Case study: How Vista Retail uses data to keep the tills ringing

Vista Retail is an instore technology company which supports clients in the retail, leisure and hospitality sectors through the supply of instore and point-of-sale technology, as well as back office infrastucture and peripherals. Everything needed to ensure a retailer is able to process transactions, in fact.

It is the largest independent retail services provider in the UK, with 120 clients - including Ted Baker, Costa Coffee and Pets at Home - who operate across 16,000 sites using 64,000 electronic points of sale.

A critical success factor for the company is keeping its clients’ operations up and running - any POS failure means no money can be taken. As Phil Budden, operations analyst at Vista Retail, explained to DataIQ, data now plays a central role in this. “We extract data on fault reporting from the hardware down to component level. If a retailer’s system goes down, that is non-trivial,” he says.

Vista Retail operates via field engineers who are assigned jobs and order replacement technology (when required) from a series of warehouses around the UK which are in turn supported by a central warehouse. A core system supports this operation and had been generating conventional management data. But as Budden says: “The business was keen to leverage more data and see where it could find efficiencies.”

The existing SQL-based business intelligence tool had supported the previous demand for reporting, but was not capable of taking the company’s data and analytics to the next level. “We were looking for something better that was also easy to use,” recalls Budden.

Alteryx proved to have the capability and flexibility which Vista Retail needed, from extracting data from the operating system to presenting it and supporting drill-downs by users into specific issues. In particular, it offered automated data blending which removed a lot of manual workarounds. Says Budden: “Our operations team holds data in one system, finance in another. We have data stores in SharePoint and Excel. The business wanted to be able to look at all of those sources and blend them together for analytical output.”

Before adopting the Alteryx solution, many key business intelligence tasks were labour intensive. Comparing data sets across a time series, for example, had to be done by extracting data for each time point and then compiling them in Excel. With the new data wrangling solution, this is a simple one-touch query. Generating better time series reports has also allowed Vista to reduce its spend on Profile from £8,000 to £2,500 every week. Automating routine recurring reports has further saved an estimated 315 hours a year in human resource.  

Repair scheduling is one of the main challenges for the business. “Previously we were doing it using a moving average model using data on faults we had been seeing over the last three months and using that to predict expected repairs in the next three months,” says Budden. While some demand levels looked highly predictable, such as Christmas being very busy for retailers, the level of detail was limited. “Once we started to look at that using Alteryx, we could look at faults over the last year or two and identify that Christmas was not, in fact, our busiest time. Instead, the Summer holidays are extremely busy, while for specific retailers, like betting shops, it is events like Cheltenham and the Grand National,” he says.

Pull-through on stock from these different demands has allowed Vista Retail to optimise its just-in-time stock availability, raising it from 88% of target to 97%. As Budden explains, this brings significant operational gains: “Where possible, our engineers will do a repair on-site, but if they can’t and something needs replacing, that item will be sent back to be repaired by our in-house engineers before it gets returned to stock. That previous stock target was not relevant because it was based around Christmas when that was not the critical point in the year.” 

With the business aiming to hold one week’s worth of stock at any given time, hitting target levels when those items are really needed has a big impact on the business and its clients. It means Vista Retail can deliver higher service levels, such as sending out parts on the day they are ordered - the number of parts not sent in that time frame fell by 67% between July 2014 and July 2015, while the number of unfilled orders each day has fallen from an average of around 100 to around 16, with an all-time one-day low of just six. 

Budden has every reason to take personal credit for these kind of uplifts since, when he started in his current role three years ago, the entire BI and analytics team was just him. Before that, he had worked for seven years at Vista Retail as an engineer. He has since added three team members as the demand for intelligence and insight has grown with the new capability. “It feels good to have been leading that and to be responsible for those changes,” says Budden.

The next step is to loop that intelligence back into the operating systems so that, for example, call centre staff get visibility on stock levels and engineer locations so they can advise clients when a repair will take place more accurately. “We want to have a real-time view for the call centre and Alteryx is making that possible,” he says.