In Lifetime Analytics the minimum option to use the date filter is the days. From what I was able to tell looks like the data is summarized by day and hour.

Is there any configuration that would allow me to keep the data summarization by minutes?

Can we change the Analytics filters to show also hours?


Carlos Freitas

Hello Carlos,

there's no such option to keep data with a minute precision; there's also no way to filter data by hour. Nevertheless, I would love to hear why you're asking for such capacities. 

Hi Pedro,

I need to create what will be a lifetime plugin to showing the daily trend of application usage, and I was trying reuse as much as possible the existing data. The ability to drill-down to a specific part of the day, at the minute level, would help to fine tune the information and if were looking to the current day will give a overall idea of how close are we from real-time.

I see. Thanks for the explanation. 

Do you really need real-time statistics about app usage? If the app usage drops in the last minute, what are you going to do? If it goes up a lot, are you going to create a new front end and attache it to your environment? That's to say, I get it, you want real time, but why do you want real time?

I know people love to have real-time stats, but that comes with a huge cost: being able to collect and transform data in real time is a different ball game! The architecture would have to be a lot different to be able to cope with that requirement. My 2 cents...

I'm not truly taking about real-time, but close to real-time with the current data. Currently lifetime already harvest information of every single request, each 15m. Keeping this information, with more detail for at least 30 days it would give better understand of the several application usage patterns. In terms of cost, from my understanding we would be talking, at a worst case scenario 59 times more rows for the last 30 days. At least in our scenario (about 70K requests per day), it shouldn't be a huge problem for the database storage.

In our scenario I'm not really worried about a increase of application usage, since currently it is pretty constant every business day, but building some alarms about the application getting slower would be very important for the business, since it would help us to understand and react much quicker to any issue, limiting the users that would face a performance downgrade.

Ah, so what you want is to keep the data at a minute precision, rather than having it grouped into hours. I got it now. 

Well, Analytics does not support that use case and to be honest, it is not the kind of use case that we tried to build into it. If you have a sudden impact in a live application, it's likely that it's not your code: a machine went down, a sudden spike in a database CPU, memory perhaps, all for all those use cases you have lots and lots of tools to help there. 

What Analytics gives you is trends on user satisfaction. Still, if you want to keep an eye on it, you can use the reports and have them delivered to your email automatically. Other than that, there's an API were you can grab the raw data and build your own tools!