Personalization with Privacy — It’s possible

One of the biggest problems faced by internet-tech industry is how to balance personalization with privacy.

I was thinking of a way which could potentially make this happen. For context, currently a very small portion of data (cache) resides on the client-side server on our desktops, tablets and phones. When we search for goods to buy online on Amazon for example, Amazon uses our cache to identify the user, looks up our previous purchase and search information which resides on its data-centers (server-side) and then using that data to suggest a personalized list of items it would like us to buy.

I would tweak this model a bit. If we’re able to segregate the meta-data about the user from platforms such Amazon, Google, Facebook, etc. and keep it on the client-side instead of the server-side, then every time a user lands on Google for example, Google will run the server-side code to display its basic website features. On top of that, it will identify the client-side data and use it to present a personalized web page to the user. This will have one big disadvantage in terms of customer experience— it will slow down the speed of loading the personalized web page. It will have one big disadvantage for the companies — they would not be able to analyze customer data (their search and purchase patterns) in the way they currently can, and would not be able to use cookies to cross-advertise products on similar websites using ad networks such as Google Ads.

Few things will have to be kept in mind to implement client-side storage of personalization data. It would have to be synced across multiple devices including desktop, tablet and phone. To achieve this there could be two ways — one, all devices sync with each other using wi-fi or 4G/5G networks and keep data updated at all times to not hamper the user experience. Two, we could establish a private cloud for each user — which means any one of the devices such as desktop, home assistant, mobile, etc. which could provide storage and processing power to process the client-side storage and programming.

There would be one other disadvantage — users would not be able to login from any foreign device and still experience the same website version as they are able to do currently.

I believe even with the user data residing on client-side, Google and Facebooks of the world should be able to target advertisement to users thereby not directly impacting their ad-based business model.

The one thing to ponder from the consumer standpoint is- how much do we value privacy as opposed to a seamless experience of personalization across multiple devices.

Despite all of this, I think there is immense value of user data residing in data-centers of the big tech players for their product and business analytics needs, because of which I would never expect big tech players to initiate a move to move user data to client-side. This means if we truly want to make this giant shift happen, we would need a lot of regulations such as GDPR, to be able to shake the big tech companies enough to care about data privacy.

Early Adopter // Amazon, Wharton, Bain