Using niche keywords

[Lysmata Debelius]

I always struggle with what keywords to use when uploading a new book to Amazon. I have looked at the recommendations about which keywords get you into which categories. Is it a bad idea to use relatively lesser known words like "Biopunk" as keywords ? How might that help? Is it only useful if someone actually searches for the term? Do people search for books on Amazon using keywords?

[AmHere]

People search on Amazon all the time. The words they use in the search box are our keywords. If you go to Amazon, select Books in the search drop down and slowly start typing it will populate with the most popular search terms with the characters you've typed in. 'Biopunk' shows up when I typed in 'biop'. So it looks like folks are using it and it may help you to use that as a keyword.

Edited To Add: Choose "kindle store" in the search drop down for eBooks. 'Biopunk" still shows up when I type in 'biop'.

[Lysmata Debelius]

People search on Amazon all the time. The words they use in the search box are our keywords. If you go to Amazon, select Books in the search drop down and slowly start typing it will populate with the most popular search terms with the characters you've typed in. 'Biopunk' shows up when I typed in 'biop'. So it looks like folks are using it and it may help you to use that as a keyword.

Brilliant. Thanks. I still have a lot to learn about keywords.

[Simon Haynes]

I bought KDP rocket and immediately discovered every single keyword I'd been using on all of my 30+ published titles for the past X years were completely useless.

Not that I've noticed any difference in sales since applying hundreds of NEW keywords, but it sure felt like I was doing something proactive and useful ;-)

[AmHere]

I bought KDP rocket and immediately discovered every single keyword I'd been using on all of my 30+ published titles for the past X years were completely useless.

Why did you think your original keywords were useless?

[CoraBuhlert]

I do use niche keywords like specific subgenres (dieselpunk, cozy space opera, etc...) that Amazon has no categories for and mix them with the keywords recommended to get into certain categories. Not sure if it helps though.

[NathanBurrows]

Related but slightly off topic, but is there any guidance or advice on using single keywords in each of the 7 fields versus multiple words or phrases? I’ve looked but not been able to find any useful advise.

[Lysmata Debelius]

Related but slightly off topic, but is there any guidance or advice on using single keywords in each of the 7 fields versus multiple words or phrases? I’ve looked but not been able to find any useful advise.

I would love to know this too. I've seen articles where people say that phrases or multiple words are more effective ("strong female character") but they didn't provide any evidence, and the rest of the advice was so dubious that I don't know whether or not to trust them.

[Tom Wood]

Related but slightly off topic, but is there any guidance or advice on using single keywords in each of the 7 fields versus multiple words or phrases? I’ve looked but not been able to find any useful advise.

I would love to know this too. I've seen articles where people say that phrases or multiple words are more effective ("strong female character") but they didn't provide any evidence, and the rest of the advice was so dubious that I don't know whether or not to trust them.

There is a discussion of this very topic at David Gaughran's blog here:

https://davidgaughran.com/2018/07/21/amazon-category-hacks-kindle-store-hack-keywords-categories/

Scroll down where I weigh in, and David confirms that you can/should use keyword phrases:

You can put as many seperate words in each of the seven keywords slots as you like, up to a maximum of, I think, 50 characters. So you could, for example, do “young adult epic dragon fantasy” as one of your seven keywords (ditch the commas as Tom suggests).

And yes, as Tom also suggests, that will count as a search trigger for all variations of that, so you should appear – somewhere – for “young adult” “dragon” “epic fantasy” and “young adult epic fantasy.”

When discussing 'keywords' note too that there are the seven category keywords, and then AMS ad placement keywords. Those are two different beasts.

[AmHere]

Tom thanks for that link.

Another one I found interesting and am testing

https://www.thebookdesigner.com/2018/09/words-gone-wild-kdp-keywords-revisited/

[Lysmata Debelius]

Thanks!
This is very useful.

[YouMeWe]

People search on Amazon all the time. The words they use in the search box are our keywords. If you go to Amazon, select Books in the search drop down and slowly start typing it will populate with the most popular search terms with the characters you've typed in. 'Biopunk' shows up when I typed in 'biop'. So it looks like folks are using it and it may help you to use that as a keyword.

Brilliant. Thanks. I still have a lot to learn about keywords.

Make sure to do your Amazon keyword research in private/incognito mode to save the results being corrupted by your search history.

[YouMeWe]

Related but slightly off topic, but is there any guidance or advice on using single keywords in each of the 7 fields versus multiple words or phrases? I’ve looked but not been able to find any useful advise.

Not guidance or advice, but I did try it once. A four figure book rank promptly dropped to digits longer than my telephone number. 

[AmHere]

Not guidance or advice, but I did try it once. A four figure book rank promptly dropped to digits longer than my telephone number.

Just to make sure I understand, are you saying the following?
- originally each of the 7 fields contained one word each
- this was changed to multiple words i.e. phrases
- this change caused the book's rank to go from 4 digit to N digits where N > 4

p.s. I am wondering if there was anything else that changed that could have caused the change in rank.

[sstreet]

I've used KDP Rocket as well to discover if particular keywords are useful (meaning they are actually typed in by people to look for items), the problem I'm always having is that if people aren't using the ones I think they are (which I discover by using a tool like KDP Rocket) then what in the world keywords are they actually using? I guess I am not creative enough to understand what people think when they are trying to find a book to read. But keywords are a huge stumbling block for me, because everything I read says put them in your description, title, ads, 7 keywords when you publish, etc. It's easy to find the ones that don't work, but finding new ones that actually do? Not so much.

[YouMeWe]

Not guidance or advice, but I did try it once. A four figure book rank promptly dropped to digits longer than my telephone number.

Just to make sure I understand, are you saying the following?
- originally each of the 7 fields contained one word each
- this was changed to multiple words i.e. phrases
- this change caused the book's rank to go from 4 digit to N digits where N > 4

p.s. I am wondering if there was anything else that changed that could have caused the change in rank.

Oh God, no. That's completely ass about face. 

Originally, each box was filled with a long tail keyword. For example: Box 1 = murder in a Las Vegas toilet, Box 2 = kidnapping an American gangster dwarf,  etc etc.

I changed that to fill each box with one word. For example: Box 1 = toilet, Box 2 = dwarf, Box 3 = gangster, etc etc.

This change is what caused the rank to plummet.
Reinstating the original 'long tail keywords' recovered the ranks.

Apologies for not making that clearer.

(For the avoidance of doubt, those aren't my keywords. Not that there's anything wrong with them, they probably work for someone ...)

EDITED TO ADD: When I say 'plummet' I mean in a bad way. Like, the starting rank, with full boxes, was at 3,000. The single word in a box, dropped the rank to 100,000.

[Lysmata Debelius]

Thanks for clarifying that. I had misunderstood.

[AmHere]

Oh God, no. That's completely ass about face. 

Ha, thanks for clarifying. I misunderstood

[CoraBuhlert]

Not guidance or advice, but I did try it once. A four figure book rank promptly dropped to digits longer than my telephone number.

Just to make sure I understand, are you saying the following?
- originally each of the 7 fields contained one word each
- this was changed to multiple words i.e. phrases
- this change caused the book's rank to go from 4 digit to N digits where N > 4

p.s. I am wondering if there was anything else that changed that could have caused the change in rank.

Oh God, no. That's completely ass about face. 

Originally, each box was filled with a long tail keyword. For example: Box 1 = murder in a Las Vegas toilet, Box 2 = kidnapping an American gangster dwarf,  etc etc.

I changed that to fill each box with one word. For example: Box 1 = toilet, Box 2 = dwarf, Box 3 = gangster, etc etc.

This change is what caused the rank to plummet.
Reinstating the original 'long tail keywords' recovered the ranks.

Apologies for not making that clearer.

(For the avoidance of doubt, those aren't my keywords. Not that there's anything wrong with them, they probably work for someone ...)

EDITED TO ADD: When I say 'plummet' I mean in a bad way. Like, the starting rank, with full boxes, was at 3,000. The single word in a box, dropped the rank to 100,000.

Well, I certainly want to read the book about a murder in a Las Vegas toilet and a kidnapped American gangster dwarf now.

[Jeff Tanyard]

Well, I certainly want to read the book about a murder in a Las Vegas toilet and a kidnapped American gangster dwarf now.

Lol... I was thinking the same thing.   

[R. C.]

I bought KDP rocket and immediately discovered every single keyword I'd been using on all of my 30+ published titles for the past X years were completely useless.

Not that I've noticed any difference in sales since applying hundreds of NEW keywords, but it sure felt like I was doing something proactive and useful ;-)

Hmmmm... I have KDP Rocket and have come to be highly skeptical of the results.   Data ALWAYS has value but applying the value from the KDP Rocket data seems... to me almost obtuse.   

Cheers,
Ruairí