This has nothing to do with advertising but should be mentioned anyway.
Once your app has been launched it is very hard to determine why people are not converting. You may have a bunch of downloads but nobody is buying the content. What is going on? It is essential you put some metrics in your app. Some way of determining how far users get and where they give up. For instance; The first versions of Learnia had people dropping out almost straight away. In order to figure out what was going on we let the app periodically write a file to a server (anonymously). The file contained the exercises completed, the tutorial elements completed, the time played, the average frame rate and a few other things. Every time the app stored the file it would overwrite the one already there. We were now able to determine how far people got before they dropped out. Learnia has 2 paths with skillstones for kids to master. Both paths start at a central island. It turned out most people just started to work from the island attacking the first skillstones they encountered. These skillstones were way too easy for most kids (SK level). They did not want to work through several years of math before they got to their own level. The tutorial explained that they could just skip to their own level (the skillstones don't all have to be mastered) but I guess most people did not listen to the tutorial. To remedy this we added an introduction sequence. Basically a fly-by over the Land of Learnia leaving the student at the correct level for him/her. To our surprise this did not help much. Thanks to our metrics we could see kids still started at the beginning and not at their own level. Why? We looked at the download times and realized parents tend to do the downloads when the kids are at school or in bed. Parents got to see the nice fly-by. The following day they would give kids the app but the kids had missed the fly-by and so they started at the skillstones that were too easy for them. Now we changed it so that the introduction plays every time you start the app unless you completed 10 questions. We would never have figured this out without detailed metrics.
2 Comments
Let's say you spend $10 on advertising. If you have set everything up correctly. Your ads look good, your app look enticing and everything is set up correctly you may see something like this:
100 people click on your ad. 80 people actually arrive at the store page of your product. 5 people actually download your app. Why is it not all people who click on my ad arrive at the store page? We think some people give up if they have to way too long for your store page to load. On iOs there is another problem. People clicking your link are presented with a window. The window says something like "Open this page in App Store? Cancel/Open". Not everybody will say Open. Hopefully Apple will remove this step at some point as it seems pointless. Instead of pointing your ad to the store page you could point to a web-site first and let people click a link to the store from there. We tried this and for us the results were similar. The number of people who actually download an app after visiting the store page is (for educational apps) around 5%. It obviously depends on your screenshots, description, price of the app and a million other things. The average is around 5% for educational apps but it will be different for games, utilities etc. In our example every download costs us about $2 (and in actual fact you are doing quite well with $2) so it only makes sense to advertise if we make at least $2 per download on average. With Learnia we hope to reach this as parents buy content per year per child. Most parents will end up not spending anything but some may end up spending $40 or more. As we are currently (May 2018) opening up all content for free it is hard to say whether we are above water on our advertising. We thought parents looking for apps for their kids would want to see a review or 2 before committing to downloading. There are a number of good websites around that review educational apps. Here is a short list of the sites we think are worth pursuing:
askatechteacher.com theimum.com funeducationalapps.com bestappsforkids.com appedreview.com smartappsforkids.com commonsensemedia.org teacherswithapps.com edsurge.com geekswithjuniors.com edshelf.com techwithkids.com educationalappstore.com There are a lot of small web-sites that will do a quick review for $25-$50 and these reviews may be ok but generally don't go in depth and are not really that trust-worthy. You should look for sites that use real parents or even teachers to do reviews. As these sites get sent hundreds of apps each week and they pay their reviewers they have no choice but to charge the developer for their reviews. This does not mean they will guarantee a good review. You just have to pay for their time. We did notice they often hide the score if it not favorable. (They may display 3, 4 and 5 stars but simply omit 1 and 2). The typically charge between $50 and $500. In case you are interested; these are the reviews we paid for. www.educationalappstore.com/app/learnia www.bestappsforkids.com/2018/learnia/ www.theimum.com/2018/03/the-land-of-learnia-review/ Were these reviews worth it? We're not sure. Whenever a review appeared we got a small spike in downloads but only 10 or so. The reviews do appear when you do a search in google so it does help that way. As it turns out most parents don't check reviews and simply download an app as long as the store page looks good. When deciding which reviews to go for or where to advertise it is worth checking the number of monthly hits a web-site receives. You can do that here: www.alexa.com/siteinfo Make sure to check where traffic is coming from. Some sites have 90% traffic from India. Presumably those hits are from bots and not real people. Google adwords display ads will accept ads in a variety of formats. We decided to produce ads in the formats 728x90, 468x60, 300x250 and 200x200. In practice the algorithm almost exclusively used the 728x90. This must have had the best results or be the cheapest to place. It is definitely worth producing an animated gif like our 728x90 below. You produce a number of images and use a gif-maker to combine them into one animated gif. We used https://ezgif.com/maker . It is nothing too fancy but it does the job and it is free.
Your gif has be less that 150k in size. The gif-maker will let you compress the gif if you have gone over the 150k. With a slider you can control the quality vs size trade-off. In addition to the 'fixed resolution ads', adwords also uses 'responsive ads'. You supply an image, a logo and some header and description text. Adwords will generate an ad and arrange there elements based on the advertising space. My responsive ad generated clicks cheaper than the fixed resolution ads so it is well worth the effort setting this up. After our disappointing results from advertising directly on web sites we decided to give google adwords a go. Google adwords has 3 components. Search terms. This is where you bid on search terms so that you come up when people do a google search. We had a play with this but could not get this to work for us. We think the competition for search terms having to do with educational apps are too sought after and therefor too expensive. A problem is also that many people are searching on a pc/mac and are less likely to be interested in our iPad/Android app. You-tube campaign. This seemed to attract a fair number of people watching our video but very few of them were engaging. We abandoned this approach. Maybe people just didn't like our video: https://www.youtube.com/watch?v=F8F63YZ2hPg Display ads. These are ads that are displayed on a wide array of web-sites as well as on you-tube. Display ads are a fair bit of work to set up but you have a lot of control over their placement. You can specify the age of your target group, interests, gender, income bracket, parent, web-sites they may like. You can specify the countries they are from or have in interest in. You can also specify which device they are on. We limited our campaign to Android and the latest models of the iPad. Once you have set everything up the algorithm will start figuring out where your ads work best. This is why our initial CPC (cost per click) was around 0.50 but after a few days it went down to around $0.13. We decided to initially spend $20/day. This is enough to get around 150 clicks which is enough to figure out what other bottlenecks there are further down the line. A typical day for a $20 budget on display ads.
Our first idea was to advertise on web-sites directly. We used buysellads.com. It lets you pick from a large number of web-sites. It will then show you the available slots and deals. Typically the sites would charge an amount around $50-$100 for a particular slot on their site for a month. Sometimes they would charge for a fixed number of impressions (ie 30,000) and sometimes your ad would be in a carrousel with others. Everything is very clearly lined out and straightforward. We picked some sites that deal with educational apps so their visitors would be interested in Learnia. As you can see below the results were very disappointing. The CPC (The cost for someone to click on our ad and follow the link to our site) was in all cases well above $2. Others may have better results but this approach did not work for us. We also advertised on a couple sites not affiliated with buysellads but the results were similar.
There was no buzz around 'the Land of Learnia' when it was first released. We were 100% focussed on development and simply did not have time to spend on social media. A lot of people will tell you you should market well before the release but I don't think that works for educational products. There is so much competition out there it is very difficult to engage people without a working product.
Maybe pre-release campaigns can be successful for games where a screenshot or 2 can give a reasonable impression of the game but educational apps are often a bit more complicated. When we released Learnia we expected at least some organic downloads but we got absolutely none in the first week. There are so many educational Apps out there (> 500K) that it was impossible to get in the top 100 of any search term initially. The reason is that both the App Store and Google Play organize the results based on number of downloads and on the reviews. If you don't have any downloads and reviews you tend not to show up. So how do you get started? There are thousands of excellent apps teaching a particular skill. For instance Addition, Subtraction using borrowing, Multiplication tables or Clock Reading. These apps typically cost 99c or $1.99 and can be great filling a certain skill gap.
Parents and teachers can end up buying a large number of these apps. I believe the model of a library of 99c apps only works if a parent or teacher is closely watching and encouraging kids to progress to more challenging apps. For teachers/parents it is a huge challenge finding the next suitable app that connects to the app a child has just finished with. If the classroom is quiet during app hour and the kids are not asking any questions then that is an indication that there is not a lot of learning going on. Kids are most likely doing apps that are below their skill level. Kids who are learning typically have questions. A better approach would be to have a single app containing many different skills placed in the right order and following a curriculum. This way kids can smoothly progress to the next skill, once they have mastered the old one. By remaining in the same app, kids will continue with the same reward structure (coins, game time etc). By staying in the same app, the user interface remains constant so that no time is wasted learning new buttons. Apps covering multiple years of learning have been around for a long term but their graphics are out of date and the variety is lacking. IXL is a good example of a complete package that can be boring and result in grinding. Students work through many similar questions without much variation or challenge. Learnia is an attempt to combine the best of both worlds. Each lesson is interesting and uses colorful 3D graphics. Children can often move objects around with their fingers and use handwriting to give the answer. Each exercise is different and usually presents 3 questions of a certain type before moving on to another type of question. Each skill is presented as a stone that students can master. Once they have mastered a skill they can try the next one along the road. This is a low threshold way of progressing and encourages kids to move on when it is time to do so. It is hard for children to focus and be productive for a long time. When learning math we would recommend using an app for 30 mins per day as a maximum. Any longer and kids will gravitate towards less challenging / less taxing activities. |
ObbeObbe used to write games and is now bringing these skills to educational apps. ArchivesCategories |