Being one of the world's largest offline media consumption repository, Zapr understands and embraces the responsibility of using and analyzing data collected from millions of smartphone users across the Zapr user-base.
We firmly believe in the philosophy of 'privacy by design' and have upheld this from conception to execution, keeping user privacy protection at our very core. With this policy, we intend to answer all your concerns on our philosophy of privacy and you can always reach out to us for any further queries.
Zapr is a media-tech startup, that leverages proprietary audio-content-recognition (ACR) technology built across digital signal processing, big-data analytics and mobile ad-tech, to analyze offline media consumption. Think of us as the 'Shazam' for all media – live TV, music, radio, movies-in-theater etc.
However instead of having a separate user-facing app, our technology is packaged into an analytics Software Development Kit (SDK) and integrated into numerous apps, through partnerships with dozens of app publishers.
Once our partner app is updated with our tech, the users are asked to share microphone access with the app, in order for the app to identify and analyze offline media consumption. The users are explicitly informed of the same, and they can choose whether or not they wish to share their media data with the app.
We'd like to be really clear here. It is literally impossible for Zapr's technology to listen in or record any conversations. Today, there are a lot of voice-recognition technologies that we interact with – Siri, Alexa, Google Assistant etc. However, the tech that powers Zapr – audio fingerprinting – is starkly different from these, and is in fact similar to music-detecting services like Shazam, Soundhound etc. Alexa actually needs to 'listen' to you to understand how to respond. Zapr does not.
Zapr's tech converts actual ambient audio into a tiny digital file, called a fingerprint, in real-time. This is a critical point, because it means that at no point in time, ever, is any audio actually saved on the phone or sent to any server. Moreover, the fingerprint itself contains less than 0.1% information from the actual audio, and therefore it is technologically impossible to replicate audio from it.
Finally, these fingerprints are sent to our servers for analysis, where we run them against pattern recognition algorithms against original media content – TV shows, movies, songs etc. to see if there is any match. And that's it.
On the server, if the fingerprint doesn't match anything, because perhaps there was no media playing around, then the fingerprint is discarded because it's useless. If the fingerprint does match something, for e.g. the IPL match, then we attribute that cricket match to the advertiser ID of the phone. And then delete the fingerprint after sometime.
At Zapr, we hold ourselves to an incredibly high standard when it comes to the privacy of our users. So to start with, we don't even know who our users are. We capture no Personally Identifiable Information (PII) about any of our users – no names, phone numbers, email addresses etc. The only ID we link to our user, is the anonymized advertiser ID (this is a non-permanent ID provided by Android and iOS to each device). The only data that our tech is capturing through our partner apps, are the TV shows, movies, songs etc. airing offline around the phone. Apart from media consumption data, linked to an anonymized ID, we also get the following basic Android data points such as – device model type, location (if the partner app has permission for it) and telecom operator, etc.
Our stringent security measures (physical, electronic and managerial) protect against loss, misuse and alteration of the information under our control.
Our data is secure – all our servers are hosted on AWS that conform to ISO 27001, 27017, 27018 standards. We also ensure adequate data encryption at the time of data collection to render the information collected as undecipherable.
The audio fingerprints we collect remain in the same format. The fingerprints are NOT reverse engineerable and are sent to Zapr servers securely. Our precise algorithm ensures that only unique aspects are captures and the remaining data is immediately discarded.
We never analyze your audio for speech, language or keywords.
Only authorized personnel have access to the website's and integrated app's servers. Your user information will be shared with them only on requirement basis.
We have third party associations, including advertisers and partners and we share your information with them only on an aggregate level to understand your TV viewership patterns. We share your ad IDs to provide you relevant ads on mobiles.
We may also share your data with some companies who advertise through our apps and may require you to provide your personal data. Any third parties' use of your personal data is governed by their own privacy policy. To learn more about behavioral advertising or to opt-out of
some types of behavioral advertising for participating ad networks, you can visit the Network Advertising Initiative (http://www.net- workadvertising.org/choices/) or the Digital Advertising Alliance(http://www.aboutads.in- fo/choices/).
Zapr has also partnered with the industry's best research bodies such as Nielsen and Kantar Millward Brown to enhance our partnered research offerings to our clients. You will always know when you are a part of a survey and the response you share is always kept anonymous. The user information data which you might choose to share as a part of a survey is shared at an anonymous, aggregate level and never at an individual level that is personally identifiable.
For demographic, behavioral, and geographic ad-targeting or to provide you with specific, localized services. For example, a sports gear brand can reach all IPL viewers, no matter which part of the country they are in.
Analyse your TV viewership patterns and interests, and share aggregate data at a city/state/national level with the media & advertising industry, to identify trends and consumption patterns. For example, out of 100 people who watch IPL, we can identify those who also watch English Premier League for insights on 'sports' audiences.
Conduct surveys that help understand the impact of a TV show, channel or the effectiveness of a brand TV Commercial. For example, out of every 100 people exposed to a new car commercial on TV, we can identify the number of people who can recall the same commercial.
Zapr has deleted ALL historical user data of EU users from all SDK versions integrated with publishers.
EU users are detected by SIM country API and/or IP address. Zapr does not collect or store any data of these users.
Publishers are recommended to upgrade to the latest Zapr SDK version 3.2.0 so that they can proactively stop the SDK from running on the devices of EU users, and ensure that Zapr cannot collect or store any user data.