Kmart Australia Integrates “Data Translators” Across Core Business Areas – Strategy – Cloud – Software
Kmart Australia has hired a team of 10 “data translators” who are embedded across its three core operational areas to improve analytics traction and adoption.
At an AI summit last month, CIO Brad Blyth outlined the retailer’s efforts to build its internal data analytics maturity and increase usage of its Sophia data platform.
Sophia – revealed by iTnews last year – is largely assembled from Snowflake, Kafka and Power BI.
Blyth said Kmart is in a constant “three-phase development cycle” for its analytics capabilities – “trying to understand problems and establish solutions; increase that; and then really trying to get value out of everything we’ve built there.
He said the data team had matured its focus over time, moving from creating all of the reports and data models itself, to facilitating the teams to come up with ideas and offering them free capabilities. -service to create reports themselves.
“Like all good tech organizations, as soon as we had access to funding, we went out and started building things, and we had a ‘build it and they’ll come’ mentality that we tried to push through “, Blyth said.
But the company was often indifferent to what was presented to it.
“In the beginning, we really didn’t have a lot of traction,” Blyth said.
“We started to think about what we were missing here, and there was really a communication problem.
“The people who had access to the solutions and who could create them didn’t really understand the problems, and the people who had the problems didn’t really understand what was possible.”
The solution was to hire and embed “data translators” into key parts of Kmart – people whose job it was to understand the issues and articulate them in a way that made them understandable as data issues.
“We have distributed data translators across our three main areas of operation – online, stores and merchandise,” Blyth said.
“They understand this area of the organization. They’re close to the P&L and close to the problem spaces, and they have the skills to articulate it in a way that we can start coming up with solutions and coming up with a hypothesis of how we could potentially solve the problems for it.
Blyth said there are now 10 data translators. He said their impact was immediate.
“The minute we put them on, something amazing happened. There was an unlock,” he said.
“We had 400% growth in three months in our backlog – that’s all the ideas that we collected, potential things that we could pursue and drive value.
“For each of them, we noticed a dramatic increase in the amount of benefits we thought we could seek. It really helped energize this establishment.
With a business-led pipeline of work established, Kmart’s next challenge was scale.
Blyth said when the data team is presented with a hypothesis or requirement, they go away and develop a report or data model to present to the business.
The first iteration was generally around “80% there,” Blyth said.
But the back and forth to tighten it meant the data team couldn’t keep up with the demand for its services.
In response, the team “took a step back” and decided to “shift our service model to a self-service model.”
“What that meant was that we had to go back, look at our architecture; implement controls, new tools and guidelines; the Data Translator function has begun to educate and perfect areas of our organization; and we refocused the data team on a data platform team, not necessarily just a solutions team,” Blyth said.
“We started building different things, focusing on building cubes, data loaders, data catalogs, giving them decent access to the things they needed. It really increased usage.
Blyth said the data team now has to build less to get value or ROI.
$4 Million List Enhancement
He also said the data team was able to experiment with some of Kmart’s key reports and drive additional value to the business that is not possible by simply consuming a report.
Blyth said the data team took Kmart’s store listing report – “which helps store managers understand who they should sign up for shifts” – and ran an automated decision engine in more to see if he could optimize the list beyond what was possible with a report alone.
He noted the company’s challenge: “Store managers are responsible for [rostering] – if they make a mistake in the list, it goes to their KPIs. They’ve been doing it for a long time, so they’re pretty good at it to be honest.
“The problem with an automated decision engine or a model is that it’s a lot more complex than something you could possibly put in a report anyway.
“So the particular model we used used thousands of data points – something that wasn’t really understandable in a Power BI report. It was able to take into account trends in individual stores, trends in industry as well as individual actual personnel tendencies.
“He managed to have a feedback loop so he could correct himself on some of the decisions he was making over time.”
Blyth said that with a bit of historical data and a short learning cycle, the engine immediately produced “a 6% improvement” in listing accuracy.
“Six percent might not sound like a lot — that’s about $4 million — but to give you an idea, it’s six percent with no real trials, no real learnings,” he said.
“So we kept on running [the engine] and that six percent increased.
Blyth said the engine was an example of how the data team could build something once that could drive value over time, without requiring the same ongoing investment of time and resources.
“We built something where the increase in value unlocked itself,” he said, adding that “that’s really driving some of the increase in value.” value” for analysis at Kmart.