| 13 | 0.89x | 0.81x | 204 / 298 |
| 14 | 0.98x | 0.81x | 198 / 299 |
| 15 | 0.93x | 0.80x | 190 / 302 |
| 16 | 0.99x | 0.77x | 193 / 309 |
| 17 | 0.87x | 0.70x | 186 / 309 |
| 18 | 0.90x | 0.72x | 189 / 310 |
| 19 | 0.85x | 0.69x | 188 / 307 |
| 20 | 0.84x | 0.69x | 183 / 302 |
| 21 | 0.92x | 0.66x | 192 / 306 |
| 22 | 0.95x | 0.64x | 189 / 303 |
| 23 | 0.96x | 0.58x | 183 / 305 |
| 24 | 0.94x | 0.57x | 173 / 305 |
So what posting consistently bought these channels wasn't a rising line. It was a line that didn't fall.
Almost nobody keeps it up
The bar is low: two videos a month, for 12 months in a row. That's 24 videos in a year, or one every two weeks.
Of 1,772 company channels, 211 ever did it between 2019 and 2024 (Figure 1, above).
Most consistent channels still didn't grow
Keeping a streak did help (Figure 3). 45% of the channels that kept posting ended their year with more views per video than they started with, against 32% of channels without a streak. With 211 and 310 channels, a gap that size is very unlikely to be chance.
Most channels that kept posting still didn't grow
Share of channels whose new videos got more views in months 10 to 12 than in months 1 to 3
But 45% is still under half. The median channel that kept posting ended its year at 0.90x: its new videos got slightly fewer views than its first ones. It just lost less than the rest.
It takes a year to show
In the first six months, the two lines in Figure 2 are almost on top of each other. Channels that kept posting were 1.01x the channels without a streak by month 6, 1.13x by month 12, 1.25x by month 18 and 1.65x by month 24.
There is no month where it clicks. The difference builds slowly, mostly because the channels without a streak keep sliding while the consistent ones hold.
One viral video doesn't carry a channel
Most channels that kept posting had at least one video with 10 times their usual views: 182 of 211. And a channel's single best video earned a median 18% of everything it got that year (over half, on 26 channels).
You'd expect a hit to bring in new subscribers who watch what comes next. It didn't (Figure 4). The 90 days of videos after a hit did no better than the 90 days after an ordinary video on the same channel.
The videos after a hit did no better
Median views of the next 90 days of uploads, against the 90 days before (1.0x)
A hit is good for that video. It doesn't change what the next one gets.
It isn't harder now than in 2019
Streaks that started in 2023 and 2024 grew about as much as streaks from 2019 and 2020 (Table 1). The raw numbers look different by year, but the older streaks started from more views, and channels that start high tend to fall back. Once starting views are held fixed, the years are the same.
Streaks by the year they started
| Streak started | Channels | Median growth | Share that grew | Without a streak, same years | Starting views (median) |
|---|---|---|---|---|---|
| 2019 to 2020 | 49 | 0.63x (0.52–1.23) | 47% | 0.86x | 930 |
| 2021 to 2022 | 59 | 0.99x (0.85–1.14) | 47% | 0.89x | 298 |
| 2023 to 2024 | 103 | 0.85x (0.70–1.06) | 42% | 0.76x | 264 |
What they posted mattered less than you'd think
We sorted every video by its title, using the same rules as Study 01, and tested whether the mix made a difference (Figure 5). We also tested how often they posted, how long the videos were, Shorts, how evenly the uploads were spaced, and early likes and comments.
Almost nothing about what they posted was clear of chance
Effect on growth, with starting views held fixed · dot = estimate, line = 95% interval
Show the numbers
| What the channel did | Compared | Effect on growth | 95% interval | Channels |
|---|---|---|---|---|
| Product videos (demos and how-tos) | all vs none | 1.30x | 0.87–2.01 | 207 |
| Like rate on the first videos | per 10x | 1.27x | 1.00–1.65 | 205 |
| Comment rate on the first videos | per 10x | 1.10x | 0.94–1.28 | 205 |
| Uploads per month | per 10x | 0.97x | 0.61–1.58 | 207 |
| Shorts | all vs none | 0.95x | 0.62–1.49 | 207 |
| Uneven gaps between uploads | per step | 0.95x | 0.51–1.68 | 211 |
| Length of the long videos | per 10x | 0.93x | 0.70–1.21 | 207 |
| About-the-company videos | all vs none | 0.80x | 0.52–1.18 | 207 |
| A recurring named series | all vs none | 0.73x | 0.37–1.61 | 207 |
| A video with 10x the usual views | had one vs not | 0.67x | 0.44–1.01 | 211 |
| Several uploads on the same day | all vs none | 0.48x | 0.26–0.92 | 211 |
The directions match Study 01. Channels that posted more product demos and how-tos grew more (1.30x going from none to all), and channels that posted more videos about the company grew less (0.80x). But with 207 channels, neither is clear of chance.
Posting more often made no difference within a streak (0.97x for 10 times as many uploads). Neither did video length or Shorts.
One result did clear chance: channels that posted several videos on the same day grew less (0.48x going from none to all). It is one result out of eleven, so we want to see it again before calling it a finding.
What the biggest growers had in common
We took the top quarter of the channels that kept posting, by growth with starting views held fixed, and compared them with everyone else on everything we could measure (Table 2). If there were a formula, this is where it would show.
The biggest growers against everyone else
| What we compared | Growers | The rest | Clear of chance? |
|---|---|---|---|
| Product videos (demos, how-tos), share of sorted videos | 40% | 28% | No (p = 0.15) |
| Videos about the company, share of sorted videos | 59% | 72% | No (p = 0.19) |
| Webinars and podcasts, share of long videos | 16% | 25% | No (p = 0.41) |
| Titles that name the company | 20% | 32% | No (p = 0.28) |
| Titles that ask a question | 6.5% | 3.0% | Only under one definition (p = 0.02, 0.06) |
| Channels running a named series | 28% | 41% | Only under one definition (p = 0.01, 0.22) |
| Most-watched video about the audience's world, not the company (hand-read) | 55% | 41% | No (p = 0.26) |
| Shorts, share of uploads | 24% | 21% | No (p = 0.47) |
| Uploads per month | 7.9 | 7.7 | No (p = 0.99) |
| Length of the long videos, minutes | 6.5 | 5.9 | No (p = 0.60) |
| Several uploads on the same day, share of uploads | 20% | 24% | No (p = 0.24) |
| Starting views (months 1 to 3) | 512 | 323 | No (p = 0.37) |
| Software companies, share of channels | 77% | 84% | No |
| 51 to 200 employees, share of channels | 68% | 75% | No |
| Subscribers today | 10,500 | 2,500 | Yes, but it is a result of growing (p = 0.001) |
There isn't one. Every lean points the same way: the growers posted more product videos, fewer videos about the company, fewer webinars, fewer titles with the company's name and more titles that ask a question. But no single difference is clear of chance, and the two that looked clear under one definition of a grower disappeared under the other. Shorts, how often they posted, video length, starting views, industry and company size made no difference at all.
Read one at a time, the top channels look like they share a recipe (Table 3). Jerry, a car insurance app, posted stories about celebrities' cars. Fly.io explained what developers were asking about that month. National Equipment Dealers showed how to run the trucks it sells.
Five of the channels that grew the most
| Channel | Months 1–3 | Months 10–12 | Growth | What they posted |
|---|---|---|---|---|
| Jerry | 232 | 3,573 | 15.4x | Short videos about celebrities' cars, like why Ferrari blacklisted some celebrities. The best got 20.8 million views. |
| Fly.io | 804 | 7,705 | 9.6x | Explainers on what developers were asking about that month: MCP, vibe coding, building an AI app. |
| OX Security | 83 | 726 | 8.7x | Security explainers and long interviews with security chiefs. |
| National Equipment Dealers | 348 | 2,287 | 6.6x | How to inspect and operate the trucks it sells, in English and Spanish. |
| DW Drums | 1,676 | 10,785 | 6.4x | A factory tour and full-length coverage of its own drum event, for an audience of drummers. |
So we tested that recipe blind. For the channels that doubled their views and 58 others drawn at random, we read the most-watched video of each one's last 3 months without knowing which group it came from, and asked one question: is it about the audience's world, or about the company? 55% of the growers' top videos were about the audience's world, against 41% of the others. A lean the same way, but not clear of chance with 87 channels.
What the growers shared for certain was the result: a median of 10,500 subscribers today, against 2,500 for the rest.
How we measured it, in full
The companies. We drew 4,697 US companies with 51 to 500 employees at random from lists of private software and industrial firms, checked each homepage, and kept the 2,181 that link a YouTube channel. Through the YouTube Data API, we read the date of every upload on 1,772 of those channels: 241,060 videos. The other 409 were left out before any result was read: 358 because the API's daily limit ran out before we reached them, and 51 because the link didn't lead to a channel we could read. We read software companies' channels before industrial ones, so the 358 are all industrial: we read 97% of the software channels but 64% of the industrial ones.
The streak. A channel has a streak if it posted at least two videos (not livestreams) in each of 12 calendar months in a row. The streak had to start between January 2019 and October 2024, so every video in it is at least a year old. Where a channel had more than one, we used its first.
Growth. For each streak, the median views of the videos posted in its last 3 months, divided by the median views of the videos posted in its first 3. Above 1.0x, it grew. Medians, so one viral video can't decide it.
The comparison. Older videos have had longer to collect views, so a channel's later videos will look worse even if nothing changed. To measure that drift, we took the same ratio on every 12-month span from 2019 to 2024 on channels that never kept a streak. That is the line without a streak in Figure 2.
Holding starting views fixed. Channels that start with very few views tend to rise and channels that start with many tend to fall, whatever they post. Every comparison in Figure 5 and Table 1 is a regression that includes starting views, and every interval is a bootstrap over channels, so each channel counts once.
How we checked it
We found three results that looked real and weren't. We report them so you can see what the numbers above survived.
- "If you're ahead by month 6, you'll probably grow." Channels ahead at months 4 to 6 grew 65% of the time, against 28% for channels behind. But channels without a streak show the same split (62% and 30%). It comes from sharing the same first 3 months, not from posting consistently.
- "A hit lifts the next videos 1.22x." A first version compared each hit with the 90 days before it. A video stands out most against a quiet stretch, so that version picked quiet stretches, and the next 90 days looked better simply because the quiet stretch ended. Measured against the channel's whole streak, the lift is 1.03x.
- "Recurring series cut growth in half." A first version counted a company's own name ("… | Acme") and plain video types ("Webinar: …") as a series. Read by hand, about half weren't series. With those removed, the effect is 0.73x and inside chance.
What this does not show
- Views, not sales. We measured views on YouTube. A channel that holds its views may or may not be bringing in customers.
- No subscribers over time. YouTube only gives today's subscriber count, so we couldn't track it through the streak.
- Kinds of video come from titles. Many titles match no rule, which blurs the content mix.
- Shorts are inferred from length. Videos of 60 seconds or less before October 15, 2024, and 3 minutes or less after, when YouTube raised the limit.
- Mostly software. 174 of the 211 channels with a streak are software companies. The 37 industrial ones grew a little less (0.75x against 0.90x), too few to say more. Industrial companies are also under-read: the API's daily limit cut off 358 of their channels, so the averages lean toward software more than the companies we drew do.
- Companies with a channel. We only looked at companies that link a YouTube channel from their homepage, a little under half.
Our takeaways
This is where we stop reporting and say what we think, as a company that makes videos for B2B teams. We've marked how sure we are of each part.
What the data shows clearly.
- Commit to a year. Posting at least twice a month is what separated channels that held their views from the ones that faded, and the difference took a year to show. Judge it at 12 months, not 3.
- There is no one format. The channels that grew used Shorts and didn't, posted 3 times a month and 15, made 2-minute videos and 20-minute ones. If someone sells you a format that guarantees growth, this data doesn't back it.
- Don't plan around a hit. A viral video is good for that video. The next ones didn't do any better.
What the data leans toward. Every one of these points the same way in this study, and Study 01 found the same with a different method and different companies. No single one is proven here.
- Show the product doing real work. Demos and how-tos did better than average. Videos about the company did worse.
- Answer the questions your buyers already ask. The channels that grew put more questions in their titles, and their most-watched videos were more often about the buyer's world.
- Talk about yourself less. Fewer webinars, fewer event recaps, fewer titles that start with the company's name.
Our read, which we can't prove with titles. The channels that grew the most each made videos for one specific audience that would watch anyway: car fans, developers, equipment operators, drummers. Sometimes that meant the buyer's world, sometimes the product itself. The common thread looks less like a format and more like knowing exactly who the video is for.
Suggested citation
Pingale, C. (2026). Does posting consistently grow a B2B YouTube channel? Harloop Research, Study 02. harloop.com/research/youtube-consistency
The data behind every channel in the study is linked under "Data" at the top of the page: every channel with a streak. Upload dates and view counts were read from YouTube on November 9, 2026.