Airbus Operations has shed some light on the business challenges of calling time on a long-standing IT contract, after moving away from an outsourcing deal with HP.
Speaking at the Gartner Datacentre, Infrastructure and Operations Management Summit in central London, the aircraft manufacturer’s head of hosting services, Peter Radig, said the decision to wind down its 12-year contract with HP was made two weeks after he joined the firm in 2012.
“The contract with HP was 12 years old and was renewed three times, with exclusive negotiations that led to a very difficult contractual situation, as well as a weak operational situation, because as that grew we never changed the definition of the statement of work,” he said.
This also led to the emergence of some accountability grey areas, as situations arose where neither HP nor internal staff at Airbus knew who should intervene when problems occurred.
“We had a situation where HP would say this is done internally, and someone else would say it was done by HP. It was very uncomfortable,” he said.
To resolve this, the company decided a fresh start was needed and issued a request for proposals (RFP), the overarching aim of which was to reduce some of the operational complexity that had developed within the Airbus IT estate over the years.
At the time, Airbus was operating 12,500 servers, all of which would need to be migrated over to a new provider, as well as 2,900 databases and around 670 applications.
All of this IT contributes towards the creation of its aircraft, which can – in the case of its A320 vehicle – take up to nine years to build using more than 2.5 million parts procured from around 16,000 suppliers, Radig explained.
To steer the company through the RFP and tendering process, it enlisted the help of a third-party sourcing advisor, ISG.
“We wanted to send a strong signal to the market that we were serious this time. We didn’t have the best reputation in the market [when it came to procurement] as we would always ask for an RFP and then get people to work to that, and then enter into exclusive negotiations with HP,” he said.
Having ISG there meant prospective suppliers also had an independent party they could quiz and seek assurances from about the sincerity of Airbus’s intentions, said Radig.
“We also wanted to include benchmarks in the contract, but had no idea how to put those in, and – finally – the time schedule we were working to didn’t allow for an in-house way of working,” he added.
Appointing the successor
Airbus eventually appointed Atos as HP’s successor in July 2014, and migration of its systems from the latter to the former was officially completed in June this year, but – Radig admitted – it is still something of a work in progress.
For the record, HP still has a presence within its IT infrastructure, as its technology is still used to underpin the company’s HPC and SAP activities.
“What we’ve learned now is that the transition phase is not the end of the transition. Officially, the transition phase ended in June, but we’re actually still working on making it work,” he said.
For enterprises considering embarking on a similar move, Radig said there are a number of steps the IT team should take to ease things along.
For example, his advice is to start documenting some key performance indicators (KPIs) for the incumbent provider at least a year before the migration starts.
“Your internal customers will not believe the service is as good or as bad as it used to be before, but you can show that in KPIs, and that can be very helpful,” he said.
“Otherwise everything is based on gut feeling, and [KPIs make it easier when] you need to make your decisions on priorities.”
Radig also warned about the potential for “brain drain” in projects such as this, as – at the first sign of change – it is not uncommon for the IT people involved to start looking elsewhere for work.
“As soon as you start announcing the contract is being changed, the best guys will leave first,” he said.
“You also don’t want to be in a position where the incumbent knows they are going to lose the contract when the first good guys are starting to go.”
Furthermore, it is also important to ensure that a thorough audit of the IT estate is undertaken before it is decommissioned and handed over, he warned, to ensure there are no nasty surprises later down the line.
“One of the biggest challenges we had was we started our decommissioning projects after we defined the numbers and sizing of the contract,” he said. “My team did better than expected and we ended up with fewer servers, which didn’t make Atos happy.”
This was easily rectified, he added, as the Atos contract allowed for some degree of fluctuation. Without a clause like that in place, the company may have had to renegotiate the entire contract, which is why this type of due diligence is essential.
“My key message is that a change like this is doable. You need to have an excellent team and top management attention, but it is doable,” he added.
Read more datacentre case studies
- Irish telco Eir shares details of its datacentre consolidation project and explains why investing in a robust UPS is a must.
- Tesco is six months away from completing a three-year programme to deliver a single datacentre operation to support UK and international operations.