More and more people are engaging with retail and fashion in a multi-channel manner, resulting in greater business scrutiny of the availability and performance of core IT services. Consequently, retailers are now attempting to identify ways that their IT systems could work more efficiently and effectively with their internal and external delivery teams, in order to support changing business needs.
It's important for retailers to work with their suppliers during contract discussions to drive improvements in availability and service robustness, whilst at the same time, creating contracts that flex with retail business requirements. When preparing any service level regime, it's generally helpful to bear the old adage "quality over quantity" in mind.
The service level metrics need to be easy to work with a small number of "key" service levels. It's important to identify issues which will create a "pinch" for retail businesses, but similarly, ensure that any service level regime is inherently flexible, so as to allow for the promotion and/or demotion of services levels, service credit points or percentage pool allocations to reflect changing business need. From a supplier perspective, however, this may need to be subject to a cap on the aggregate number of service levels, or the extent of changes that can be "imposed" rather than mutually agreed.
It is important to identify what the service requirement actually is. For example, maintenance windows for a core application in the early hours of a Monday morning may be acceptable to a retail business during most of the year, but the sales cost of downtime for the same period in December may well lead to adverse scrutiny of the IT department and its suppliers alike.
In the same vein, it may be that most of the year, on a 24/7 measurement basis, 99.9 per cent availability (which equates to roughly 10 minutes of downtime a week) for a given IT service/application is sufficient. In peak periods, however, when customers may go elsewhere if they can't make their purchase, the lost sales cost of the same 10 minutes may be significantly more. Whilst a business may not require "gold plated" availability (with a corresponding price tag) for most of the year, one should accordingly consider whether, like other parts of retail businesses, IT service delivery arrangements can be structured around peak trading periods.
At its simplest, this can mean having a higher service level at different times of the year, or changing the requirement altogether. This might mean fewer, or indeed no priority 1 incidents that remain unresolved within a given fixed time, or that the fix time for the same incident becomes shorter. Equally, one can look at minimising "down-time" by placing a freeze on all non-essential maintenance unless signed off by both parties' IT change board.
Another means of achieving this may involve looking at the consequences of failing to meet the service level, rather than the service level itself. Often, we see customers seeking to apply a multiplier to any service credit points, percentage allocations or flat rates, with the effect of increasing the service credit accruing as a result of any service level failure, if that the failure occurs during a defined "peak" period.
It doesn't have to be all stick as opposed to carrot though. In some cases, agreeing a positive incentivisation scheme which applies during peak periods, can work, provided that it is clearly structured and easy to administer. Taking the example of the 24/7 availability service level described above, it may be that 10 minutes of weekly downtime is tolerable for most of the year, but during seasonal sales periods there is a tangible (and possibly even quantifiable) business benefit of a service being delivered to 99.99 per cent availability (so only just over 1 minute of downtime over the week), with a resulting increase in online trade.
Rather than pegging the service level at the higher level and paying for a quadruple nine service all year round, one may accordingly want to consider whether the service level could be pegged at the lower level, with the ability for a supplier to earn a bonus payment, if it can deliver to the higher level, during the customer's busiest period. Alternatively, if the platform or application is simply not able to function at these levels, retail businesses could consider rewarding suppliers via gainshare, if a supplier is able to identify and successfully implement initiatives that improve the level of operation.
Creating a model which flexes and adapts, can be a useful means of ensuring that IT systems meet commercial expectations and trade requirements. Service incentivisation at its best, encourages both retail businesses and their suppliers to work together and the means of achieving this described above, are just examples of the levers and dials to consider.
Chloe Hersee is senior associate at law firm DLA Piper
Sometimes, the power of the mainframe is the most cost effective answer. Computing's Peter Gothard puts Computing's readers' questions on the future of the mainframe to IBM's Z13 expert Steven Dickens.
This Dummies white paper will help you better understand business process management (BPM)