---
title: 'Does A/B Thumbnail Testing HURT Your Channel?'
source: 'https://youtube.com/watch?v=-rBAZR6_-a8'
video_id: '-rBAZR6_-a8'
date: 2026-06-14
duration_sec: 296
---

# Does A/B Thumbnail Testing HURT Your Channel?

> Source: [Does A/B Thumbnail Testing HURT Your Channel?](https://youtube.com/watch?v=-rBAZR6_-a8)

## Summary

The hosts discuss whether YouTube's A/B thumbnail testing feature can hurt a channel's views. They argue that while useful in specific scenarios, it's often misused and can lead to lost clicks if a poor thumbnail is tested against a good one.

### Key Points

- **Introduction to A/B thumbnail testing** [0:00] — The hosts introduce the topic of A/B thumbnail testing and whether it hurts views.
- **Fan question** [0:10] — A fan asks if A/B testing might be giving fewer views and whether to end it early.
- **Strong feelings on misuse** [0:34] — Jen has strong feelings because creators ran with the feature without setting guidelines for when to use it.
- **Similar percentages make testing irrelevant** [0:55] — On their main channel, 9 out of 10 times the percentages are so similar that testing is irrelevant.
- **Pointless without drastically different thumbnails** [1:14] — A/B testing is pointless unless you have two drastically different thumbnail styles.
- **Both thumbnails must be good** [1:21] — Testing one good thumbnail against a poor one is bad because the poor thumbnail still gets shown and loses clicks.
- **Losing thumbnail not resurfaced** [2:01] — Travis asked YouTube's V. Richie: people who didn't click on the losing thumbnail won't automatically get resurfaced with the winner.
- **When to use A/B testing** [2:31] — Use it when you have two really good thumbnail ideas and aren't sure which to choose.
- **Not for every video, especially small channels** [2:40] — Smaller channels often get inconclusive results due to low impressions.
- **Testing doesn't end after 48-72 hours** [2:54] — Different audiences see the video over weeks and months, so a single test may not be definitive.
- **Multiple A/B tests recommended** [3:07] — Jen suggests three tests: initial community test (first 3 days), casual audience test (rest of week), and a highly discoverable option later.
- **Community vs. discoverable thumbnails** [3:47] — A thumbnail with a face may win early because subscribers know the creator, but it may not be best for long-term discoverability.
- **No definitive answer** [4:19] — Just because a thumbnail works today doesn't mean it will work tomorrow; the winning thumbnail may not be the best for the video's lifetime.

### Conclusion

A/B thumbnail testing is not inherently harmful, but it should be used strategically with two strong thumbnails. Small channels may not benefit due to low impressions, and even for larger channels, the test results may not reflect long-term performance.

## Transcript

today on the Boost we're going to talk
about the ab thumbnail testing and
whether or not you should actually be
using it is it hurting your views we'll
talk about it here on the Boost by vid
IQ all right J we just got a text
message from one of our fans here let's
take a look at what it says sup Travis
and Jen I love you guys thank you for
being awesome I was wondering what you
guys think of the ab thumbnail testing
feature on YouTube I'm currently using
it on my new video and I'm afraid that
it might be giving me less views and
hurt my potential of the video going out
should I just end it early and let it
cook it's a good question Jen but do you
think about this do we let it cook I
have strong feelings about this actually
because I think this came out and we all
kind of ran with it because we were so
happy and so excited that I don't know
if we really set guidelines for
ourselves as to when to use it and when
not to use it cuz I do think there is
1,000% times where you should not use
this one of the things is we've been
testing this on the main channel for a
while nine times out of 10 the
percentages are so similar that it's
like irrelevant right like it's pretty
irrelevant if you go into making a video
knowing in advance uh what your
thumbnail title are going to be then
they're probably going to connect pretty
well and unless you're doing drastically
different thumbnail Styles which by the
way is kind of what AB testing is for
then it's kind of pointless right like
otherwise what's the point it's also
pointless to not have two really good
thumbnails like AB testing is not to
have one really good thumbnail that
you're like this is awesome and then
just you know another thumbnail to throw
in there no matter how different they
are because you have to think about it
that thumbnail is still going to be
surfaced to people and you will lose
that click if it's not a good thumbnail
just because it's adding you know it's
making the better thumbnail look better
doesn't mean the worst option of the two
is not still being shown to people and
potentially losing that click cuz
they're not going to get this video
resurfaced you know most likely with the
other thumbnail than shown yeah this
actually came up a couple of weeks ago
and I had to ask V Richie about like
what happens if the losing thumbnail the
people who did not click it uh will it
get resurfaced and after the thumbnail
is decided on on the winter and the
answer is kind of no not necessarily
like it could if it eventually does well
it could get resurfaced to those people
who didn't click on in the first place
but it doesn't automatically get
retested against those people despite
the fact that the newer thumbnail might
be better so does that mean you should
not use it no I mean I think there's
instances where you might have two
really good thumbnail ideas and you're
just not sure which one to go with
that's a great opportunity for it do you
need to use it on every single video no
and especially if you're a smaller
Creator a lot of times you'll get
inconclusive results because you won't
have enough Impressions it takes a lot
of impressions for the process to even
figure out which is quote better and the
other thing is thumbnail testing doesn't
end the first 48 or 72 hours after you
do it the reality is you get different
people coming to your channel over the
course of weeks and months Jen uh speak
to that if you want to really really get
into AB testing then you need to have
like three AB tests like set up you
should have that initial Community test
set up within those first like three
days you should have it set again for
when your casual audience the rest of
your subscribers are seeing it within
that you know the rest of the week and
then a highly highly disc able option
past that time of course this means the
video has to be getting views that
entire time and surfacing for that to
actually make sense so we're probably
talking about this with um larger
channels not necessarily a strategy for
smaller Channels cuz that first one
probably still won't even be done but
when we think about what makes a good
discoverable thumbnail versus what makes
a good Community thumbnail and when that
video is being shown to both those
thumbnails look entirely different so if
you have a thumbnail that has your face
on it and then you have a thumbnail that
doesn't within those first three days
most likely the thumbnail with your face
on is going to win because your audience
knows you and they're reacting to that
thumbnail but that doesn't necessarily
give us the answer that we maybe wanted
for the you know longevity of the video
if you will I love that that was that
was really well explained and and again
nothing is definitive just because the
thumbnail works today doesn't mean it's
going to work tomorrow to Jen's point it
doesn't mean even that the thumbnail
that YouTube tells you is the best one
is actually the best one especially when
you're thinking about the videos live on
YouTube potentially forever majority of
people who ever see that thumbnail are
actually would maybe potentially connect
more with the thumbnail you didn't end
up even using which I think is probably
the case more often than not but if you
like questions like that and getting
your answers from us Jen and Travis here
at the Boost check the link in the
description where you can check out the
entire podcast and some other answers
we've given see you next time