After frustrations with traditional approaches to big data — and trying to gain a 360-degree, understandable, actionable view of each member — the credit union re-engineered its whole data storage and analysis process using what it calls an analytical data utility (ADU).
Tom Wilson, OneAZ’s senior vice president of enterprise data and analysis, says the credit union coined the term to describe its approach to providing associates easy, self-service access to data sources.
That self-service ADU approach uses a data lake and integrating software for data cataloguing, enrichment, analysis, reporting, and connecting to other solutions.
To the non-technologists, that means the ability to target marketing campaigns far more easily than, say, sifting through 15,000 leads in Excel spreadsheets. To the IT staff, that means more time to focus on other priorities.
OneAZ has undertaken a new CRM strategy that puts member experience at the top (as the illustration above shows), with technology tactics supporting that experience.
“Business stakeholders don’t know everything they need to know today, and there’s too much dependence on IT to know all the use cases,” Wilson says.
At the root of the problem are the many disparate data sources that don’t integrate well. Plus, according to the senior leader, the data quality itself is “always an underestimated problem.”
OneAZ’s new integration software solves the problem of data types, and the CRM-ADU team now delivers the information from the data lake to business line analysts and to member service associates in a format that provides them with a complete view of the member.
Marketing staff began using the new tools for campaigns a few months ago, and the call center now uses them for member support. Up next is rolling the functionality out to OneAZ’s network of 21 branches.
Wilson says the credit union expects the new approach to power a $2.5 million boost in net income per year through marketing campaigns that will generate new members, more business, and better retention at projected growth rates of approximately 10% to 15% a year.
Wilson says the entire process is secure, transparent, reconcilable, and repeatable, giving the credit union long-sought ability to really understand who specific members are and what they really need or want from the credit union.
“As our analysts adopt ADU, it is becoming foundational for our reports, dashboards, and analysis,” Wilson says. “This moves us toward OneAZ Credit Union’s goal of one version of the truth.”
The CRM-ADU Team at OneAZ Credit Union includes, from left, SVP Tom Wilson, business and data integration analyst Robert Escalona, senior business intelligence analyst Rachel Smith, and branch operations support analyst Tim Hatch.
Wilson came aboard in 2015 as OneAZ was building an enterprise data warehouse that could work with its existing CRM solution. As is often the case, Wilson says, it didn’t work. Too many data types made importing and processing information slow, awkward, and inaccurate, and the work was thwarting the primary objective.
Wilson and his team were already researching the ADU concept and working with a software integration specialist when they switched gears and decided to abandon the EDW and go with the cloud-based data lake concept.
Wilson says EDW projects typically take 18 to 30 months to fully implement, with a failure rate of more than 50% or significant reduction in scope. That’s not the results OneAZ got with the data lake.
“We implemented our data lake in seven weeks,” he says.
After that, the credit union completed the entire CRM project in four months.
“Now we’re delivering on more than two dozen initiatives and measuring their effectiveness,” the OneAZ SVP says. “That would be impossible without our platforms working together.”
OneAZ Credit Union uses Unifi Software and Microsoft’s Azure cloud and Dynamics CRM to manage its data lake in a Hadoop open source framework. Find your next solution in the Callahan & Associates online Buyer's Guide.
The data lake approach also is costing approximately one-fourth of the spend expected for the original EDW project.
According to Wilson, setting up the data lake and integration processes went quickly; however, it was complex and can be unforgiving if designed incorrectly.
To avoid that, Wilson suggests securing complete management buy-in and then plan, plan, and plan. Initial planning should include business users, data experts, and platform engineers to document all the data requirements, dependencies, and flow.
Then, on the front end, the business users and analysts work with the integration interface.
From there, the data goes to front-end associates.
“Then they can interact in a more efficient way with our membership,” Wilson says.
It’s as easy as jumping in a lake.