Since talk about Facebook and data responsibility is abuzz right now, let’s jump in on the conversation. After all, data responsibility plays a huge role in market research, and we want to make sure we are using data to benefit consumers, instead of…well…creeping them out.

What happened?

Let’s start with a quick review. What did Facebook and Aleksandr Kogan do, and why was it wrong?

In short, Aleksandr Kogan created a personality quiz app that granted access to users’ public-facing Facebook data including ID, name, first name, last name, gender, link, locale, timezone, updated time, and verified.

The main scandal came within the API’s permission group “Extended Profile Properties.”

These endpoints gave access to data about your friends—data your friends may not have granted permission. This data included everything from “about me” and actions to website and work history. Another endpoint allowed the developers to even access private messages.

Kogan then provided that data to Cambridge Analytica, who used it to develop techniques that could influence voters.

Yikes.

Why was it wrong?

The problem with this data debacle doesn’t lie in gathering data. In fact, it’s awesome to gather data on people (when they explicitly grant permission). Data collection is helpful for both brands and consumers when done correctly

The problem was they used sneaky tactics to collect data on app users and app users’ friends and family…data the friends and family didn’t grant permissions for access. Then, they used that data to try an influence voter outcomes.

So, what do we do as data scientists?

As aggregators of data, we have a responsibility to protect consumers, explicitly state what data will be collected and how it will be used, and make the consumer feel secure about data collection methods.

How to personalize ads without being creepy

Here are some top ideas on how to personalize ads in a helpful and ethical manner:

  • Get permission to gather data. This should go without saying, but if there is anything to learn from Facebook/Cambridge Analytica, it’s that you need permission. When consumers know their data is being collected and how it’s being used, it doesn’t come across as espionage. In fact, consumers are often likely to share personal data in exchange for more personalization.
  • Be sensitive. Just because you have data on a consumer doesn’t mean you have to use it. For example, if there is no reason for you to personalize on location, don’t call out your users out on their location. Similarly, you may know what their interests are, but don’t shame them or make fun of them for their interests. Remember when Netflix did that? It didn’t go so well for them.
  • Focus on relevance. The more you can to make your messages relevant, make the shopping experience more fun, and make life easier and more convenient for your users, the more spot on you know your strategies are.

Remember, data collection is awesome when used with explicit permission and when used to improve the lives of your consumers.