Do you know what a datapoint is and why it is important to authors? No, neither did we until a few weeks ago. But knowing about them may explain why some social media advertising is less than effective. Essentially a datapoint is one tiny bit of data that an algorithm can use to connect it to another datapoint. Think about it the way you would think about a pattern matching game such as “Mahjong” (the computer game, not the ancient Chinese tabletop game). If you find tile A in Mahjong and then find a second copy of tile A, you click on the two of them and you remove them from the puzzle, at the same time as winning points. Every time you do something on social media, you create datapoints which the platform then uses to direct advertising towards you. You are creating multiple sets of Mahjong tiles which the algorithm will use to match the same tiles that advertisers have used. Every time you “like” a post; every time you comment on a post; every time you share a post, you create a datapoint for the platform to work with. Even the words you use in your posts and comments become datapoints. For example, a friend posts an image of a donut they are about to eat. If you “like” the post, you tell the website’s algorithm you that you like donuts. If you comment on the image, you create more datapoints related to donuts. It should, therefore, come as no surprise if you start to see advertisements for donut brands or donut shops on your timeline shortly afterwards, because the algorithm now knows that you have an interest in donuts, and it can make money for the platform by targeting you with donut related adverts. The reality may be that you “liked” the post because you like the friend, not the donut (you may actually hate donuts), but the algorithm doesn’t know that. It has made up its mind that you like donuts so that is that. Knowing this can help you when you use social media for advertising, because the platform wants to know which users to send your advert to so that they see it. Because when they see it they may click on it and then the platform can charge you for that click. So sending the ad to the right people is not only important for you, it is important for the platform. So, this is something you need to know about if you want to use social media efficiently for advertising. For example, if you pay for an ad on Facebook, its advertising algorithm goes looking for the right people to whom it should show your ad. It could show it to everyone on the platform, but that isn’t efficient because not everyone will interact with your ad, so it won’t earn Facebook any money. First it scans all the content in your ad (the image, the text, the link etc) and converts it into data points. Then it goes looking for users to match your ad up with. To extend the Mahjong analogy, if your ad is tile A, the algorithm looks for users who also have tile A. How does it know? Your book is the equivalent to the donut we used in our analogy. If your ad includes references to fantasy, the algorithm goes looking for users who have fantasy related datapoints recorded against them. But the algorithm won’t send the ad to people who have “romance” in their datapoints. We will use Facebook (Meta) as an example, but all the social media advertising algorithms work the same way. Your ad for your book works like the ad for the donut brand or donut shop. It looks for people who (a) like books, (b) like books in your genre and (c) anyone who has ever interacted with posts that use any of the words in your ad and who therefore might be interested in your book. Point C in the para above is the most important if you advertise on social media, because it can make the difference between your ad being seen by the right readers and it not being seen at all. But success breeds success for the algorithm. When it finds a user who clicks on the link in your ad, it then uses that user's other datapoints to find more users like them. It "learns" from its successes to improve its targeting of users. Typically, authors use the blurb for their books as the primary text for their social media ads. That’s fine, it makes sense to do that. However, it can also cause problems. A good blurb will be between 250 and 350 words long. Take out all the common words: prepositions, pronouns, conjunctions, adjectives, adverbs etc which are useless as datapoints (because they are so common) and you are probably left with less than 100 datapoints (proper nouns, nouns and verbs mainly).which can be used to connect your ad to the right audience. Let’s say you write fantasy or sci-fi and you use a lot of made up names for characters and locations in your books. Now let’s say that you use some of your made up words in your blurb, which forms the text for your ads. They are useless as datapoints because they won’t appear as datapoints that Facebook users have reacted to in the past. They have never heard of the land (or planet) of Gigariga and so Facebook’s algorithm won’t be able to find any datapoints that relate to Gigariga. Which means that using your blurb may have reduced your potential number of datapoints from around 100 to a much smaller number, which means that Facebook will show the ad to fewer people. It may even show the ad to the wrong people, the ones who would never think of buying your book. The lesson to be taken from this, therefore, is the same one that applies to using Keywords on Amazon. Make sure you use the right datapoints. This may mean re-writing your blurb. If your book is a fantasy, then you have to use that word and as many synonyms for it as possible in order to maximise the number of fantasy readers who will see your ad. The same applies to other datapoints, such as action, adventure, mystery, challenge and thousands more words that you might use. You don’t want to be too general in your wording, so that your ad is shown to readers who never read fantasy, but at the same time your datapoints mustn’t be so narrow (like using your made up words) that Facebook can’t find anyone to which it can show your ad. Just copying and pasting your blurb into your Facebook ad may seem to be quick and easy, but it may also be sabotaging your efforts and causing you to waste money. So, think about those datapoints and make sure they are present in your advertising copy. That is what professional copywriters do every day; which is why their advertising works. If you have enjoyed this blog, or found it informative, then make sure you don’t miss future editions. Just click on the button below to sign up for our newsletter. We’ll even send you a free ebook for doing so.
0 Comments
Leave a Reply. |
AuthorThis blog is compiled and curated by the Selfishgenie publishing team. Archives
November 2024
|