Data Experience Design Changes Product Development at Nasdaq
In 2016, Nasdaq’s search to add data expertise to the product design team led us to a hybrid role between data visualization and user experience design. We called this role “data experience designer.” If you’ve never heard of that title, you’re probably not alone.
It doesn’t bring you LinkedIn popularity like Data Viz or UX design, and as far as I know there aren’t any conferences or courses explicitly targeting professionals of data experience design, yet this hybrid role helps us tackle very specific challenges we face designing financial-data products at Nasdaq.
So what is data experience design, or more specifically, how is it different than data visualization or user experience design? In short, a data experience designer considers the full data experience — not just the chart or visuals used, but the real-world questions the customer needs the data to answer. The designer must consider everything from how the data are collected and structured to the actions a customer needs to take once their questions are answered. From this holistic view, we can start to tackle challenges that plague data-centered products, like siloed data, meaningless dashboards (with charts that often don’t say anything at all), and interfaces that make essential comparisons or actions impossible.
This is not a skill set every company needs, but there are a few reasons data experience design is important at Nasdaq, specifically our products, customers, and data. If your company or teams face similar challenges, adding data experience design might be a valuable investment.
Nasdaq is a financial technology, trading and information services provider, operating 26 exchanges globally, with a presence in every capital market. Our products support the entire trading lifecycle — pre and post trade, surveillance, governance, risk and compliance, and market intelligence. Clients include exchanges, clearinghouses, central securities depositories(CSDs), brokers-dealers, regulators, and public and private companies. Tackling data experience is critical at Nasdaq, where the value in many of our products is in delivering data insights that give our customers an edge. Our products are judged quickly and harshly based on this value proposition, so by extension, these challenges are priority for the product design team.
Nasdaq has a huge suite of products across multiple business silos, with different technology stacks and diverse customer bases, which can lead to very disjointed experiences from one product to the next. Along with tackling product-specific data challenges, we are using data experience design to improve the customer’s interactions with the data across our products, looking at things like defining universal chart patterns, unifying interactions, and designing with principles of data flexibility and fluidity at the forefront.
Our customers have high standards for data consumption in our products. It’s not that they don’t want to be delighted with elegant icon animations and beautiful transitions — it’s that first and foremost they need to extract value from data. Our customers care about things like latency, data density, source quality, compatibility and exportability, and they often don’t mind using a tool that looks like it was designed in 1999, as long as meets those data-consumption standards.
The primary way our products delight our customers is by painlessly delivering the right information in a digestible, understandable format. While this is 100% within the purview of traditional UX design, we are taking some data-specific approaches at Nasdaq to do this more successfully for our customers.
One way we do this is by adding data interviews to our customer research. This pushes us beyond “tell us what data you need to see” to a conversation about specific stories in the data — what context is essential, what are your thresholds and limits, and how do you spot risk? Since displaying the data in different forms brings out vastly different stories, data sketches asking the customer to respond to these different views early in the process help us find our “a-ha” visual, and avoid dooming a product with a bunch of meaningless charts. Nathan Yau at Flowing Data has a great article on data sketching here.
Prototyping with real data that tells a real story is essential to delivering on our promise to create the best data experience possible.
To say data at Nasdaq is complex feels like an oversimplification. From quantitative tick-level exchange data, to qualitative meeting ratings, big data to small — the sheer volume is so vast that I haven’t even begun to take a full inventory of all of the data we deliver to our customers. But we have added a data deep-dive to the discovery phase of new projects. That requires the design team to genuinely nerd out about the data, learning not just how and why it is used but the real-world patterns the data represents, the limitations of the data, and the nuances between different data sources.
Prototyping with real data that tells a real story is essential to delivering on our promise to create the best data experience possible. When a customer or stakeholder believes the data in a prototype is fake, there is a tendency to gloss over or even ignore a visual. This can be crippling for a data-centered product, especially if the entire UI is a data visualization. When the data is real, people search for meaning in the visuals, and we can elicit the feedback we need to create the right product.
Sometimes we learn that we need to go back to the data structure, and bring together data from different databases for the story to be useful, or slice our aggregates into more representative buckets for realistic comparisons. Without applying a data experience design process to the data, this feedback can come too late in the process, when patches and feature requests are the only solutions.
I’d love to tell you we started out of the gate with all of this, but since my background is in data visualization, I have had quite a learning curve (and there is still a lot of room to develop). If you are adding data experience design to your team, it may be helpful to hear the phases we have gone through here at Nasdaq.
Phase 1: Reworking Visuals for Specific Products
When I started at Nasdaq, I was assigned to redesign specific charts for specific products. While this limited the data experience contributions I could make, there were several benefits to starting this way. Not only did these projects offer quick wins and a fast-track to delivering a return on investment, but they gave me a chance to learn the company, data, products, and customers. A lot of this work involved more traditional data visualization tasks like improving the data story of specific visualizations, or adding an analytics section to an existing product, but this work has formed the foundation of our data experience pattern work.
Phase 2: Data Experience Patterns
After working on a handful of existing products, we’ve started to see some successful patterns emerge, so the next logical phase has been to try and record these patterns in guidelines and documentation so we can scale and validate both within our team, and across other teams.
It’s not realistic to think a strict style guide could work across a such a diverse suite of products, so we have focused on providing pattern resources and high-level guidance for our internal design teams, developers, and product teams.
This includes things like:
· Data experience principles
· Style and interaction guidelines
· Best practices
· Code and design resources
Future Phase 3: New Product Ideas (with existing and future data)
With a solid inventory of data, and patterns defined, we will be able to turn our attention to the data itself, exploring the stories in existing datasets to design new features or product ideas. While it can be harder to carve out resources to work on projects like these, the upside potential is significant. We have already learned some surprising things about the value of UIs for products that are strictly data feeds, useful so that the customers can validate the value in the data.
We will also investigate ways to generate and acquire new data for our customers, leveraging new technologies to improve transparency and insights within our products.
Future Phase 4: Internal-facing products
While it may be the hardest sell of all, improving the data experience design for internal data would yield huge benefits for most companies. We have already seen some potential projects around customer insights, product feedback and releases, product migrations, and sales targets. This is an area we will continue to explore as our team grows.
Although data experience design is still in its infancy here at Nasdaq, it has already added some valuable tools to our process, helping us bring out essential data stories and workflows across our products. If your team is on a similar journey, we’d love to hear what tools and resources you are using to tackle data experience design.