The first month of the promotion of a free non-game application: rakes, cones, good luck, conclusions

So, behind the first month (well, a little more), as our first application appeared on Google Playlist. Not only is non-fiction, also in one of the saddest categories - books and reference books. Promoting such a thing in our troubled and unreadable time is also a challenge, although the application is free. Below, I tried to collect everything that might be useful to other such negeymdevs - statistics, exhausts from different sites, thoughts out loud.

Immediately - introductory, which I myself drew, still planning to enter the market. First, I will not spend a penny on promotion of the application. Not because they are not (although ...), but for now from the principle - to collect the maximum yield from free options. Secondly, to limit only installations through Google. And immediately the spoiler - the first is still possible (0 p. 00 kopecks spent), the second - is gone.

And another spoiler - under the cut analysis of the conversion before and after attracting an audience of torrentofilov ...
')


So, everything began according to the plan - with the placement of ads on our sites: the site, social networks, newsletter. To be honest, I was hoping to gain at least a couple of dozen installations with them, because the audience is loyal, everyone knows us, the total number of people even with doubles is more than 700 ... Schaz! The exhaust is almost zero. Conclusion - on "your" hope, but with the plan "BE" do not delay.

Then I began methodically to pile up the IT platforms, where reviews are posted, and in parallel - where you can publish the news with a link. On Habré, I found a list with recommendations, she was joking with popular resources. As a result, scored 12. Below - briefly about each.

www.ferra.ru
Published: yes
Location features: Publications of new users only in the sandbox.
What is done: Placed a review in the sandbox

android-magazine.ru
Published: yes
Location Features: Publication conditions are not clear: android-magazine.ru/about . Answered: sent materials 1500-2000 characters for free, their review for a relatively small money.
What has been done: 11/20/2014 - I sent a question about the possibility and conditions of publication on Partner (@) android-magazine.ru. 11/21/2014 - Submitted a quick review for publication.

www.android4all.ru
Published: yes
Accommodation Features: Reviews write for free. Read more: www.android4all.ru/about
What is done: 11/19/2014 - Sent a letter with a request for review. No answer. But looked over!

android.mobile-review.com
Published: no
Accommodation Features: Reviews write for free. Learn more: android.mobile-review.com/news/53
What is done: 11/19/2014 - Sent a letter with a request for review. No answer

droider.ru
Published: no
Accommodation Features: There are no details on terms and conditions.
What has been done: 11/19/2014 - sent a request for roy (@) droider.ru. No answer

droidtune.com
Published: no
Accommodation Features: Publish submitted articles: droidtune.com/o-proekte-2
What is done: 11/19/2014 - Submitted an article "About the project." No answer

trashbox.ru
Published: no
Accommodation features: Publish for a fee, but they can review especially interesting things for free. Worth a try
What was done: 11/20/2014 - I sent a request for a free review on rost (@) trashbox.ru, m1 (@) bobylov.ru They answered almost immediately what could be, but then silence.

www.proandroid.net
Published: no
Location features: Recommended on Habré, which is a good review done. But a year ago.
What is done: 11/20/2014 - I sent a request for review through the form on the site

android-zone.info
Published: no
Accommodation features: Recommended on Habré. It seems to be free: android-zone.info/razrabotchikam
What is done: 11/20/2014 - Submitted the article "About the project"

gagadget.com
Published: no
Features of placement: Publish for free articles with a die “for publicity” and signed by the Advertising robot Vasya. The rest is paid: gagadget.com/another/6422-partneram-i-reklamodatelyam
What is done: 11/20/2014 - I sent the article “About the project” as a press release to editors@gagadget.com. No answer

androidmag.net
Published: no
Accommodation Features: Post news and reviews with moderation: androidmag.net/contact-us
What has been done: 11/20/2014 - I sent the article "About the project" through the form on the website

www.androidtalk.ru
Published: no
Accommodation features: Recommended on Habré. It seems to be free: www.androidtalk.ru/about
What has been done: 11/20/2014 - I sent a request for review and the article "About the project". Mail not delivered

Conclusion on placement - it is worth throwing the network widely, because it is not clear to whom that seems interesting. I will not draw conclusions on efficiency - I will give further specific statistics with publication dates of materials. Match yourself who is interested.

Simultaneously with the hilling of androids, I tried to cover several literary sites with which I myself have long been friends. Someone came from AstraNovy - our partner. A good result was given by Fantlab - with us we are generally in tune with close cooperation. I already pull up the ratings ratings in the catalog, I am planning to work with reviews. In store I have a couple more links, which need to go with some already but some baggage. T.ch. the implementation of the plans is not complete (see below). Nuance - as opposed to IT sites with application overviews, the sites that are specialized in topics are already established contacts with the administration or top users. I suspect that a couple of references can give zero exhaust, because there I am only one of a host of users, albeit a long-standing one.

Toward the end of November, preparing a version with the Fantlab ratings and sadly looking at the installation statistics, I decided to abandon the second paragraph of my introductory and prepared the applets with markers for third-party sites - to track the installations from them through the Google Analyst. By the way, a banal, common truth, which I ignored for a headache - the analyst should be embedded in the very first release! I only got it in the second, so some of the data is simply lost. Small part, but offensively. And it is necessary to think through the formulations of actions and events in advance, and from the position of convenience of analysis. Now I have a lot of crap precisely because of bad language. OLAP-cubes are terribly inconvenient to build, you have to manually redraw the data in Excel. And their volume is growing ...

I posted the api on w3bsit3-dns.com and androidforums.ru. For some reason I decided that a file is also needed on androidpit.ru, so the material was placed there much later than the first wave. However, in vain. So ...

With w3bsit3-dns.com the file has spread over a bunch of other forums, and, in a couple of places, interesting reviews have appeared without my initiative. Only installations with a marker (i.e., from files pulled from there) already for fifty dollars, and this also had a positive effect on Gugloplei. Conclusion - yes!

But androidforums did not give anything. No installation, no effect on Google Play. Perhaps the zero settings are due to the fact that the main audience that pulls files for their sites has already merged them with w3bsit3-dns.com, but this gap in performance still does not explain. Or maybe the audience itself is so different that our application is just completely by-passing the second ... Conclusion - if it was only to be, although I wouldn’t have completely discounted it, another application may have a different effect.

Further - more interesting. "And there is no figly?" - I thought, and went to deal with torrents. The application is free, and looking at the change in conversion with the advent of supporters of free dissemination of information is extremely interesting. Whether or not the hypothesis is confirmed that if they are offered a convenient service for choosing and buying books, will they actually buy? The conversion chart and comparison of average values ​​before and after see below.

I included two in the list of torrents: NNM-club and Rutracker. On the first, information is added without pre-moderation, but newly registered users get the right to place posts only after 3 days. On the second - pre-moderation, which lasts about a day. At least 24 hours later, I wrote a request to the admins in the subject, which rather quickly was considered positive.

The difference in time in the publications on torrents I got exactly two days. And again the same picture - NNM-club gave both native installations (25 with a marker), and from Googleplay, and the file spread to other torrents (including porn ...% -0). Rutracker practically didn’t notice the publication (2 installations), and considering that this torrent, unlike NNM-club, is forbidden to place links to other sites, then you can’t put a direct link to the market and hope for an increase in installations on Google . Here I fully admit the large intersection of the audience of these torrents, t.ch. I place the apks there in a different sequence, the result could be different (although it is unlikely on the effects on the installations in Googleplay). Spending power on two torrents is meaningless. Which to choose? If such an audience fits, then NNM-club is more likely yes, Rutracker is more likely not. The inability of the second to add a link to the market, in my opinion, leads to the unconditional victory of the NNM-club.

Well, now it's time for statistics:



Together with the dates of publication of materials in the network:

11/14/2014 - Ads on our social networks, our newsletter from the site
11/19/2014 - ferra.ru and livelib.ru
11/21/2014 - AstraNova
11/22/2014 - Fantlab
11/24/2014 - android-magazine.ru
12/03/2014 - android4all.ru and w3bsit3-dns.com (coincided)
12/05/2014 - androidforums.ru
12/07/2014 - coandroid.ru (review done by yourself)
12/08/2014 - prodroiders.com (review done by yourself)
12/11/2014 - androidpit.ru
12/13/2014 - nnm-club.me
12/15/2014 - rutracker.org

And now about the conversion ... I confess that I am an optimist and I try to think about people well. Therefore, the hypothesis voiced earlier adhered to itself. But the numbers below made me take a couple of steps towards the pessimists' camp ... Or realists?



Yes, there is a splash (although it could have come from w3bsit3-dns.com too - see the figure above), but if you look at the average numbers before and after, you can see that it is a one-time (curiosity?):

Average number of active users per day 02-12.12.2014: 19.6
The average number of sessions per day 02-12.12.2014: 26.9
Average conversion 12/02/12/2014: 9,4%

The average number of active users per day 13-22.12.2014: 23.6
The average number of sessions per day 13-22.12.2014: 38.0
Average conversion rate: 12-13 December 2014: 3.2%

Much could be explained by a small amount of data, but not three times! :( Forum users had a negative effect on the fall in conversion - if the first period was divided into “before 08.12 and after”, then its average value would be 10.6% and 8.04%, respectively.

This is the situation at the moment ... What are the plans?

1. Reach the remaining sites where our profile audience is a handful of readers.
2. Try to use the media, since there are already some results, there is something to write in the press release.
3. Trying to get publicly the opinions of experts in our publishing industry (we’ve already started to beat the thresholds, we’ll see), or maybe not only in ours - such an advancement tool may also be interesting for marketers, and people who read hudlite are also found among them ...
4. Look at the results of advertising in a paper magazine - at the beginning of the year, FanCity’s number should come out with our module right on the page (many thanks to the editorial board and Andrey Bocharov for the support!)
5. Release version 1.0 in January with significant updates and re-try to bypass the sites where applications are being reviewed.

Something from the free, but worthy use I missed?

And finally, I would like to identify a few points that have been important and useful for me, even if, at times, at the price of stuffed cones:

• Acquaintance, even if not very deep, with the project management methodology helped me a lot in my work. Especially if you consider that I acted in the role of both the customer and the performer - it is easier to agree between each other easier;)
• Add pre-thought analytics in the first release - already mentioned it. And it is also useful to make a global switch constant for a complete ban on sending statistics — useful for testing, in order not to clutter the real database with left data.
• Adding the ACRA library to catch bugs is also available on the very first release. Moreover, it is necessary to set up sending logs for reasons of balance in the volume and importance of errors. If you send everyone, for example, you can get 100,500 messages about Internet unavailability from the user ... And it is also useful to make a global switch constant to completely prohibit sending logs - useful for testing, in order not to clutter the database with left messages that are tracked in the debugger window.
• Choose the icon and the name of the application from the following considerations (there are a lot of tips like “pay special attention to them” on the net, but how it’s not said):

The icon in the market should stand out and immediately catch the eye - maybe with brightness, and maybe with laconicism. But stand out. This is the first chance to hook the user. We need to see successful examples in the top of the target category. Then - as the bottom and the middle looks at the moment, and do something differently.

The name (alone or paired with the icon) should immediately open the curtain from the list of applications in the catalog - and what's inside. A beautiful, but abstract name will not push the user to open the application page and read its description. In 30 characters it is difficult to keep within, but it is possible. Albeit at the expense of beauty and sonority - this is not the purpose of the name in the market. The application itself can be called anything. Here I will still be thinking with my own and maybe experimenting.

• When developing it is clear to fix - what functions to include in the nearest version, and what can be postponed to the next version (in terms of project management - content management), otherwise you can get stuck in endless improvements and additions. If possible, set yourself a clear deadline, and even read them publicly, so as not to be tempted to move the release a little more, then a little more, and again endlessly. This approach helped me a lot, otherwise it wouldn’t still be launched ...
• Keep on the table, read, re-read and use as reference books the following books:

Herbert Shieldt. Java Complete guide
Programming for Android (O'Reilly, Peter Publishers)
Eric Freeman, Elizabeth Freeman. Design Patterns (O'Reilly, Peter Publishers)

It seems to me that this set is a Mastrid for every novice developer for Android / Java. In a complex, balanced and methodically competently, all the necessary material that teaches the right is given. The rest is already in addition and the expansion of this base, depending on where you want to specialize. I will now be looking for something about the work of Android with clouds and for Spring in particular. Course on Coursera to begin to review, listen again ...

Well, all of the coming, of course!


* my plans for the New Year holidays covered with a copper basin ... I will prepare a new version ...