Vista DnA launched data product teams to revolutionize the way analytical insights are generated and scaled for internal customers and transform the user experience of small businesses everywhere. Design Data and Analytics Lead Amy Edwards has a birds-eye view of progress so far.
We’re democratizing data by offering data products to all. It’s an evolution from primarily offering analytics as a service and using machine learning models to elevate the customer experience, inside and out. But how do you perform and transform at the same time? It’s a challenge all DnA teams at Vista are working through.
What’s the difference?
Vista now has two types of data analyst:
- Embedded within a Vista product team – which may focus on producing a web page on the Vista site, for instance. These analysts or ‘analytics consultants’ provide analytics-as-a-service to teams across Vista.
- Part of a cross-functional data product team with data engineers, scientists and product managers to develop scalable, self-service analytics products – AKA, the big structural shift.
Their work is complementary and interdependent.
A cool kind of ecosystem surrounds this setup:
- A data product team gets busy delivering data products that make life easier for the embedded analyst and the Vista product team they support. Up close, you’d see new access to improved data automating tasks like data cleaning and visualization, with algorithms revealing insights that simply wouldn’t be available otherwise.
- The embedded analyst reciprocates by using the data products and supporting stakeholder adoption of these data products. They also define requirements for recurring needs that go into the data product team’s backlog.
The payoff? The data product team lightens the embedded analyst’s load, freeing them to focus on complex, higher-value business questions.
What’s a data product anyway?
Here’s a quick, broad sweep:
- Curated datasets by domain, primarily for analyst use.
- Self-service dashboards for all. These empower specialist and non-specialist Vista audiences to access, visualize and derive insights that inform decision-making.
- Algorithms that leverage vast customer data – largely for digital product teams working on tools like recommendation engines.
The big reframe
Establishing data product teams has reframed data and analytics at Vista from the top down. Data projects have become data products, ripe for continuous improvement. It fits with how we think of ourselves now: as a customer-centric technology company.
Why reform was needed.
There were natural constraints to being solely in project mode:
- Analysts doubled as data product managers: embedded but without the bandwidth to debug dashboards to make them fully scalable.
- Data scientists were also embedded but didn’t have data engineers on hand to ‘productionize’ their algorithms.
- If a proof of concept caught on and became a ‘data product’, it meant crossing your fingers that you could keep it from breaking. It was short-term thinking.
What productizing data really means for us.
Vista DnA can now create sustainable offerings and make room for constant enhancement at scale. Creating that reality meant breaking down silos between data analysts, engineers, and scientists with multidisciplinary teams. Strategically: restructuring to make data product teams fly shows our determination to become a standout data-driven business.
Bridging the gap
The new design domain was established in 2021. Up-and-running data products are vastly outnumbered by the amount in development (stretching to 100+ data products across all DnA teams). So we have to be thoughtful about how to meet business needs and what gets prioritized. That can mean choosing between development of a data product or giving ad hoc support through an embedded analyst. It’s a fine balance as we perform and advance in parallel.
Evolution is thirsty work.
We have so much data to drive value and fantastic customer experiences from. Basti, our CDO, likens data to water: the resource of growth and vitality. Like water, we want data to be readily available for everyone – knowing that requires scouting more sources on a constant basis.
I now think of DnA as setting up for exploration ‘out back’, with our embedded analysts charting new territory. What happens when we find a new world? An exciting location oozing with potential? It’s time to build a house: get proper plumbing, add real infrastructure. Here’s when we call in our data product teams to get this prized resource flowing with ease and scale operations. Imagine curated data tables as the water pipes, dashboards as faucets, algorithms as lawn sprinklers – none are strictly necessary, but they sure make life better.
Investing in scalable, easy to maintain solutions upfront is enabling DnA to focus on producing more advanced analytics. Soon, algorithmic data products will make Vista designs super-simple to find, powered by semantic search and computer vision. Today, our dashboards already mean Vista stays ahead of which styles customers love and what trends inspire them, so we can offer what small businesses want on tap.