When you are trying to offer good service, too much data or the wrong type can be more problematic than not having enough data in the first place.
Enterprise software specialist Freshworks and the Service Desk Institute researched this topic, and found that over half of the teams they spoke to suffer from this problem.
A massive 55 percent of businesses said they do not find all of their service desk metrics useful or valuable. Only 17 percent measure cost per incident. We spoke to Simon Johnson, General Manager, UK&I at Freshworks to find out more.
One of the main findings was that more than half of businesses don’t find their services desks are offering them useful data. Why are so many getting it wrong?
It’s important to understand why this data is not useful to the business. Is the fault here that people on the business side don’t understand the data they are getting, or is there a misunderstanding of what the business wants? Whatever the reason, this direct information around customer service and experience should be incredibly useful to other teams, from marketing and sales through to product development or operations. It’s therefore important to review the data that gets provided over on a continuous basis, so that it is always useful and valuable.
The fault is therefore less in the data itself, and more in deciding what data points are important for the ITSM team and for the business. Once these points have been discussed and confirmed, it’s then important to spend time on how that data is shared. For example, some teams don’t have an accurate, “apples to apples” comparison across all their channels. If you can’t get a consistent approach across all your channels, this makes the data hard to understand.
For external-facing helpdesk teams, it can be hard to get email, phone and online chat conversations ranked alongside each other. Each of them involves an interaction with a customer, but are the interactions being tracked equally? For internal service desk teams, the problem can exist too.
Another issue is when teams only get asked for numbers – for example, how many tickets get dealt with. You can reply with a number – “2000 tickets” – to that. Originally, that might have been enough. But people change roles, or get replaced, and that number loses its context. If you can’t help someone to understand that this figure is a great result given that half the team were out sick or on holiday, or this is the busiest time of the year, or it’s actually lower than normal due to a new approach that helped offer more self-service … that is a big miss. Context is vital to decision making.
Is this a case of service desks not moving with the times? Can they make changes, when their job involves change management?
The challenge here is that companies want to adopt new approaches to working with customers, and teams have to develop new ways to meet what the business wants around IT. However, this speed of change can make it difficult for existing teams to evolve their processes. This is particularly difficult if your ITSM tools don’t support easy customisation and configuration changes that you can carry out yourself.
ITSM teams should be able to use their existing change management skills to improve. Better planning can help avoid some of the simple challenges. Even just being a fresh pair of eyes on a plan can make things easier.
Are new developments like DevOps helping or hindering service teams?
DevOps was developed to bridge the gap between developers and IT operations – the processes around automation and continuous deployment have improved because of the collaboration here.
The issue is where the boundaries are around collaboration between teams. Getting people involved across the business – so from service and support as well as operations – should be a natural step at the start, but it is all too easy to miss. Opening up discussions with other teams – for example, marketing around customer experience – can help you get involved earlier.
DevOps aligns well with the Agile development philosophy – so ensuring shorter release cycles, and making changes to priorities so that services meet business needs now, rather than what they were expected to be months ago. This emphasis on change should mean that DevOps will help these service teams as they provide input on their requirements as well, and those changes should be implemented quickly too.
I am reminded of Conway’s Law, which was first developed in 1967 – that companies build IT systems and services that match how their teams communicate internally. If you run in silos, then your IT systems will tend to be siloed and stand-alone; if you focus on cross-department initiatives, then your IT will be the same. DevOps can help that process, but it can involve taking baby steps at the beginning.
How can teams learn to use their data more effectively?
Context is really important. According to our research with SDI, around 80 per cent of service desk teams provide monthly reports to the business. However, only 38 per cent of companies provide context on the results that they offer. If you want more teams to make use of your data, then this kind of insight should be added.
It’s also worth looking for stories in your data. When you have common issues come up that affect multiple people, can you find a good example that puts a human face on that problem? This can help paint a picture for others within the business. Combining data with stories can help people understand some of the impact from the decisions they make.
Will AI automate all this anyway?
AI can help automate some of the processes that service and help desk teams go through. It can make it easier to link up the right people to fix a problem faster based around skills.
Similarly, search-driven analytics can help businesses prepare for the future based on the results that they are seeing today.
These results can be used to automatically spot where common issues are coming up and what resources can be provided to help those problems – whether this is new content that can be provided to help users solve their own issues, or additional staffing resources for specific channels that are seeing increased demand from customers. By forecasting resource requirements, you can improve efficiency across the team using AI and automation together.
However, AI is not the answer to everything and skilled staff will be necessary to keep this running. For example, AI can be used to power chatbots, but this relies on your approach to knowledge management. If you don’t have good content that is well ordered and searchable using AI, then your chatbot won’t be effective. This is where your existing staff have a significant role to play on understanding your customers and what their issues are, and how to develop content to meet their needs. Without this, your investment in AI will be wasted.