To the Left

The first time I heard about Shift Left is when I read an IBM whitepaper on the optimization of tech support. It was amazing, obvious and, for most IT organizations, way ahead of it's time. It took two years before I had a customer asking for this concept, and it's an instant hit with every IT manager who hears of it.

The basic premise that support tickets become increasingly expensive if multiple and/or specialized parties are involved in resolving them. It is therefore advisable to optimize IT infrastructure and support in such a way that the bulk of incident tickets can be resolved quickly and cheaply. The following picture sums it up pretty well:

ShiftLeft - Picture by OGD ict-diensten, used with permission.

If you take, from left to right, IT infrastructure, end users, IT support, system administrators, and technical experts, you want the largest volume of tickets to be resolved mostly on the left, rather than on the right side. After all, when something fails the end user cannot work. If it's not automatically resolved of fixed by the user himself, he just has to wait around while the service desk starts burning time and money. When the service desk cannot resolve the issue, the more expensive system administrators get in gear and start billing, all the while leaving the end user unable to do part of his work. In short, the further 'right' the problem goes before it's resolved, the higher cost in time and money. Hence, the drive to 'Shift Left'.

There are a number of things a company can do to enable this shift left. First is very good knowledge management, where the service desk learns to perform as much of the system administrators job as possible without significant error rates. Secondly, the end user can gain knowledge of how to resolve common problems on their own by making a knowledge base and relying on community support, training for commonly used devices and applications, and digital literacy.
Both the service desk and the end user need to be assigned sufficient rights to resolve common problems on their own for these two measures to work. Thirdly, modern IT infrastructure can be configured to be highly fault-tolerant, and if you switch to cloud services the issue becomes moot and you only have to worry about internet access.

The purpose is to optimize for quantity. Back-end systems can fail over to each other. Users can find their own 'any key'. The service desk is well able to execute common changes and resolve issues on it's own, given proper guidance and training. Sysadmins would much rather hack away at a difficult issue once in a while instead of being inundated in relatively common and easy tickets. Ideally, the cost, volume and resolution time of tickets is greatly optimized.

Thankfully the technology like cloud services are evolving to give a seamless experience even if single components fail. Digital literacy is increasing quickly and end users are quite happy with devices and apps that are intuitive to use and manage without reliance on IT support. As more people use tablets, phones and purpose-built apps to do parts of their work, the IT department takes on the role of a facilitator rather than a break-fix oriented club of technicians.

* He/his has to serve for all genders here :)