I’m puzzling over metadata — apparently this is the data that helps search engines find books to suggest when folks like you and me enter keywords into Google, Amazon and other online services.
UNRAVELLED — due out this September — is a story about Ann and Edward Jamieson that begins in 1935 with flashbacks to WWI and ends in 1944. Part love story, part war story, Unravelled explores the consequences of two world wars on one couple’s marriage. The tag line is Two Wars. Two Affairs. One marriage.
I believe it’s a great read, the kind you might get absorbed with on a lazy weekend or take along on holidays or suggest to your book club. Both women and men have enjoyed it.
So the question I’m considering is how readers will find it when searching. On Amazon, I tried the following searches:
- WWI fiction – 294 books (a smallish number is good). But an emphasis on war is usually a turnoff for women readers.
- WWII fiction – 1704 books. Same problem as above.
- Historical fiction romance – over 54,000 books. And really, Unravelled is much more than a romance.
- Book club recommendations – 137 books. Small number, but who really searches like that?
- Great summer reads – 345 books. Not bad, but summer is over in this part of the world by launch time.
- Fiction for women – more than 118,000. Impossible to get noticed in that crowded scene.
- War romance novels – 882 books.
So, when you’re looking for a good read, what do you search for?
14 Responses
I searched the Kindle store for “historical fiction Louisville KY” which fits the novel I am writing. Only one book came up, which is good for me, I guess, (six books came up in all ‘Books’), but I wonder how many people would search by city?
Carmel, the tag words “historical fiction Louisville KY” are a dead end, with only 5 results. You won’t generate much business using those words. You would be better off switching to “Kentucky History” which got 10,130 results, or “Southern Fiction” which got 26,963 results. Just using “The South” got an amazing 995,469 results!
Well, I just search Ancient Rome or Roman Legions. Or the author i know like Simon Scarrow and check his books out then note the ones that Amazon puts up that are ‘lilke’ his.
There is a book outHow to sell Fiction on Kindle by Michael Alvear that is really helpful with this. Besides that. This marketing monster is killing me.
Marketing monster is a good term, Judy. Let us know if you have any good techniques that work!
Reblogged this on Dreamin'.
I am so looking forward to your novel. I personally think search terms are more generic than writers imagine. Thus those big numbers you cited like “historical fiction romance.” But if your novel has elements of a romance/love story it might be wise to include that in at least one of the two choices you get to categorize your book.
It’s really informative to look at novels that might compare to yours and note what choices are made. (You probably already know this. ) And not to forget, you can use up to five tags to further clarify your novel.
Who would have thought writing would morph into this second , and necessary, preoccupation? And we all thought it was just about the writing.
My very best wishes
Thanks for the encouragement, Judith. These days I feel like one of the Hindu gods who has many arms!
Tags are a form of metadata. However, the hardcore metadata that search engines regularly harvest are taken from a list hidden in the of every webpage. The better website design programs like Weebly provide easy access to edit metadata. Ask for help from your Internet Service Provider if you can’t locate this feature for your website. A Google search about “metadata” will provide even more information. Good luck
Jack -Very interesting. I feel I need to know a lot more about meta data, and it’s elusive (at least for me) I have done similar searches as Mary has and still feel just as at sea with the categories.
Many thanks for the information, Jack. Where on earth do folks come up with names like Weebly?
That’s not how it works. The numbers you see are the number of times people have searched for that item using the Amazon’s search engine and BOUGHT the item. It is not the number of books they have under that term. Ideally, you would want to use tag words (or search words) that have over 50,000 searches meaning people who have searched for those terms have spent money on those items. Then you take those terms and incorporate them into your a) novel’s product description and b) tag words when you input the novel into Createspace and Kindle Direct Publishing. For instance, the term “historical fiction romance” yields 54,803 search results (pretty good) but the term “book club recommendations” yields only 142 results. That means that in the last month, only 142 people searched under “book club recommendations” and bought, but over 54,000 people searched under “historical fiction romance” and bought those items.
Here are some other results I generated using a variety of other terms:
The term “war” alone yielded 974,230 search results.
The term “espionage” alone yielded 67,385 searches.
The term “war action” yielded 72,965 searches.
The term “action adventure books” yielded 160,579 searches.
The term “lost love” yielded 193,603 searches.
The term “love and romance” yielded 142,834 searches.
The term “literary fiction” yields an astounding 398,912 searches!
Watch this great video for a tutorial:
http://www.youtube.com/watch?v=Do5IGzhIvdM
Thanks for the clarification, Rachel.
Mary,
Metadata is one of the factors that search engines uses to find documents. Metadata is data about data.
What this means is when you buy a can of peaches the label on the can is metadata. It tells you that it is a can of peaches, the type of peaches, the number of calories in the can, where it produced and by whom. Without the label you would not be able to tell whether you bought a can of peaches or a can of beans.
Metadata is the labelling you apply to web pages such as the title and description.
For books some of the metadata labelling is the title, author, publisher, price and subject categories.
Metadata is a generic term as there are various standards used for various industries. Onix is the standard used in the book trade.
The way that a web search engine works is that it visits a website it crawls through the site and looks at the text in the web page both the visible text and the metadata fields in html code and places the text in a database called an index.
When you input a search term in a search engine. The search engine searches its index and says I found your search term on this web page on this web site and delivers back to you the search results in a rank order.
The way that search engines calculates the rank order is a closely guard secret. Very roughly it assigns points to the following and calculates a ranking score:
– title metadata field,
– the description metadata field,
– text in the HI header on the page,
– where the search term is found, higher on the page scores better
– links to the page
There are other factors in involved. For example, if you search Google.ca and Google.com you will get different results because the results are geared to different audiences.
These principles apply to most search engines.
I hopes this helps.