1.1 A successful system
We will now look at an example of an IT system that was successful.
Box 1: A successful IT system – pub chain Mitchells & Butlers’ top-to-toe IT overhaul
Graeme Burton
Running a pub is a challenging business, with perishable goods to store and sell, staff to roster to make the most of the British summer (if it turns up) and, most important of all, treasured customers to serve in a timely and friendly fashion. Running a pub chain, therefore, must be an order of magnitude more challenging.
So the challenge of efficiently running the IT behind Mitchells & Butlers (M&B), one of the UK’s largest pub companies with more than 17 different ‘brands’ – as well as one in Germany – must be off the scale.
Partly as a result of having grown by acquisition on the one hand, while divesting itself of non-core businesses on the other, M&B had a sprawling IT estate and, in some instances, inadequate bandwidth into a number of sites.
‘We had a lot of legacy technology and therefore planned to go through this “advancement of technology” process,’ says Martin Taylor, director of business change and information technology at Mitchells & Butlers.
That is why, three years ago, the company decided to embark on a five-year, top-to-bottom IT infrastructure refresh, taking in its data centre, its systems and network topology, the systems it runs at its headquarters and regional offices, retail systems, the level of bandwidth into individual outlets – the whole lot.
At the heart of M&B’s new IT infrastructure is a private-cloud-based architecture hosted by technology partner Fujitsu, which took over from IBM as M&B’s primary technology partner in 2011, just pipping Vodafone-owned Cable & Wireless to the contract.
In the process, IT at head office was almost entirely virtualised, while a new operational data store (ODS), based on Oracle, was installed to take trading data from outlets on a just-in-time basis, radically improving executive access to operational data. Everything connects via the ODS and now head office can see how individual units are trading on an almost minute-by-minute basis.
The project is a radical departure on the old system, whereby much of the company’s core IT was hosted on either IBM e Server p Series (RS/6000) AIX Unix-based systems, or System I (AS/400) mid-range servers, largely hosted in an IBM data centre. At head office, the computer room buzzed with some 20 different servers to handle email, file and print, and some key business applications.
In most cases, existing applications – such as the all-important labour scheduling and supply chain management applications – could be virtualised and were ported to the new architecture. Where applications couldn’t be virtualised, this was normally due to their age and, therefore, a good excuse to ditch them and migrate to all-new applications.
The initial tender process was worked out with a consultancy called EquaTerra, which is now part of KPMG.
‘We asked ourselves what we wanted to achieve,’ says Taylor. ‘What’s the output we want to get to? Then, we worked with EquaTerra on how we would perform the benchmarks, because the issue with benchmarking is that everyone calibrates it in some sort of way.
‘So we said, if we look at the outcome in mind, why don’t we benchmark about how those outcomes have been achieved using EquaTerra’s databases of contracts won and tendered for? That gave us a good feel for what could be achieved and the size of the prize – what it would cost us to do and the likely running costs.’
Once an IT partner had been tendered for and identified, a solid plan for the migration needed to be worked out.
‘We took all the elements of our IT and categorised it into six key areas,’ says Taylor. ‘That meant the data centre, network, outlets, end-user computing, central services and applications. Then we looked at the outputs that we wanted from each of those areas, the cost base and the areas where we wanted to get our innovation from.’
Initially, the data centre and the network were prioritised as they provided the platform that supported the rest of the business – the applications running in the data centre, while the network supports the outlets, which in a modern hospitality business also require 24/7 logistic and IT support.
Infrastructure renewal was not the only consideration. M&B was also keen to consolidate applications and, as far as possible, to share applications across its business units as far as possible.
‘Our finance system, human resources, payroll and property management systems are common throughout the organisation,’ says Taylor.
‘It doesn’t matter whether it is a Toby, Harvester or an All Bar One, we get the synergies of having one application and streamlining our processes so that they are common throughout. Occasionally, we will have to use something slightly different for the odd brand, but in the majority of cases, everything is consistent.’