We tend to think of UX metrics as tools for seeing how people use our designs. However, there’s real power and magic when we take data we’ve collected and put it back into the designs we’re delivering. Sometimes, a small addition of behavior can be worth billions of dollars of additional revenue.

In this live discussion, we’ll explore how we can discover hidden value for our users, by looking into the data we collect. We’ll look closely at examples of how this is done well and talk about how you could mine your data for undiscovered value.

Notes for today's discussion:

Value Discovery Metrics

What data do we have (or could collect) to provide additional value to our users' experiences?

Definitions:

A measurement is an observation of a tangible quantity.

A metric is a measurement that we track.

A analytic is a measurement that computers track.

We use metrics to detect and report important change.

For Value Discovery metrics, we're using the data inside our designs to enhance the users' experience.

UX leaders often only think of metrics as for reporting goals and progress.

Value Discovery metrics are different, as they are more for the users' benefit than the team's.

Examples of common Value Discovery metrics:

• Amazon's Customers who bought this have also bought? recommendation engine
• Spotify's Weekly music discovery tool
• Waze's traffic routing capability
• Netflix's recommendation engine.

Search engines have been using value discovery metrics since the start.

We can think of Value Discovery metrics a design medium.

We're designing with data.

As designers, we need to understand:

• What data is available to us?
• What data isn't currently available, but could be?
• How the data relates to other data we already have?
• What is the value within the data we could surface?

Value Discovery metrics create operational efforts.

Data maintenance and cleansing is resource consuming.

Value Discovery metrics will play an increasing role as artificial intelligence and machine learning capabilities become more commonplace.

Teams will need increased data science and machine learning skills.

Designers will need to understand the underlying data schemas and architecture.

Ethics require we focus on improving people's lives.

Build the metrics for UX outcomes, not pure business outcomes.

UX Outcomes: If we do a good job on our design, how will we improve someone's life?

Designers will need to surface how the algorithms work.

We'll need a new framework for the ethics of using this data this way.

Not everything can be solved with an algorithm.