How to monitor your app in first few days? - Analytics

Hello everyone,
We have launched an android app, Factor Monsters (by Knowledge Platform), yesterday. Google Play is not showing downloads as for now and I came to know that it will show number of downloads after 2-3 days.
I am curious that, in such case, how can I benchmark our app's performance? We have received 30 ratings so far. Are they a good start? Is there any benchmark or tool to determine if our app is going fine or not?
I tried to do Google Search but was unable to find any answer which can give this information.
I shall look forward to learn from you in this regard.
Best,
Hira

Once and for all. Remember one thing. You should integrate analytics services into your app to get prorper information and on time.
30 is rather good but the amount of marks rarely indicates the amount of downloads. I would suggest 100-500 in your case.

Gooldy Games said:
Once and for all. Remember one thing. You should integrate analytics services into your app to get prorper information and on time.
30 is rather good but the amount of marks rarely indicates the amount of downloads. I would suggest 100-500 in your case.
Click to expand...
Click to collapse
You would suggest 100 - 500 of what in his case? Downloads?
Usually, you can expect that less than 0,5% of the users will review your app, and that 1 - 3% will give you the rating. Top games (Clash of Clans) get up to 10% ratings. But with such a low data, anything could happen. Also, many things could affect them:
- have you reviewed your app? With 30 reviews, your review is 3,3% of all your reviews...
- have your friends or family reviewed your app?
- have you asked people to review your app?
Anyway, good indicator would be to check how many impressions have you generated. Ad providers are much quicker with their dashboard updates. Also, integrating Google Analytics or Flurry could give you detailed insight in your users count, session length etc.
If you got 30 unincentivised reviews from general audience (not your facebook friends etc) it would mean you are most likely in 5000+ downloads range. But I guess you got these reviews from your friends or people that are somewhat familiar with you, and you are in 100 - 500 downloads range. Anyway, you can't expect any detailed and accurate data with such a low reviews count.

Hey hzainab,
30 ratings is a good starting point for just 2/3 days on the store, but as others have stated, I would wait until you have reached a higher volume if you want to use ratings as a metric.
I would focus on getting to know my users and the retention. Ratings may be good, but have you wondered how many of your new users have returned to your app after the first session? Or even if they are using the app the way you intended? Who are they? How much are they using? These KPIs will be more important on this stage since they will allow you to fine tune your users experience.
Google Analytics or Flurry can be an option and they are free services, but they will just give you some (limited) information. Since this is a newly app and you probably don't have a gigantic user community already, why don't you try other services which will also be free for your current numbers?
I know I am biased because I work for Liquid, but I would suggest that you take a look at what we do.
Regards and much success.

vaktu said:
You would suggest 100 - 500 of what in his case? Downloads?
Usually, you can expect that less than 0,5% of the users will review your app, and that 1 - 3% will give you the rating. Top games (Clash of Clans) get up to 10% ratings. But with such a low data, anything could happen. Also, many things could affect them:
- have you reviewed your app? With 30 reviews, your review is 3,3% of all your reviews...
- have your friends or family reviewed your app?
- have you asked people to review your app?
Anyway, good indicator would be to check how many impressions have you generated. Ad providers are much quicker with their dashboard updates. Also, integrating Google Analytics or Flurry could give you detailed insight in your users count, session length etc.
If you got 30 unincentivised reviews from general audience (not your facebook friends etc) it would mean you are most likely in 5000+ downloads range. But I guess you got these reviews from your friends or people that are somewhat familiar with you, and you are in 100 - 500 downloads range. Anyway, you can't expect any detailed and accurate data with such a low reviews count.
Click to expand...
Click to collapse
Yep, downloads. Usually you get bigger review\download rate in the beginning because of friends review , friend of friends review - etc. One crappy game of mine got 60 reviews and less then 700 downloads

Hi, You can choose App Annie, App Figures or Flurry to track your apps performance.

We used Google Analytics and BugSense.

Tips to monitor your app -
1.Firstly – don’t only focus on app downloads!
2.Want to improve your app? The real data is about engagement!
3.More behavior to track? Push notifications!
4.Track those all important reviews and ratings
5.Depressing uninstall analytics – not to mention crash stats!
6.User Acquisition – where did a spike in downloads come from?.

Some steps
hzainab said:
Hello everyone,
We have launched an android app, Factor Monsters (by Knowledge Platform), yesterday. Google Play is not showing downloads as for now and I came to know that it will show number of downloads after 2-3 days.
I am curious that, in such case, how can I benchmark our app's performance? We have received 30 ratings so far. Are they a good start? Is there any benchmark or tool to determine if our app is going fine or not?
I tried to do Google Search but was unable to find any answer which can give this information.
I shall look forward to learn from you in this regard.
Best,
Hira
Click to expand...
Click to collapse
Below is the data you need to track immediately after launching:
New users from acquisition sources
First-week engagement rates
First-week retention rates
Most used features
Least used features
App-store views and ratings
App installs vs. churn
Glitches/bugs

Related

[Q] I am looking for a good gym app

There are a few gym and workout apps but I can't find any that seem to describe what I want and since the market now only has a 15 minute trial, it's not enough to check the non-free ones out.
Basically I want to be able to put in a few different routines I have, each with a variety of cardio exercises, weights and stretching and then set the reps and weight etc.
Each time I go I'd like to be able to pick which routine I did then have the app, over time, suggest changes. For example, if I've done 2x10 reps of 40kg bench presses a few times, it will suggest I put the weight up on that. Or when I've run 10 minutes on the treadmill at 10kph a few times, it might suggest upping the time or speed etc.
Anything out there that can do this kind of thing?
idk jefit pro
has free version and is good but pro much better
I tried the free one and it does seem ok but doesn't do any kind of advising. That's what I'm really after
Gymprovise
Gymprovise is what you need. It is the best! No other gym/bodybuilding app comes close.
-Looks awesome. Beautiful, even. Other workout apps are butt ugly by comparison.
-UI is intuitive, clean, smooth, and most importantly, FAST. Doesn't need to connect to internet while you are tracking.
-Big database of exercises. Got all the usual suspects. Pictures too which are top notch. You can create your own too.
-You can plan and track routines with strength, cardio and stretching, and then view your
-Got weight and body measurement tracking as well
-Does not have social rubbish built in which IMO is just a freaking distraction from the Workout - although I think you can share completed Workouts to FB etc
Been using it for over 6 months and I have not looked back.
JEFIT is completely overrated. Steer clear. You'll be pulling your hair out trying to get it to do what you want.

[POLL] Help collect information about Nexus 7 I/O performance

Thought I'd try to get some information about the lag caused by the Nexus 7's storage. The reports of individuals here are not detailed enough to provide a way of figuring out if the lag only effects certain versions of the Nexus 7, is a result of age, or might only effect early shipments (like the 8 GB model). Here's a link to a Google drive spreadsheet to facilitate the collection of data (edit: Fixed sharing to allow others to edit and added a few more categories. Sorry about that):
https://docs.google.com/spreadsheet/ccc?key=0AksJrQ4EFpS4dFRPcXlXVFc1QlJaOElZTmljRFF0LXc&usp=sharing
Categories are: storage capacity, storage used, age, 3G model, # of apps installed, ROM, and scale of lag.
Sent from my N7
DroidRunner said:
Sent from my N7
Click to expand...
Click to collapse
And I welcome those posts. This is a grown up way of collecting data and all samples are accepted. It'll give a more complete picture of why some people see the lag and some don't. By the way, next time why not actually contribute to the thread as well instead of posting a humorous image only?
armada786 said:
And I welcome those posts. This is a grown up way of collecting data and all samples are accepted. It'll give a more complete picture of why some people see the lag and some don't. By the way, next time why not actually contribute to the thread as well instead of posting a humorous image only?
Click to expand...
Click to collapse
This is my contribution:
http://www.reddit.com/r/Android/comments/1gkiuv/one_year_later_the_nexus_7_has_gone_from_the_best/
Read the comments. Sometimes LagFix will fix it for a bit but will slow down again. This is why everyone is looking forward to the 2nd generation model. Compiling data as to what model and space available is good fun, but it comes down to the crappy internals they used to make a $200 tablet.
Sent from my Nexus 7 using xda app-developers app
Contribution #2:
Ran LagFix and re-ran benchmarks and saw a jump in I/O performance. There.
Contribution #3
It wasn't a problem when the device was released because the flash chips hadn't been used yet. All flash memory has a limited number of erase/write operations that can be done before it can no longer be reliably read. And as the IO controller has to do more and more error correction to compensate for the increase in error rate your read rate goes down.
There are a number of things that manufacturers can do to mitigate the time it takes for this process to happen. SSDs typically have a reserve of unused blocks that can be made active to replace blocks which have an uncorrectably high error rate. They can also use wear leveling to make sure that the drive degrades evenly over time.
Another contributing factor is that the controller in at least the Nexus 7 apparently isn't doing garbage collection on the flash memory. This causes a problem because unlike with magentic-based memory, flash memory must be empty before it can be written to. The technical reason for this being that although flash memory can set individual bits to 0, it can't set those bits individually back to 1, in order to set them back to 1 the entire block has to be reset to 1 first. Garbage collection is supposed to go through and empty sectors that are no longer in use but if it's not doing that then it increases the write time because it has to erase whatever is on the memory first before it can be rewritten. See this issue: https://code.google.com/p/android/issues/detail?id=37258
I don't think companies are doing this out of malice as you suggest, they probably just assume that since most people replace their phone and tablet every 2-3 years anyway, degraded performance after 2 years is an acceptable trade off for cheaper and less complex memory and controllers.
Click to expand...
Click to collapse
Aaaaaand #4:
http://www.anandtech.com/show/6073/the-google-nexus-7-review/6
Really good article.
TL;DR: Crappy hardware
Sent from my Nexus 7 using xda app-developers app
DroidRunner said:
This is my contribution:
http://www.reddit.com/r/Android/comments/1gkiuv/one_year_later_the_nexus_7_has_gone_from_the_best/
Read the comments. Sometimes LagFix will fix it for a bit but will slow down again. This is why everyone is looking forward to the 2nd generation model. Compiling data as to what model and space available is good fun, but it comes down to the crappy internals they used to make a $200 tablet.
Sent from my Nexus 7 using xda app-developers app
Click to expand...
Click to collapse
This thread is for collecting a set of empirical data to uncover some more reasons why or if it is really only crappy internals. If you're not going to contributea sample then its not relevant to the scope of this thread.
OP just wants to collect data on a perceivable issue, and has a spreadsheet for contributions to the thread.
Not to read chat logs.
If any folks don't wish to contribute as stated, then it's best to just move along to another thread. Works out easier in the long run.
Thank You.
MD
armada786 said:
Thought I'd try to get some information about the lag caused by the Nexus 7's storage. The reports of individuals here are not detailed enough to provide a way of figuring out if the lag only effects certain versions of the Nexus 7, is a result of age, or might only effect early shipments (like the 8 GB model). Here's a link to a Google drive spreadsheet to facilitate the collection of data (edit: Fixed sharing to allow others to edit and added a few more categories. Sorry about that):
https://docs.google.com/spreadsheet/ccc?key=0AksJrQ4EFpS4dFRPcXlXVFc1QlJaOElZTmljRFF0LXc&usp=sharing
Categories are: storage capacity, storage used, age, 3G model, # of apps installed, ROM, and scale of lag.
Click to expand...
Click to collapse
On what scale is lag measured in? I used a scale of 0-10 where 0 is no lag and 10 is so much lag that it's unusable.
I have done my part. I'm interested to see the result.
veeman said:
On what scale is lag measured in? I used a scale of 0-10 where 0 is no lag and 10 is so much lag that it's unusable.
Click to expand...
Click to collapse
The comment in the spreadsheet states it's a scale of 0-5. I can change this to make it more visible. I felt that a scale of 10 was too granular.
armada786 said:
The comment in the spreadsheet states it's a scale of 0-5. I can change this to make it more visible. I felt that a scale of 10 was too granular.
Click to expand...
Click to collapse
That's my comment LOL
Sent from Nexus 7 3G using Tapatalk 4
stfudude said:
That's my comment LOL
Sent from Nexus 7 3G using Tapatalk 4
Click to expand...
Click to collapse
I actually made one at the top too! That's what the yellow corner signifies; a comment.

Google Now

Forgive me if this has been discussed but I have not seen any serious discussion about it. Many reviews of the Nexus 5 mention inconsistent battery life. In my humble opinion most of these are due to Google Now. I think the idea is fantastic and it is amazing at figuring out what you are doing and where you are going. I firmly believe in a few more development cycles it will become a major part of many people's lives. However if you analyze it with a good battery monitor you will find that it sometimes hammers the data connection and more importantly wakes up the phone almost continuously. Other times it is very parsimonious in its use of wakelocks or the data connection. I am certain Google is working on improving this but I absolutely believe that if one wishes to have guaranteed all day battery life the only way to ensure this is to disable Google Now.
I am curious if others have had or similar experience or wish to violently disagree with me. I actually wish someone like Anandtech would do an objective analysis of this. Some of my conclusions come from hard data from GSAM battery monitor but other observations are far more anecdotal.
Agreed, many of googles own apps and services are terrible on battery life. Even worse then google now is google maps. It runs continuously in the background even if you try and kill it.
jlevy73 said:
Agreed, many of googles own apps and services are terrible on battery life. Even worse then google now is google maps. It runs continuously in the background even if you try and kill it.
Click to expand...
Click to collapse
That's why you can root and remove it (maps) from /system and then install it from the playstore as a normal app, and freeze it with greenify
Sent from my Nexus 5 using XDA Premium 4 mobile app

[Q] App isn't showing in top new free at Google Play

I've released a game at deс 22th in "adventure" category. It was reviewed by youtube blogger and it got a boost of installs. Now it is around 1200 (4.5 stars based on 173 ratings) Since a week I can't find my game in "top new free adventures" for my country (all installs are from there). But same time I see there games with 0-10 installs! And the top in not full, it contains only 450/500 games.
Few days ago google manually changed content rating to "Medium Maturity" for this game. Could it cause disable participating the goole tops for my game? Or maybe there are new rules for rating new games?
I think their algorithm considers other things as well along with your rating and downloads . e.g the store listing, compliance with their policies. Not sure but we have also faced same scenario in past.
WakingBat said:
I think their algorithm considers other things as well along with your rating and downloads . e.g the store listing, compliance with their policies. Not sure but we have also faced same scenario in past.
Click to expand...
Click to collapse
And what did you perform to solve this? I'm afraid to publish new app if it will be invisible in tops.

Help! My Google Play app is being spammed with hundreds of 1-star ratings!

This has been a nightmare for me tonight, watching my app go from a 3.9 rating to a 3.7 in a matter of hours.
I regularly monitor my app's statistics in both the Google Play Developer Console and an app called Andlytics. I look at my average rating, reviews, downloads, app crash logs, etc. Being intimate with my daily numbers, when they start to change drastically, I can easily spot when something is off.
On 11/7/2015, I got hit with numerous fake ratings and reviews. I received a total of 52 1-star ratings for that day, which is about 30-40 above average (I average about 5-20 1-star reviews a day). Other than the countless negative ratings, nothing else was abnormal (normal downloads, normal crash logs, etc). After reading a few of the reviews they happened to leave, the picture became clearer... all of them came from the same handful of devices, and every single one of them had the en/RU language set on their device. I went ahead and flagged those reviews as spam -- so far they are still there, but fingers crossed that they will go away. As far as the ratings go, I am unsure of how to view and flag the individual ratings (reviews left without comments), but I am absolutely sure that they left some of those as well (by my count, at least 10-20). Here are some screenshots I was able to take of the reviews...
https://drive.google.com/open?id=0ByK_ywEzSR7GSjJ4RHVpRXAtbUk
https://drive.google.com/open?id=0ByK_ywEzSR7GNUtrSl8tbXZKM0k
https://drive.google.com/open?id=0ByK_ywEzSR7GQjdfdFNxVG5LYUU
https://drive.google.com/open?id=0ByK_ywEzSR7GalU5eTNRdHM1ZHM
https://drive.google.com/open?id=0ByK_ywEzSR7GOHVsSE5oczdRRU0
https://drive.google.com/open?id=0ByK_ywEzSR7GdUpQVTJEUDZOZDg
https://drive.google.com/open?id=0ByK_ywEzSR7GOTJzMGxPSWdtakU
A few days went by with no strange activity. Until on 11/15/2015, I again was spammed like crazy with a whopping 613 1-star ratings! No reviews were left this time -- only ratings. And again, besides the avalanche of bad ratings, other app statistics looked normal (normal downloads, normal crash logs). This time around, I knew exactly what was going on, as I had read about it a few days prior... Apparently there are services out there where users can pay for thousands of fake reviews on a competing app, to ultimately lower its ranking in search results. Notice how in the screenshots below, the 1-star ratings go from a 277 at 9:20pm, to a staggering 609 at 10:12pm! In less than an hour my rating had drop 0.2 -- from 3.9 to 3.7 -- in comparison, my usual rating change for a day is a +- 0.001... By the end of the night, the final number of 1-star reviews settled at 613, as seen in the third screenshot.
https://drive.google.com/open?id=0ByK_ywEzSR7GOHAzVmdLVm1FcFU --- 9:20pm
https://drive.google.com/open?id=0ByK_ywEzSR7GbXhMeThydDd0Y2c --- 10:12pm
https://drive.google.com/open?id=0ByK_ywEzSR7GQlctVGlDejBndW8 --- 12:00am
As you can see, I am clearly being attacked -- most likely by a competitor who wants to get their app ranked higher than mine. I believe that 100% of these ratings (minus the normal 5-20 daily average) are falsified, and would like help in getting them removed from my app's rating. Every moment they stay, it hurts my app's reputation on the Google Play Store.
Does anyone know the best course of action to report these fake ratings? There are several resources on the Google Play Developer help section, but they simply tell you to flag the review -- unfortunately you cannot flag reviews that were left without comments -- which the majority of mine are. I did fill out a generic form for support, but I have no idea if I'm actually sending to the right people. Anything you guys know would help!
Thanks,
-Jake
You can describe your problem to moderators:
support.google.com/googleplay/android-developer/contactflow
If you can, post here news about your progress in solving this problem

Categories

Resources