The experience of the withdrawal of the application in the App Store Top Russian: numbers, graphics, investigations
Hello!
I always enjoy reading the case study of developers for iOs and want to make a contribution to understanding the magic of the App Store - tell our story of creating and promoting an application in the Entertainment category for the iPhone.
Some teasers:
1. Most of the large sites in Russia and Ukraine wrote about us. All analytics and conclusions are presented.
2. We have risen to 3 places in the general Top App Store of Russia and Ukraine, 1-2 place in the category. At the moment we are already the 4th day in the Top 5 with more than 60K installations and about 500 reviews / ratings on the App Store.
3. We carefully documented the whole process and made many screenshots along the way.
')
The idea was suggested by the wife - she is an active user of Vkontakte, but because of the abundance of groups and publics in the tape she began to miss the news of friends. At the same time, I did not want to turn off the news of the publics, and it was not so easy to configure Vkontakte filters.
Time and desire was - together with ptashcka they quickly made an application to which you can transfer news of interesting groups and read them there, hiding from the main VC tape. Thus, friends remain on Vkontakte, and public subscriptions can be read through our application, when convenient.
We posted it on the App Store and found that users like it. We decided to continue the development and focus on 2 things: simple and elegant design, without an abundance of unnecessary details, as well as a selection of interesting groups by categories.
Developed the theme of minimalism and visualization of information. Hidden likes and reposts under the cat, by default, we show only the first 160 characters from the text block, the interface elements are hidden, so that the maximum screen space can be given to the content. For the popular format of posts with several pictures, we developed an issuance algorithm, which we are proud of - the images are formed very harmoniously and create their own mood for each news.
For a long time they made an icon, went over a lot of options. I didn’t like everything until I thought of putting a cat on it - the most popular object of photos on the Internet:
Once again convinced that a good icon and name are extremely important - after changing them, organic plants grew 5-6 times!
They took the selection of groups seriously - they analyzed all VC groups with more than 100K participants and each gave a rating based on the main criteria: more likes, reposts, less advertising and links, unique content and a minimum of bots. They made their own analytics, which regularly monitors the quality of groups in the compilations. Including set up filtering of advertising posts in particularly annoying groups. If the habrasoobshchestvu interesting - write, make a separate post on this topic.
I want to note the importance of proper analytics within the application and setting up user funnels. Ilya Krasinsky, who led the mobile development of LinguaLeo, helped us a lot. Ilya immediately pointed to our big miscalculation - after launching the application, the user saw all the groups at once and was forced to understand them, which is why only 74% got to the tape. After we introduced a simple intermediate screen for choosing interests, this indicator increased to 85%.
Put up a funnel to activate users, the ultimate goal - the entrance to the tape and view multiple posts. We created 3 base cohorts based on frequency of use: Heavy (every day), Regular (once and more per week) and Infrequent (less than once a week). By analyzing the use of Heavy, we try to understand how the application can be improved, Infrequent - where the problems are in the interface and why users do not reach Regular. And, of course, they looked at the most important parameter - retention - how many users remain in the application and continue to use it. Flurry in this regard, in my opinion, is better than Google Analytics, because immediately calculates these metrics.
For analytics we use Flurry and Google Analytics. Flurry is good free of charge and a cohort analysis out of the box - you can instantly see how changes in the application affect conversions. Google Analytics can also be commended for the online mode and the quick update of statistics, as well as the convenient sorting of events into categories, which, unfortunately, is not in Flurry.
To manage the project, they used simple and free Trello, where they also tracked the bug.
Before launch, we optimized search words using sensortower.com, everything is quite simple and convenient. As a guide, they used this article on Habré .
After the functionality was implemented and tested, we were satisfied with the conversion and retention, and the search words gave the best results, it was decided to move to the Holy Grail of developers - Top App Store.
To reach the Top, it is necessary to have as many installations as possible in the shortest possible period of time, so our marketing activities were focused on time and aimed at 10-12K installations per day. According to the promises of luminaries, this should have been enough to get to the Top. That's how it turned out.
The plan was:
1. Reviews on thematic resources.
2. Advertising in the public VK.
3. Facebook Mobile Ads, as a backup, if necessary, quickly throw traffic.
4. Partnership with other applications and co-promotion.
As day X was chosen the 7th of September. First, the people returned from vacations and holidays, and secondly, we would have slipped in front of the news wave about Apple updates. But then, unexpectedly, Apple itself intervened.
The final version of the application was uploaded to the app on August 23rd. According to our calculations and previous experience, this should have been enough for consideration, but on the 5th the application was still in “Waiting for review” status. A request for an Expedited Review was submitted, it was confirmed, but the application was only reviewed on the 8th. Here, by the way, is a useful resource that calculates the time of consideration based on data from other developers from Twitter.
I had to renegotiate with all review sites and partners on the 14th of September. On September 9, we appeared in the App Store and, taking advantage of the resulting break, we worked on reviews and ratings: in the update we asked our users to leave a comment on the application, and also used the service for this. Judging by the latest news , ratings and users seriously affect the rating.
At first glance, planning reviews on Saturday looks illogical, when traffic drops significantly compared with weekdays, but we were guided by the considerations that fewer publications and reviews spend more time on the main pages, so more users will see them. No conclusions on the effectiveness of such an approach have yet been made, I will be happy with expert opinions in the comments.
In the end, agreed with iphones.ru, planetiphone.ru, macosworld.ru, ukranianiphone.com, macdigger.ru and tech-touch.ru. From the advantages it can be noted that on iPhones.ru a link with an icon in the right navigation bar hangs for quite a long time, while planetiphone.ru has a mobile application where they send push notifications. Asked to use our links to track campaign performance. The main idea is to count not only clicks, but also installations, as well as the quality of traffic from each individual resource. There are many good solutions for this, such as mobileapptracking.com and appsflyer.com, which are considered the most accurate, but we ended up using the free Flurry. In the User Acquisition section, Flurry can generate the required number of links and link them to your funnels and cohorts. Thus, you can see what percentage of users who have installed from each specific marketing channel converted to Heavy users.
In fact, it turned out like this:
+ UkrainianiPhone.com used its tracking links and gave a total of 684 clicks.
I do not know how correct it is to evaluate the effectiveness of clicks, because I usually take the phone myself and find the application on it without following the link. Settings on clicks and the quality of users have not yet miscalculated, lay out in the comments as soon as it will.
Android users there are on average 2 times more than iOS, so there was little point in advertising in general public. We selected a few dedicated to iOS applications and located in them. After the launch, in order to maintain traffic, we made several more posts in groups. According to the observations, groups up to 500K participants work best - there is excellent involvement, people comment, like and repost.
Sometimes even holivory were ignited, where without them:
The problem is that it was not possible to calculate the efficiency of this channel normally - many public owners basically do not post short links, apparently fearing the possibility of a substitution and a ban from the administration. An important point, check the public for the presence of bots before you buy ads. Make it simple - go to vk.com/adscreate, create an ad for users of this group and select those who come from iOS devices. On average, on Vkontakte, this indicator is 11%. If it is significantly lower (of course, if the group is general, and not the community of Android owners), it is highly likely there are a lot of bots.
Good, but rather expensive thing. When targeting Russia, one click costs $ 0.25-0.55, while the installation goes about $ 0.8-1.2, depending on the CPC. We had a few coupons for advertising, which we spent during the campaign. The most effective ad with a CTR of about 9% looked like this:
We tried to agree on co-promotion with other applications that are relevant to our target audience, but nothing really came of it. While our application was unknown and unpopular, few people wanted to go for a partnership, so they were postponed.
In general, they spent on marketing, not counting FB coupons, about 40 thousand rubles.
Started at 10-00, September 14. Here is a graph of new users on Google Analytics for the weekend of September 14-15:
After the first peak of users passed, it turned out that we didn’t get to Top, so we made a decision in addition to enable FB mobile ads to full. Totally spent coupons for about $ 300, received about 1500 clicks. App Store now recounts ratings every 3 hours, for us it looked like this (general / category, Russian App Store):
Sat, 9:00 am - 563/53
Sat, 12:00 - 29/4
Sat, 15:00 - 25/3
Sat, 18:00 - 17/3
Sat, 21:00 - 15/3
Sun, 00:00 - 13/3
Sun, 3:00 - 11/3
Sun, 12:00 - 9/3
Sun, 15:00 - 8/2
Sun, 18:00 - 7/2
Sun, 21:00 - 6/2
Sun, 00:00 - 5/2
Accordingly, conclusions can be drawn about the effect of the number of installations on the rating. After we climbed to the 15th position, it went much easier, as organics helped.
Rating:
Settings for the weekend:
During Monday and Tuesday, we tried to keep in the Top-5, because It’s exactly the five first places that users see by default when they go to the iTunes homepage. Fueled by traffic from Facebook Mobile Ads, publicity ads and several reviews that came out later than the main wave.
To test the monetization of the disconnection of advertising, an Admob / iAd ad banner was installed, which can be disabled for 99p. I’ll say right away that there are few purchases, pennies go on advertising - on Saturday, September 14, about $ 30 in total. At the time of promotion they made promotional codes, by entering them it was possible to turn off advertising for free. Unfortunately, there was a bug that did not allow activating the promo code of a part of the users, so the ads were turned off altogether so as not to upset anyone until we correct the error. Now we stopped on another version of monetization, more about it later, when there will be numbers.
Obviously, the application is not for everyone: someone grew out of VC content, someone has already formed a circle of interests, but they have their own audience and are satisfied - the average rating is 4.5 out of more than 500 left.
Thanks to everyone who wrote, it was very nice to receive your ideas and wishes. We have released the first version and are actively improving the application.
Ask questions and share your experiences in the comments!
Good luck in conquering the App Store!
I always enjoy reading the case study of developers for iOs and want to make a contribution to understanding the magic of the App Store - tell our story of creating and promoting an application in the Entertainment category for the iPhone.
Some teasers:
1. Most of the large sites in Russia and Ukraine wrote about us. All analytics and conclusions are presented.
2. We have risen to 3 places in the general Top App Store of Russia and Ukraine, 1-2 place in the category. At the moment we are already the 4th day in the Top 5 with more than 60K installations and about 500 reviews / ratings on the App Store.
3. We carefully documented the whole process and made many screenshots along the way.
')
Entry - wife asked
The idea was suggested by the wife - she is an active user of Vkontakte, but because of the abundance of groups and publics in the tape she began to miss the news of friends. At the same time, I did not want to turn off the news of the publics, and it was not so easy to configure Vkontakte filters.
Time and desire was - together with ptashcka they quickly made an application to which you can transfer news of interesting groups and read them there, hiding from the main VC tape. Thus, friends remain on Vkontakte, and public subscriptions can be read through our application, when convenient.
The juice
We posted it on the App Store and found that users like it. We decided to continue the development and focus on 2 things: simple and elegant design, without an abundance of unnecessary details, as well as a selection of interesting groups by categories.
Design
Developed the theme of minimalism and visualization of information. Hidden likes and reposts under the cat, by default, we show only the first 160 characters from the text block, the interface elements are hidden, so that the maximum screen space can be given to the content. For the popular format of posts with several pictures, we developed an issuance algorithm, which we are proud of - the images are formed very harmoniously and create their own mood for each news.
For a long time they made an icon, went over a lot of options. I didn’t like everything until I thought of putting a cat on it - the most popular object of photos on the Internet:
Once again convinced that a good icon and name are extremely important - after changing them, organic plants grew 5-6 times!
Creating collections
They took the selection of groups seriously - they analyzed all VC groups with more than 100K participants and each gave a rating based on the main criteria: more likes, reposts, less advertising and links, unique content and a minimum of bots. They made their own analytics, which regularly monitors the quality of groups in the compilations. Including set up filtering of advertising posts in particularly annoying groups. If the habrasoobshchestvu interesting - write, make a separate post on this topic.
Training and analytics
I want to note the importance of proper analytics within the application and setting up user funnels. Ilya Krasinsky, who led the mobile development of LinguaLeo, helped us a lot. Ilya immediately pointed to our big miscalculation - after launching the application, the user saw all the groups at once and was forced to understand them, which is why only 74% got to the tape. After we introduced a simple intermediate screen for choosing interests, this indicator increased to 85%.
Put up a funnel to activate users, the ultimate goal - the entrance to the tape and view multiple posts. We created 3 base cohorts based on frequency of use: Heavy (every day), Regular (once and more per week) and Infrequent (less than once a week). By analyzing the use of Heavy, we try to understand how the application can be improved, Infrequent - where the problems are in the interface and why users do not reach Regular. And, of course, they looked at the most important parameter - retention - how many users remain in the application and continue to use it. Flurry in this regard, in my opinion, is better than Google Analytics, because immediately calculates these metrics.
For analytics we use Flurry and Google Analytics. Flurry is good free of charge and a cohort analysis out of the box - you can instantly see how changes in the application affect conversions. Google Analytics can also be commended for the online mode and the quick update of statistics, as well as the convenient sorting of events into categories, which, unfortunately, is not in Flurry.
To manage the project, they used simple and free Trello, where they also tracked the bug.
Before launch, we optimized search words using sensortower.com, everything is quite simple and convenient. As a guide, they used this article on Habré .
Blitzkrieg “App Store”
After the functionality was implemented and tested, we were satisfied with the conversion and retention, and the search words gave the best results, it was decided to move to the Holy Grail of developers - Top App Store.
Planning
To reach the Top, it is necessary to have as many installations as possible in the shortest possible period of time, so our marketing activities were focused on time and aimed at 10-12K installations per day. According to the promises of luminaries, this should have been enough to get to the Top. That's how it turned out.
The plan was:
1. Reviews on thematic resources.
2. Advertising in the public VK.
3. Facebook Mobile Ads, as a backup, if necessary, quickly throw traffic.
4. Partnership with other applications and co-promotion.
As day X was chosen the 7th of September. First, the people returned from vacations and holidays, and secondly, we would have slipped in front of the news wave about Apple updates. But then, unexpectedly, Apple itself intervened.
The final version of the application was uploaded to the app on August 23rd. According to our calculations and previous experience, this should have been enough for consideration, but on the 5th the application was still in “Waiting for review” status. A request for an Expedited Review was submitted, it was confirmed, but the application was only reviewed on the 8th. Here, by the way, is a useful resource that calculates the time of consideration based on data from other developers from Twitter.
I had to renegotiate with all review sites and partners on the 14th of September. On September 9, we appeared in the App Store and, taking advantage of the resulting break, we worked on reviews and ratings: in the update we asked our users to leave a comment on the application, and also used the service for this. Judging by the latest news , ratings and users seriously affect the rating.
Reviews
At first glance, planning reviews on Saturday looks illogical, when traffic drops significantly compared with weekdays, but we were guided by the considerations that fewer publications and reviews spend more time on the main pages, so more users will see them. No conclusions on the effectiveness of such an approach have yet been made, I will be happy with expert opinions in the comments.
In the end, agreed with iphones.ru, planetiphone.ru, macosworld.ru, ukranianiphone.com, macdigger.ru and tech-touch.ru. From the advantages it can be noted that on iPhones.ru a link with an icon in the right navigation bar hangs for quite a long time, while planetiphone.ru has a mobile application where they send push notifications. Asked to use our links to track campaign performance. The main idea is to count not only clicks, but also installations, as well as the quality of traffic from each individual resource. There are many good solutions for this, such as mobileapptracking.com and appsflyer.com, which are considered the most accurate, but we ended up using the free Flurry. In the User Acquisition section, Flurry can generate the required number of links and link them to your funnels and cohorts. Thus, you can see what percentage of users who have installed from each specific marketing channel converted to Heavy users.
In fact, it turned out like this:
+ UkrainianiPhone.com used its tracking links and gave a total of 684 clicks.
I do not know how correct it is to evaluate the effectiveness of clicks, because I usually take the phone myself and find the application on it without following the link. Settings on clicks and the quality of users have not yet miscalculated, lay out in the comments as soon as it will.
Public and VK groups
Android users there are on average 2 times more than iOS, so there was little point in advertising in general public. We selected a few dedicated to iOS applications and located in them. After the launch, in order to maintain traffic, we made several more posts in groups. According to the observations, groups up to 500K participants work best - there is excellent involvement, people comment, like and repost.
Sometimes even holivory were ignited, where without them:
The problem is that it was not possible to calculate the efficiency of this channel normally - many public owners basically do not post short links, apparently fearing the possibility of a substitution and a ban from the administration. An important point, check the public for the presence of bots before you buy ads. Make it simple - go to vk.com/adscreate, create an ad for users of this group and select those who come from iOS devices. On average, on Vkontakte, this indicator is 11%. If it is significantly lower (of course, if the group is general, and not the community of Android owners), it is highly likely there are a lot of bots.
Facebook Mobile Ads
Good, but rather expensive thing. When targeting Russia, one click costs $ 0.25-0.55, while the installation goes about $ 0.8-1.2, depending on the CPC. We had a few coupons for advertising, which we spent during the campaign. The most effective ad with a CTR of about 9% looked like this:
We tried to agree on co-promotion with other applications that are relevant to our target audience, but nothing really came of it. While our application was unknown and unpopular, few people wanted to go for a partnership, so they were postponed.
In general, they spent on marketing, not counting FB coupons, about 40 thousand rubles.
Stamping on Top
Started at 10-00, September 14. Here is a graph of new users on Google Analytics for the weekend of September 14-15:
After the first peak of users passed, it turned out that we didn’t get to Top, so we made a decision in addition to enable FB mobile ads to full. Totally spent coupons for about $ 300, received about 1500 clicks. App Store now recounts ratings every 3 hours, for us it looked like this (general / category, Russian App Store):
Sat, 9:00 am - 563/53
Sat, 12:00 - 29/4
Sat, 15:00 - 25/3
Sat, 18:00 - 17/3
Sat, 21:00 - 15/3
Sun, 00:00 - 13/3
Sun, 3:00 - 11/3
Sun, 12:00 - 9/3
Sun, 15:00 - 8/2
Sun, 18:00 - 7/2
Sun, 21:00 - 6/2
Sun, 00:00 - 5/2
Accordingly, conclusions can be drawn about the effect of the number of installations on the rating. After we climbed to the 15th position, it went much easier, as organics helped.
Rating:
Settings for the weekend:
During Monday and Tuesday, we tried to keep in the Top-5, because It’s exactly the five first places that users see by default when they go to the iTunes homepage. Fueled by traffic from Facebook Mobile Ads, publicity ads and several reviews that came out later than the main wave.
Monetization
To test the monetization of the disconnection of advertising, an Admob / iAd ad banner was installed, which can be disabled for 99p. I’ll say right away that there are few purchases, pennies go on advertising - on Saturday, September 14, about $ 30 in total. At the time of promotion they made promotional codes, by entering them it was possible to turn off advertising for free. Unfortunately, there was a bug that did not allow activating the promo code of a part of the users, so the ads were turned off altogether so as not to upset anyone until we correct the error. Now we stopped on another version of monetization, more about it later, when there will be numbers.
Total
Obviously, the application is not for everyone: someone grew out of VC content, someone has already formed a circle of interests, but they have their own audience and are satisfied - the average rating is 4.5 out of more than 500 left.
Thanks to everyone who wrote, it was very nice to receive your ideas and wishes. We have released the first version and are actively improving the application.
Ask questions and share your experiences in the comments!
Good luck in conquering the App Store!
Source: https://habr.com/ru/post/194164/
All Articles