I have read a number of posts from bloggers explaining about how their WP battery life was vastly improved when they disabled “push” or “as items arrive” email. This appeared strange and contradictory to me, as from what I understand, Microsoft maintains a constant low power connection to every Windows Phone, provided a data connection is active, or the device is being powered while on WiFi. So, as the phone is already connected, all a push message does is tell the phone to ramp up power to receive a new message and popup an alert. No extra power to constantly monitor for new messages. Theoretically, push email should actually be more efficient than; every 15, 30 or 60 minutes, as that method ramps up power whether there is a new message or not.  So I figured, why not test this out and find the truth.

I have five email accounts synced with my phone, three of which can be setup for push email. The other two accounts are setup to fetch mail every 30 minutes or 2 hours. As an FYI, I also have 8 Background Tasks turned on all the time, as well as Data (mostly LTE), WiFi and Location services. Bluetooth is off unless I need it.

TEST 1 – For the first test I left everything as is. The three push accounts were left to receive mail via push. I got my battery charged close to full and set it down on my desk at just about midnight. I woke the phone up after 8:00AM. I sent two of my accounts an email to ensure there was some activity. I received a total of 3 message overnight. The results:  95% at 12:01AM / 77% at 8:02AM – Difference of 18% or 2.25% per hour. Reasonable for me considering that I typically drain 3 to 3.5% per hour with an occasional text or reading a few messages throughout the day. My signal at work is also a tad weaker. Still working on why the 3.5%. But that’s for another article.

TEST 2 – For the second test, all the parameters stayed the same, except my three push email accounts were set to check for mail every 60 minutes. The results: 97% at 12:00AM / 73% at 8:02AM – Difference of 24% or 3% per hour. Looks like I proved my point. But you can never rely on a single test.

 

TEST 3 – For the third test, I changed my three push accounts to check for mail every 15 minutes, figuring this would represent the worst case scenario. The results: 11:59PM at 97% / 8:00AM at 79% – Difference of 18% or 2.25% per hour. What’s this, an 18% drain again. Same as push. How can this be? More testing required.

TEST 4: For this test I changed my three push accounts back to checking mail every 60 minutes. Let’s see if we can reproduce the first test. The results: 12:03AM at 99% / 8:05AM at 75% – Difference of 24% or 3% per hour. Honest, I am not making this up. Even I was surprised to get the exact same result.

TEST 5: Final test. Reset my three push accounts to “as items arrive”. The results: 12:31AM at 99% / 8:33AM at 83% – Difference of 16% or 2% per hour. Not sure of the 2% difference between this test and the prior push test, but possibly because I started the text with the battery at a higher charge. There you have it though. Verified.

 

So what does this prove. Well it’s only my anecdotal personal experience, so not sure if it proves anything. Maybe not conclusive but the testing method was objective enough to convince me. I am comfortable knowing that my push email accounts are not draining my battery. And having the battery drain more when set to check mail every 60 minutes sort of validates my theory, although I figured it might only be a couple percent difference. The real mystery though is why did my battery drain more when I set my mail accounts to sync every 60 minutes and the same as push when set to every 15 minutes. Could that be a bug? Sure, testing during the day when emails are flying in would be an ideal situation. But who has the strength to resist touching their phone while it tingles and vibrates for hours. Not me.

Note that for each test, my accounts only received three or four messages overnight. While checking for messages (15, 30 or 60 minutes) should use the same amount of energy whether or not there are any, push could theoretically use more power if you receive messages frequently. For example, if you typically download dozens and dozens of messages daily, turning off push may save you some power. Then again, if you are that important to receive all those messages, you probably want to get them ASAP. 

Not convinced. You can do your own testing, even without Battery Meter. Get your phone fully (or close to) charged and check the percentage in Settings>Battery Saver. Note the time and set the phone down overnight. Wake it up in the morning and note the battery percentage in Battery Saver. Repeat the test the next night with different email settings. But try to keep your charge level, and obviously the length of your test the same. To get a good sampling, you test should be at least 5-6 hours. As with any good testing methodology, removing as many variables as possible will offer the most accurate results. Trying to perform a test during the day; driving to work, moving around your workplace, storing your device in a steel locker, going out for lunch or running errands, will all wildly skew your results from day to day. Static, controlled tests as the only way to derive accurate, comparable results.

First truth: If you want to save battery (or don’t want to use any more), turn on that push email. And if you have mail checking set to 1 hour, change it. At least for now till someone can verify if there is really a problem. I am planning a series of Truth posts to give users a better understanding of how their battery drains. Background Tasks, Battery Saver, LTE vs. Edge, Email on/off and Feedback are all on the list. Any suggestions?

20 COMMENTS

  1. Question on data usage, though. Does the “Push” notification use more data? I can deal with a little less battery if it means that I don’t go over on one of the smaller data plans.

  2. Awesome article. I am one who went from “as items arrive” to once every hour in an attempt to save my battery. It seemed at first that it helped but now I’m not sure. I was going to say that maybe the difference in power consumption is due to attachments needing to be downloaded but WP7 requires the user to actively download the attachment file once inside the email so that’s a no go.

    I look forward to the next article.

  3. Peter S: Don’t think that push vs. a timed interval would make any difference. As Murani points out, WP never downloads attachments automatically, regardless of download method. Same goes for imbedded images. And if you have ever tried reading a very long threaded message (we get lots of those in our internal staff notify mail) you will find that WP does not load the message, waiting instead for you to open the message before it starts to sync. My guess is each message has a size threshold of a couple KB, and anything above that waits for you to actually open the message, or click on a link.

    The other alternative would be setting your mail to manual, so you could control when you wanted to sync mail.

  4. Interesting finds regarding battery life. Recently one morning at work I got a constant, steady stream of a ton of emails (think 4 digits, the emails were autogenerated) for about 2 hours. I have that account set to push and it devoured my Trophy’s extended battery to the tune of about 33% in 2.5 hours off the charger. I wonder if it would have been different downloading the emails in bulk every 15 minutes as opposed to constant downloading as the messages came in? Sadly, I don’t think I want to repeat that experiment.

    Also, regarding the first paragraph, are you implying that all push emails come through MS servers? That doesn’t seem right to me and I can imagine that many businesses would consider this a security issue. Maybe I’ll do some testing when I can find some free time.

  5. Wow! With all those emails, I would sure hate to be you for a day.

    Good question about the MS servers. Not sure. I was made to understand that Microsoft controls everything that passes to your device, but for email maybe they are opening a channel that a provider can use to push or sync to your phone. So not actually touching the data. If someone more knowledgeable than me has an answer, feel free to chime in.

  6. Note that with LTE and certain breeds of HSPA+ always-on flattened all-IP is used in which case there may be no gear shifting from no data to an enabled data connection, in which case one might get zero benefit from using conservative mail polling settings versus push or imap-idle.

    Also, not bad Jim.

  7. Push email uses activesync. Its built into Exchange and other mail systems license it. Mail goes straight from your server to your phone. There is no middle man.

  8. Rob I may be missing your point but I believe where Jim is coming from is that since his WinMo days he’s been given email polling cyclicity options and figured the only reason someone might be interested in something other than instantaneous is saving juice and maybe bandwidth from some sort of additional overhead, and that he found no apparent link on his phone to saving juice and polling infrequently, which is interesting, and if his data is both good and not unique to his phone and cellular connection standard, potentially pretty helpful.

    And by push he means any form of relatively instantaneous email including activesync on your own server, the cloud or imap idle, also I suppose something with a “middle man” if you’re referring to berries. We generally don’t bother referring to those however.

  9. Rob: Ok, thanks. Guess that makes some sense. So Hotmail, Live and GMail are all using Exchange to push messages to WP.

  10. Doug: The conversation is getting a little above my level of understanding, but yes, that’s what I am referring to. Basically, no downside to receiving mail as it arrives. I suppose if Exchange does it for a business, that same technology could be used by the big players mentioned above. But the phone needing to stay in constant contact to both MS (for alerts, etc.) and Exchange (for each push account) seems like somewhat of a burden for the device.

    As I am only using push with Hotmail & Live, I wonder if the results would be different if I was getting push from the XYZ Corp. Exchange server, presuming they take different paths to deliver mail to the phone.

  11. Doug, Jim made this comment
    “Good question about the MS servers. Not sure. I was made to understand that Microsoft controls everything that passes to your device, but for email maybe they are opening a channel that a provider can use to push or sync to your phone. So not actually touching the data. If someone more knowledgeable than me has an answer, feel free to chime in.”

    Thats what I was responding to.

    Jim, Hotmail and Gmail use activsync, not Exchange. Activsync is the push technology originally developed for Exchange and windows mobile. It was later licensed to other phones and mail systems.

  12. Rob: Thans for the explanation. I know it’s not the same thing, but activsync (or activesync) makes me cringe a bit. Bad memories from PPC & WM.

    So, why do you think an Actvsync push connection uses no more energy than a pull (timed) connection? And do you have an opinion on whether an Exchange (private enterprise) vs. Activsync connection would use more, less or the same amount of power.

    If Exchange does use more juice vs. Activsync then my conclusions above don’t apply, and could help explain the “push uses more power urban legend”. Inquiring mind want to know.

  13. Not to derail the mail tech discussion, but… where’d you get that cool Battery Status app with the History graph? Nothing like that in the Nokia Diagnostics app I’ve got, for sure…

  14. Battery Status is a homebrew app that can be found over at XDA-Developers. It is using an API not officially available for Marketplace apps, so it will never be found there. It also can update its background task every 10 minutes vs. the Marketplace minimum of 30 minutes, which is another no no.

    You can pinch to zoom in on any part of the graph to get a better view. When you tap/hold along the graph, you can see the time/percentage at every interval by dragging along. So I can see my battery percentage change every 10 minutes across the day. Easy way to see drain per hour for example. All the data is saved since the last soft reset so you can bring up a graph showing a weeks worth of data. The Live Tile updates every 10 minutes as well. Really essential for doing the kinds of tests I have been working on.

  15. :-(

    Thanks, figured that was the case, but just had to ask!

    Would sure love it if MS and/or Nokia would either provide a similar Battery Status app, plus a Task-Manager-like app… or open up the APIs to allow a 3rd party to fill the void.

  16. […] First truth: If you want to save battery (or don’t want to use any more), turn on that push email. And if you have mail checking set to 1 hour, change it. At least for now till someone can verify if there is really a problem. NOTE: This has been confirmed with Hotmail & Live “push” email. Yet to be confirmed for Gmail and Exchange accounts. […]

  17. […] First truth: If you want to save battery (or don’t want to use any more), turn on that push email. And if you have mail checking set to 1 hour, change it. At least for now till someone can verify if there is really a problem. NOTE: This has been confirmed with Hotmail & Live “push” email. Yet to be confirmed for Gmail and Exchange accounts. […]

  18. […] First truth: If you want to save battery (or don’t want to use any more), turn on that push email. And if you have mail checking set to 1 hour, change it. At least for now till someone can verify if there is really a problem. NOTE: This has been confirmed with Hotmail & Live “push” email. Yet to be confirmed for Gmail and Exchange accounts. […]

Comments are closed.