ALL YouTube keyword research tools are broken - here’s why
Written by: Dexxter ClarkI’ve used Tubebuddy for YouTube keyword research and YouTube SEO since 2017.
But when I started this channel with 0 watch time and 0 videos, and the tool give me very positive results on very competitive search terms, so I got a bit suspicious.
Let’s take a look at this example:
[100% mr beast]
You think that I would be able to rank for this search term with the competition of one of the biggest youtubers on the platform?
Or this example:
[“how to get more youtube views 2021”]
Really? Serious? a channel with 2000 subscribers against 2M subscribers…
You might think that only TubeBuddy is off.
Well, vidiq thinks I can’t even rank for my own channel name.
[sound; kedeng, bats]
So I went on a quest to find out what is going on here.
[b-roll: bossche broek, stok, hand horizontaal boven ogen, stoer in verte kijken]
[kijken of je iemand kan vinden die je stoort en aan de kant moet gaan]
[voeten filmen, dan gezicht]
[sound fx: “hero theme”]
you don’t use your gut feeling
The pitfall with using a keyword research tool, you blindly trust the tool and stop using your gut feeling.
You become lazy and stop doing actual manual keyword research, don’t think with your grey matter upstairs.
It’s that self-driving electric car that you don’t pay attention to the road anymore.
We’ve all seen the headlines of how well that turned out.
Lack of data
“Dexxter, you are talking smack about my my favorite tool.
It is not bad because I pay 50 bucks a month for it.
If you think that you can do it better, why don’t you?”
That is exactly what I did.
For the research of this video I programmed a proof of concept for a keyword research tool.
Within the matter of a day I had a tool that actually works better that what is already out there.
I will show you the prototype in a minute.
During programming I bumped into some issues, which could explain why TubeBuddy and VidIQ are so inaccurate, and why they often contradict each others scores.
All tools, including mine, use the YouTube API, the Application Programming Interface that Google make available to everybody.
You give the API a search query and it returns search results.
However this API doesn’t tell you everything you need to know, ad that’s a problem.
To determine if you should make a video you want to have 3 things
• Low competition, the smaller the channels are in the search results compared to yours, the better.
The more watch time they got on their channel, the less likely you are to beat them
• 12A high search volume, to know if there is a lot of demand for a topic.
Low competition, low search volume will only be useful for a new channel
High search volume and high compeition is not desireable.
• relevant search results to the search query.
The channel size is not a problem/
But Search volume is, the API doesn’t give search volume.
So the only thing you can do is estimate it, based on nothing tangible.
Which I think is why Tubebuddy and VidIQ often contradict.
It is extremely hard to determine relevancy.
We as humans know that boat, barge and ship basically mean the same floating device, but a computer does not.
Keyword research tools, including mine, blindly trust the returned search results to be relevant to the search query, while they may not be the case.
This gives a couple of problems:
• 16. First, If YouTube cant find results with “ship”, it will return results with “barge”, if that happens it could be an amazing search term, since nobody made a video about ships.
Or it can be that viewers that searched for “ship” have another search intent, namely “barge”, YouTube’s algorithm figured that out based on viewer behavior, tools can’t.
But if “ship” returns “barge” with only big channels, you get a bad score.
• 17 Second, YouTube returns results based on what this particular viewer is most likely to watch, which can be irrelevant to the search term.
• 18. Third, the amount of returned search results says nothing (in contrast to what tubebuddy claims), since YouTube ranks by most-relevant-to-viewer.
If out of 4 millions search results, only the first 3 results are be relevant, and the rest are big channels, you get a bad score.
• 19. Fourth, often hindi spoken videos show up in the search results, which are not a competition for an English spoken channel.
But also this can give you a bad score, when the Hindi channels are big.
• 20. Fifth, when you do manual research and look at the search results and can see which kind of audience it attracts.
When I see a lot of spammy search results or a lot of Hindi videos, I know that this isn’t my target audience I want to attract to my channel.
[broll: search term “ ”].
But with a tool you can’t see that
Let me give you are rundown of my tool:
You give in the search query.
I give the results with some extra data on the channels, so you can see with your own eyes what the score is based on to tackle the relevancy thing, so it’s never a guess.
When you encounter a channel that has exactly the same amount of subscribers and total amount of views, you have 50/50 (=50%) chance of beating that channel.
Then I give this search query a 10 out 10, because that is a very good chance of winning the battle with the competition, since newness is also a ranking factor.
I give a chance prediction for every video, so you can see where the total score is based on.
The search term “Mr beast” is accurate
Also “how to get more youtube views 2021” gives a good result
But it also struggles with my own channel name, but at least: now I can see why.
And I have some ideas to improve the accuracy even further.
But despite my best efforts, my keyword tool would never give 100% accurate results, simply because I lack the same data just like every other tool out there.
So, why do keep big youtubers promoting these tools when they clearly have their flaws?
It’s either that they’ve never tested the scores of the tools with real life search results or they don’t care.
The only thing I could think of is a financial motive.
Sure enough, TubeBuddy gives a 50% recurring affiliate commission.
The smallest plan is 10,- a month.
So when you send 100 people to tubebuddy for 5 bucks a month is 500 bucks for free every month.
But there are even more expensive plans, so this estimate is quite conservative.
api data already avalible
There is another thing that I found remarkable.
The YouTube API gives exactly the same data as what is already publicly available on YouTube.
In other words: no tool in the world can give you information that you can’t find on YouTube.com
Focus on the wrong things
What annoys me is that TubeBuddy and VidIQ focus on the wrong things.
Tags for example, the YouTube tags tools in TubeBuddy and VidIQ are so prominently integrated, a beginner might get the idea that they actually do something.
If this is new information, check out my video on the topic in the description.
Another example: SEO scores, which are totally bogus.
That SEO score only exists in the TubeBuddy realm, not in the YouTube algorithm realm.
YouTube looks at the relevance to what viewers like to watch, even if the search term is slightly unrelated.
There are also features in the tools that are in YouTube Studio by default for free.
Like replacing text in descriptions of multiple videos.
any positive side to keyword research tools
So where are those YouTube tools good at?
There are other features in those tools that could be useful.
Some features that could be useful:
If you like to know the total amount of views on the channel of the competition with 1500 videos.
Technically, that data is availble and you can whip out that calculator, but a tools has done this much quicker.
Showing a graph of the views of the last 30 days of the competition, is something that a tool could do for you.
Or showing a graph of view velocity on the best performing videos on a channel.
Useful, but not something you can’t do with a spreadsheet.
Showing your thumbnail in VidIQ between other videos to see if the video sticks out, is something you can do in photoshop, but the tool can also do this.
Also showing the thumbnail when you’ve already clicked on a video is handy.
A/B testing a title and thumbnail, again not something you can;t do by hand with a spreadsheet, but a tool can do that for you.
Making a list of your best performing videos within a click to see a pattern, again, not something you can’t do in YouTube Studio, but can be convenient.
I think of all the features of VidIQ and Tubebuddy, the most useful one: generating ideas that you might not have thought of.
These can be keywords, titles or talking points for making a new video.
A lot of tools like Tubebuddy still use the auto suggest feature of YouTube, again not something you’ll miss without the tool.
36. Use the tool against you:
I know that a lot of creators blindly follow the tools without thinking.
When Tubebuddy says not to go after a certain keyword, I still do it because manual research is much more reliable.
I’m literally using the tool against you when TubeBuddy gives a bad score.
TITLE IN EDIT:
What is your favorite feature and which tool do you use?
Let me know in the comments
Can do without
I’ve managed to get from 0 to 1000 subscribers within 5 months, without the advice of the keyword research tools.
And this is the most competitive niche on YouTube.
I’ve started focussing more on value for viewers and returning viewers to my content, and I’ve managed to double my view to subscriber ratio compared to the time I used TubeBuddy.
Share this articleIf you took value from this article, please share it on social media.
Music Producer / YouTuber
Read more about the author
Music Producer / YouTuber
Read more about the author
Read also ...
Continue reading >>
Continue reading >>
Continue reading >>