From Data to Signals: The evolution of customer insights

I’m excited about the shift from data to Signals because it represents an opportunity for businesses to improve customer experience through modern data practices, while meeting their own privacy needs and those of their customers.

It’s no secret that companies of all sizes are interested in acquiring and using consumer data. They go to great effort, incur significant expense, and take on meaningful risk in order to get and process that data, and the reason is simple: being able to make better choices in interactions with customers can be tremendously valuable for both companies and the people they serve. Used well, consumer data can inform those better choices in a number of ways.

My mission at is to build products and technologies that let businesses and their customers have better experiences through the power of ML, while preserving privacy and avoiding reliance on internet-wide surveillance of consumers. It’s an honour to be working with an amazing team that rejects the idea that business value and privacy are mutually exclusive, and I’m excited about what our Signals-based approach makes possible. Signals can provide substantial business impact, while using resources like data scientist and marketer brainpower more effectively – and keep everything safe from a privacy perspective while doing so. We’ve all seen how powerful, and how dangerous, it can be for companies to learn from each other’s data, and I have first-hand experience from my time at Mozilla and Facebook with the challenges of doing so in a way that keeps individuals safe. I’m proud to be working on powerful ways to help businesses make better use of their data, gain insights from the experiences of other companies, and avoid the risks that can come with sharing PII.

When seeking a better understanding of their customers’ needs, businesses often reach for more data, including from 3rd parties or data brokers. Working with this data can present a number of significant challenges, including data usability, privacy risks, and limited data insights.

It’s often the case that 3rd party data can only be connected to a business’ own information on a few crude elements: email addresses, phone numbers, zip or postal codes, or a global “pixel” tracking identifier that causes so much privacy concern. This requires that both the importing company and the source companies have significant overlap in their customer base, and limits the applicability of many interesting data sets. These identifiers tend to also be quite sensitive and therefore challenging to process, and having them can put substantial compliance and security burdens on the receiving company!

In addition, the companies providing these data sets usually need to scrub them substantially to avoid disclosing sensitive business details, such as specific pricing or transaction contents. While these concerns are eminently reasonable – privacy is a concern for business information, too – they can dramatically reduce the value of insights available to a company that buys access to these data sets.

Our Signals approach takes a different tack: instead of relying on privacy-invasive identifiers to connect data from different sources, we use ML to identify important insights into behaviour patterns that are manifest in a set of de-identified data, and then to connect those insights to a company’s own first party data. Our algorithms identify consumer behaviour patterns in each data set, and construct models that can apply those patterns in the context of another company’s operations. This allows us to correlate group behaviours observed by different organizations, without requiring (or knowing about!) overlap in those organizations’ customer bases, and with significantly reduced privacy impact for the consumers who are represented in the data sets.

Merely collecting more data doesn’t lead to better understanding of your customers, and can lead to significant privacy issues that reduce data utility and add operational overhead and liability. Working with Signals instead provides you with ease to use insights from your own and other companies’ consumer data, while protecting the privacy of all the customers involved. It’s time for your company – and your customers – to get the benefits of a modern, Signals-based approach to data-driven customer experiences.

To make third-party data easier to use, and help you build more powerful models, we’ve created a tool for data scientists on our Trusted Signals Network (TSN) as part of an Alpha program called Project Sandman. Want to join Alpha and see how easy it is to work with Signals?

Click here to learn more and sign up.


Want to learn more?

Responsible AI is part of our DNA. It underscores everything we do and is something we know a lot about.

Let's Talk