What their recruiters said on linkedin when they contacted me. I didn't respond, but I read this article and thought I'd heard the name somewhere...
[snip; about recruiter]
We sell software to tier 1 mobile network operators. Our software is running
on over 150mm handsets in the US. Each handset collects and reports 100's of
metrics of device and user behavior in real time. These metrics comprise 10's
of gigs of data per day resulting in Petabytes of data stored to date.
With our intelligence solutions, the Mobile Operator can for the first time,
analyze system, device and user behavior from every enabled smart phone
handset/device on their network. From this insight, the MNO can meaningfully
improve CAP/OPEX and customer satisfaction.
We need to hire someone to lead our data analysis effort for our ground breaking
solutions. This role would report to our VP of engineering.
[snip -- describing the company]
key phrase: " Our software is running on over 150mm handsets in the US. Each handset collects and reports 100's of metrics of device and user behavior in real time. These metrics comprise 10's of gigs of data per day resulting in Petabytes of data stored to date."
Do the carriers count these reported metrics against data caps? Overage fees can quickly add up I'm not sure I want my phone using my bandwidth in that manner. Granted, the amount is relatively small compared to user-triggered activities (viewing online video, etc), but the point remains...if I'm nearing my bandwidth limit and am consciously trying to limit my data use, but they're collecting and sending out as many metrics as they indicate in real time, that's not cool.
Considering 150 million devices and their quoted "These metrics comprise 10's of gigs of data per day", let's assume 10's of gigs to be 50GB (10's of gigs is less than a hundred and more than 10, so let's go with the middle ground).
50GB spread over 150 million users comes out as ~333 bytes per user and day.
Of course, the transmission of that data is likely more bursty, but even if it transmits all the data in one go, that's only 10K per month.
So your argument about the limit doesn't really fly because even if they did charge for for that data (which they probably do), considering a limit of 1GB per month, those 10k would be 0.001% of your monthly allowance, so it's probably not even detectable by their overcharge detection algorithm.
Now. Don't get me wrong: This kind of malware is really bad and shouldn't be on these phones, or if it is, it should be opt-in for the purpose of remote support.
It's just important that we hate it for the right reasons (security, privacy).
Except that number doesn't make sense. 333 bytes per user per day can't contain the level of information they purport to offer.
333 bytes is enough to send plenty of information, like your favorite apps and how long you used them. Consider something like: "com.android.browser:1.4h;1-800-HI-THERE:2.3min", which is only 47 bytes.
and keypresses too?
Since most Android phones have no physical keys...
(why mention physical keys?)
An average like that takes into account many people hardly using their phone so there is nothing much to report, or using it in a bursty manner so there is nothing to report most of the week but plenty at the weekend.
The "150m devices" claim is rather vague too. It will no doubt include devices that are no longer in use, like when a dating site claims to have X million members without mentioning the fact that all but a few thousand of them haven't logged in for many months - they can truthfully claim such devices have the software installed but that will skew the average bytes/device/day taken from the released figures downwards.
A rule of marketing (which includes selling the company to prospective employees if they are looking for them on linkedin): Never lie when you can selectively use honest statistics instead.