The practical experiences of organisations who have decided to change their businesses, especially in the area of IT, are always of interest to me as learning from someone else’s successes and failures  is common sensical. At the last ACORD Club meeting, Graham Card from Willis gave a presentation on Willis’ experience on changing their internal processes and implementing e-Accounting. I felt that this presentation raised some interesting points on what worked and some apposite observations on what is needed to make something like e-Accounting succeed, so in the spirit of learning from other’s experiences I thought I would make this the subject of this months blog.

The list of reasons why Willis felt it necessary to change aligned with what we at Scyllogis have been saying for some time now. Point one was that the industry as a whole needs to lower its service costs; no surprise there, but what I found a surprise was that Willis felt that location savings are not the answer. This was something that was discussed at the round table discussion on BPO that I chaired at the last Insurance Day Summit, where I raised the opinion that surely you can’t outsource a fundamentally broken process? Wllis certainly agreed with this enough to want to fix the process.

In addition to process improvement there were two other main reasons for change: Willis felt that the efficient handling of risk is impeded by lack of information and this is compounded by the fact that the industry is slow to leverage the potential offered by information standardisation. Consequently, they felt that they needed to improve access to information and to move towards standardisation of the information passed between themselves and their business partners.

The answer to ‘why do something now?’ for Willis was easy; ‘we must’! In 2006, Willis decided that ‘do nothing’ was not an option and initiated a major program to take them in the direction that they felt they needed to go.

Unsurprisingly in our market some of the major challenges were historic: Willis had already been ‘leaned’ twice, the last one in 2000. These two reorganisations attempted to ‘iron out’ differences between the acquired businesses rather than harvesting the synergies between them. Another interesting, if also unsurprising point, is that staff tend to stay long time at Willis, there are a large proportion that have been there over 20 years and as we all know and Willis discovered, it is hard to ‘teach old dogs new tricks’.

Willis felt that it was important to manage expectations from the start in terms of what was possible and what not. ‘Straight through processing’ for example, Willis feel that this is just a dream at the current time (although they feel that this will change in time, more of that later).  IT is only part of the answer, according to Willis, process being the other. It is vital, they say, to get the internal process correct and streamlined first and then seek external partners to collaborate and prototype with.

Willis decided that they needed a single, fully integrated system, and chose Eclipse from Sequel. Other key activities undertaken to realise the process change were: process definition and rationalisation; people were re-aligned around clearly defined roles within the new process with data captured only once – no re-keying! The location at which the process would be executed was decided after – there was no pre-assumption on this.

So, in addition to a new IT system what else did Willis do that led to success? There is quite a list: willingness to ‘bend’ the business to fit the solution; clear and detailed business requirements; challenge of unduly customized practices – look for commonality; obsessive, fact based prioritisation; utilisation of quality resource – the ‘new guard’ and lastly and vitally, unwavering management support.

Willis had a clear implementation plan of ‘buy not build’; business led solution design around an intense two month ‘conference room pilot’. They decided on a ‘trade on’ approach, with no data migration as these had compromised projects in the past. Willis invested in parallel running to ‘backstop’ testing. There was no ‘big bang’; delivery was phased, with aerospace kicking off followed by 10 further releases.

The driver for all of this work was to increase productivity and cost effectiveness. In some areas there was a 50% re-work rate. Willis found that downstream problems were symptomatic of upstream problems and solving these upstream problems led to the downstream ones going away. There were a number of key elements to the process re-engineering: disciplined processing with an emphasis on ‘right first time’ controlled by embedded real time compliance; clear defined roles and responsibilities, with removal of duplication and removal of the ‘hands off’ attitude and finally, rationalisation of processes. Two examples quoted were that some processes needed 9 authorisations that were reduced to 3 and also that every line of business had a different debit note format!

From an IT perspective, the legacy situation was a myriad of different systems. In order to get trading with some partners ‘ECA Lite’ was born which, in fact, was anything but ‘lite’. Rather it was a ‘heavy’ tactical solution involving: ACCESS databases; XIS produced ACORD messages and scanned documents. This was not going to be a scalable model.

The ‘new world’ was Eclipse, which is an integrated, single solution for all areas of the operation. The strategic model also sees an automated message gateway provided by Tri Systems and full DRI capabilities.

Now, in terms of trading externally, the strategic model gives the technical framework that Willis want, but trading externally doesn’t just happen naturally once all the technical pieces are in place. Willis’ view is that ACORD, whilst absolutely vital, is not enough by itself to make external trading happen. The reason is that standards define the structure but do not ensure quality, understandable and relevant content. ACORD does go quite a way in improving data quality but it is critical that the standards are reinforced by rules and protocols that govern the content. Willis’ feeling was that ‘straight through processing’ was only achievable with agreed, transparent and re-usable content. Also needed are SLAs with partners. The message they left us with is that external trading is not ‘plug and play’, there must be collaboration between partners.

I think that what Willis has achieved and they way they have gone about it can serve as a model that others could certainly do well to follow. Their observations and experiences on what it takes to effectively trade externally should be considered by all who chase the holy grail of ‘straight through processing’. I am not usually one for bureaucracy but I do feel that an initiative on how to control and standardise the content of ACORD messages is now vital, since there seems to be no debate left as to whether ACORD is the right direction in which to go and what we need to focus on now is how to effectively use the ACORD standards.