---
title: 'Your channel doesn''t grow if you use A/B testing. Here''s proof.'
source: 'https://youtube.com/watch?v=jn2UFbpeOgM'
video_id: 'jn2UFbpeOgM'
date: 2026-07-28
duration_sec: 521
---

# Your channel doesn't grow if you use A/B testing. Here's proof.

> Source: [Your channel doesn't grow if you use A/B testing. Here's proof.](https://youtube.com/watch?v=jn2UFbpeOgM)

## Summary

The video argues that YouTube's A/B testing feature can harm video performance, especially for small channels, by splitting traffic among experimental titles and delaying the conclusion, which damages initial momentum. The creator shares personal tests where six out of seven videos underperformed when using the tool.

### Key Points

- **Traffic split issue** [0:48] — When testing three titles, 66% of viewers see experimental titles that may be terrible, and YouTube judges the video during the test.
- **YouTube's 2026 changes** [1:26] — YouTube now allows testing up to three titles, three thumbnails, or combinations, and picks the winner based on watch time and average view duration, not CTR.
- **Test duration trap** [2:33] — If no clear winner, the test can drag up to 14 days, damaging the video's momentum and causing it to lose the initial 48-hour boost.
- **Math of the test** [3:11] — With three titles, traffic is split 33/33/33. Even if one title is a clear winner, the other two experimental titles are shown to 66% of the audience, potentially signaling to YouTube that the video is bad.
- **Recommendation for small channels** [5:01] — Only use A/B testing if you average at least 1,000 views per video in the first hour; otherwise, avoid it.
- **Alternative strategies** [6:12] — Limit tests to 24 hours and manually pick a winner, or do manual title swaps every 24 hours based on performance.

### Conclusion

YouTube's A/B testing can backfire for small channels by splitting traffic and delaying results, harming video performance. The creator advises using manual testing or limiting the test duration to avoid damaging momentum.

## Transcript

Everyone says AB testing your thumbnail
and titles is the smartest thing you can
do for your YouTube growth. YouTube
literally built the feature for you, so
of course you should use it, right?
Wrong. Because if AB testing is supposed
to help your video perform better, then
why did six out of seven of my videos
completely tank the moment I turned it
on? I was just like you. I saw the big
creators raving about it. Mr. Beast uses
it. So, I thought, "Okay, let's me go
all in." I ran AB testing on seven brand
new videos back-to-back. I trusted the
tool and one by one I watched those
videos die. But, what if AB testing
itself isn't broken? What if YouTube is
quietly punishing you every single day
the test is still running and you don't
even know about it? Here's what I found.
When YouTube runs your test, it splits
your traffic equally between all your
variations. So, if you're testing three
titles, 66% of your audience is seeing
an experimental title that might be
terrible. And YouTube is judging your
video the entire time this is happening.
So, in this video I'm going to show you
the actual math behind why this tool can
destroy your video's momentum and who
should and honestly should not be using
it. But first, let's me quickly break
down how AB testing actually works in
2026 because YouTube just made some big
changes to it. As of 2026, YouTube is no
longer just AB testing thumbnails. You
can now test both titles and thumbnails.
You can test up to three different
titles, three different thumbnails, or
three combinations of titles and
thumbnails. How cool is that? One thing
most people don't understand is YouTube
doesn't pick the winner based on the
click-through rate or CTR. Instead,
YouTube picks based on watch time and
average view duration. That's quite
interesting, isn't it? I mean, it is
called AB testing, so wouldn't they want
to track clicks instead? The truth is
YouTube is doing this because YouTube
doesn't want clickbait. For example,
take a look at these three thumbnails.
If version A gets 10% clicks, but
everyone leaves in 5 seconds because it
is clearly clickbait, and version B gets
2% clicks, but they watch for 50% of the
video, then version B wins even though
the CTR is significantly lower. And
here's what I discovered when running AB
testing on seven newly launched videos.
Based on my tests, all six out of my
seven videos of mine underperformed,
failing to gain 1,000 views. You see,
the biggest issue with YouTube AB
testing is if it can't find a clear
winner, it drags the test up to a
maximum of 14 days. That's actually a
trap because it damages your channel
each day it drags. Let me explain to you
why. Check out this test I did for one
of my videos. Nothing fancy, just a
title test. Based on the final result,
it is clear my top title is the winner
with over 46%
watch time share. But you see, if you
squeeze your eyes real good, you'll
notice that this test took over 11 days
to conclude. So, just by thinking of
this logically, the math doesn't add up.
You see, in order to conclude which
title is the best, YouTube needs to test
all three of them, meaning my traffic is
split 33 33 33. If I have one clear
winner, then that makes the other two
experimental titles. And those two
experimental titles are shown to 66%
of my audience. So, let's do a simple
logical walk-through here. YouTube
starts with the top one, found that it
works, has over 40% watch time. But,
YouTube doesn't know if this is the best
yet, so it pauses the top one, then
tests the second title. Oh, no, the
second one did not perform well after 3
days. So, let's move to the third one.
Again, oh, no, the third one failed,
too. So, that makes the first one the
clear winner. And the entire test took
11 days. So, tell me, if you knew the
first one is the winner, then you would
stop the test and just use the first
title, right? But, you can't do that if
you are AB testing and I think that's a
massive flaw in this feature. Moreover,
I have a very bad feeling that when
YouTube is testing my experimental
titles and both underperformed, YouTube
might have received the signal that the
video is not a good one. Therefore, when
it decides that the first one is the
winner, the damage has already been
done. It is already too late. YouTube
thinks this video is bad because of your
experimental titles. The video has
already lost its initial first 48 hours
boost, so YouTube decides to not give
any more impressions to it and the video
fades away. I definitely do not have
concrete proof of that, but that's what
I saw in my AB testing and I hope that
isn't true. That's not to say this is
entirely YouTube's fault. There's
certainly an angle where it doesn't make
sense to use AB testing, especially for
small YouTube channels. And when I say
small YouTube channels, I don't mean by
subscriber count. I mean by view count.
You have probably heard that AB testing
has been a game-changer, that it helped
many videos boost their CTR and
subsequently more views. What you might
be aware of as well is that those
sayings often come from big channels.
So, what this means is that you need a
big amount of views in order to conclude
your AB testing effectively because if
you don't have sufficient views, then
your test will not only drag on, but it
will also fail to come to an accurate
conclusion. From my own experience and
testing, I highly suggest using AB
testing only if you can average 1,000
views per newly launched video,
preferably about 1,000 views in the
first hour. The more, the merrier, of
course. The reason is because you want
to have enough views to really run the
AB test to the fullest and also to have
AB testing conclude as soon as possible,
preferably within the first day. Heck,
even 1,000 views is me being optimistic.
I'm actually thinking about 10,000 views
average. So, if you're getting less than
1,000 views per video, I suggest you
stop using AB testing and instead use
this strategy pivot. My first strategy
is if I choose to use AB testing, I will
let it run for the first 24 hours only.
Then, I will evaluate the test so far
and manually select the one that is the
clear winner. I will not let it drag on
for more days. My other strategy is to
avoid AB testing altogether and do
manual swaps. So, if I am testing the
first title at launch, I will check it
again after 24 hours. If it fails, I
will manually swap to the second title
and test it. After that, I will pick and
go with my gut feelings based on which
gets more views. The new video boost
usually lasts for the first 48 hours or
so, so this to me is quite a good test
provided you are disciplined enough to
do it manually. And also, if you don't
have the traffic to support your tests,
then avoid testing on brand new videos.
You have already put in a lot of effort
and time into it, so don't let the video
underperform because of YouTube's tool.
Another strategy I tried is to look out
for underdogs in your channel, such as
videos with high impressions, but low
CTR. Videos like this actually signal
that YouTube sees a market in it, but no
one is clicking because of a bad
thumbnail or title. And I think AB
testing does help when it comes to this.
In fact, I actually tested this on one
of my videos. It is rather old, a few
years old, so I just did some tests on
it just to see what happens. But so far,
it doesn't look like I managed to revive
it, although I haven't left it long
enough yet, so maybe YouTube still needs
time to follow this one up, so we'll
see. So, yes, I am very disappointed
with YouTube's AB testing. I really wish
it worked for everyone, big or small
channels, but it is very clear to me
that I should stop doing any more AB
testing for now and instead rely on my
instinct. Seven videos used to test AB
testing is a lot of videos, but hey, I
think I gained quite a lot from it and I
hope you do, too. Also, I run YouTube
channel audits as well for my viewers,
where I will audit and give my most
honest feedback to growing YouTube
channels, help give them a direction,
and help them grow their channel. If
you're interested, check out the link in
the description. So, this concludes my
AB test. The next test I will do is
going to be what happens if I were to
upload seven videos without AB testing.
Is my gut feeling more accurate than
some YouTube tool? That's what I'm going
to test next, so when the video is
ready, it will appear somewhere over
here. So, be sure to check it out. All
right, thanks for watching and I'll
catch you in the next one.