Keeping employees and the company productive is the primary goal of IT support in the sense of digital employee experience management. However, further thought should be given at this point. […]
Running or not running? Those who derive detailed insights from daily IT system administration and helpdesk operation and use key figures strategically have decisive advantages: Technically, the key figures form reliable foundations for targeted investments-for sustainably optimized IT infrastructures and efficient operating processes. From an organizational point of view, the key figures from IT support help to measure the success or ROI of IT investments, rollouts and upgrades based on employee productivity and satisfaction.
However, achieving these goals requires a change of perspective – away from reactive problem solving and towards forward-looking IT support – in conjunction with innovative technologies such as analytics, artificial intelligence or bots. It is about – so the experience from project practice – to develop five skills for IT support solutions.
The volume of tickets stagnates or decreases? Security can be deceptive. Studies show that as a rule only 50 percent of disruptions are reported at IT workstations-for a variety of reasons. If a message comes up, the service desk often lacks the necessary transparency: It knows little about the user, his workplace and what is happening on the terminal. Decreasing ticket numbers and short processing times are only meaningful KPIs if they are validated with satisfaction surveys and proactive action and the influence of certain measures can be actively demonstrated. As a rule, the entire workplace landscape is a heterogeneous blind spot. Individual faults on the devices are just as difficult to diagnose as connection problems.
Therefore, it is first necessary to “measure” the IT user experience-with hard and soft metrics. Since large amounts of data quickly come together in larger installations, only AI or machine learning approaches that automatically identify fault images based on KPIs for the Digital Employee Experience (DEX) and pattern recognition and make the control center capable of action help. Keyword control center: This must make the user’s view transparent and enable proactive intervention if the KPIs deteriorate. The fact that everything is running in the data center doesn’t help much if users can’t access the services, are unproductive or frustrated because they can’t do their job.
A classic example is performance complaints from Microsoft 365 after the switch from on Premises to SaaS: As a result, many tickets run up, but the cause remains in the dark – a network problem is suspected, but not found. In such a case, a predictive procedure that automatically analyzes all connection attempts from thousands of terminals will help. A ‘clustering’ of the devices with good versus bad connections (e.g. a latency of more than 600ms) would identify a possibly misconfigured proxy server as the cause. Instead of processing thousands of complaints at a cost of around 15 euros per ticket, disruptions can be resolved much faster and costs can be avoided from the outset.
Real measured values are often only half the truth: A service shows optimal performance data in the data center, measurement data from end devices do not provide any abnormalities-and yet employees can be dissatisfied. To get a picture of the mood here, non-specific mass surveys via e-mail with links to long questionnaires are rather counterproductive. Companies that conduct annual satisfaction surveys know the problem: low response rates due to too many questions and reservations that nothing will change. Actionable results, on the other hand, provide specific, context-related surveys that appear briefly and precisely on the user’s screen. Specifically addressed to a current topic, the response rates are over 80 percent.
Let’s take video conferencing platforms as a more recent example: If test users do not receive any significant tickets in the service desk, a rollout seems certain. It remains to be seen whether and how intensively testing has actually been carried out. Because only when technical data is correlated with concrete user feedback, a realistic picture of all advantages and problem areas is shown and you can see how new projects are generally accepted.
Contextualized communication is also relevant from the point of view of the control center: Why not inform users who are affected by malfunctions openly? Instead of accepting’ unavoidable ‘ tickets when the crashes become more frequent with a newly rolled out application, proactive targeted information of the affected users is helpful. It shows that the IT is working on it, it is not the end device or incorrect operation and a solution is foreseeable in time. The result is less frustration on both sides.
In practice, operationalizing a digital experience control center means operating a big data application from which options for action and improvement initiatives are derived (see point 1). With regard to a workplace environment, the question arises daily: What distinguishes good from bad IT experience? Classically, cluster techniques serve to find similarities or differences. Today, there are AI and machine learning procedures to identify clusters:
- Step 1: Preclassification / compression of many individual measured values into tradable, meaningful aggregates (scores);
- Step 2: Analyze which features / factors lead to good vs. bad scores;
However, just having a lot of data and thus transparency is not enough. Due to the amount of data generated, mechanisms such as AI must be used to inform a systematic control center about fault patterns at all, to name factors and to enable root cause research or correction.
Around digital workplaces, there are usually many more construction sites than the IT team can handle-it must be prioritized. But what are the criteria? It is obvious to first address the disorders with the most affected. From a big data perspective, this means that the clusters with the largest number of disrupted workplaces due to a common possible cause are at the top of the list. This is important so that the ‘low hanging fruits’ are not harvested first, from which only a few in the company benefit. Sometimes a readjustment is necessary, since the clustering mainly takes into account technical parameters. But there are critical roles and areas that have priority-such as dispatchers in logistics, finance or production sites.
This is where the advantages of a control room with AI-based clustering and drill-down options based on business intelligence come in: This enables a prioritization that quickly fixes the most urgent problems of the most important corporate functions with the available resources. Once the relevant fault pattern has been identified, stored interference suppression methods, so-called playbooks, support understanding the details and implementing the necessary steps. Ideally, these run automatically. An optimized IT control center is therefore not only pointed out to malfunctions, but also gets the procedure at hand to bring about a corresponding solution.
Automation in the data center has been an alternative scenario for years, which outsourcing service providers in particular use to efficiently avoid or solve disruptions. In the workplace environment, automation is primarily associated with software distribution and is often in its infancy when it comes to eliminating disruption to workplaces. Many service desks have scripts with which they handle known problems. This either runs through a remote session or the user is asked to do certain things. Much of this could be automated. With known common problems such as repairing Outlook profiles, the effort pays off immediately.
The same applies to major disturbances. A typical example is if a network problem could be solved with thousands of computers by a (temporary) proxy setting change: impossible to implement by hand. Inform users by e-mail and explain the steps? Not particularly efficient. Only automation helps here, preferably with the help of a stored interference suppression method.
With infrastructures constantly evolving, changing and demanding users, it is not always easy for IT to ensure productivity and ease of use for employees. While best practices and automatisms have been in place in the data center for many years, there is a lack of transparency and efficient processes in the digital workplaces.
The concept of an IT control center with a focus on the employee promises a remedy: To proactively understand where it is “stuck”, to be able to prioritize error situations using AI, combined with automated work instructions for troubleshooting – this will make the difference in the future: Save IT costs and at the same time make employees more satisfied and productive and thus have a lasting positive influence on the business.
* Holger Dörnemann is Solution Consultant Director Central EMEA at Nexthink.