We are in a productivity trap. There seems to be nothing but savings in the toolbox. But if we really want to increase productivity, we need to smartly invest more in tools and people.
Save money and work harder
Globally, the telco sector is a sad example of how companies are trying to sustain their business by saving opex and capex in turn. Now in economic headwinds, enterprises have also stopped spending on new initiatives. Investments are mostly to keep lights on and minimally revive existing services. Big enterprises are in a corporate mode which means quartal financial optimization and short-sighted vision. The main focus is on savings. Even vendors and startups which have driven innovation now think that incremental evolutionary solutions are the best choice. Vendors consolidate and buy other similar companies to shape their business to new market conditions. Startups are sucked inside incumbents by acquisitions and they adopt these corporate operating models. Revolutionary tech waves have been missing for a long time. AI has now all the chances to become a game changer but so far it hasn’t been realized yet.
Here in Finland companies are very keen to pay dividends to shareholders instead of investing in the company’s future growth. Companies have a hard time convincing investors to invest in the long-term and see the growth potential because investors are afraid that companies will make stupid investments or acquisitions. I wonder if there are problems in the management then. If companies invest, money will go to buildings, not things that would increase productivity. Healthcare is in crisis mode because of increasing load and ever-rising costs. Currently, industry and employer unions and the government are driving blindless initiatives to tighten working conditions for employees, not to mention ridiculously low wage levels and their suggested raises. The mindset is “If we just work more and harder, our productivity and global competitiveness will increase”. That’s just stupid and shows that productivity in white-collar professional work is not understood at all.
Mental problems
The goal should be “work smarter, not harder” as often cited. Productivity will not increase by pressing harder. In the engineering world developer productivity studies say that 3-5 hours a day is the optimal time to work at the best performance. It’s half a day at maximum, think about that. The reality is that many people have calendars fully booked with meetings where they arrive head empty, without preparation. Half of the meeting time is spent on recalling what was agreed last time and what should have been done already. People leave for the next similar meeting frustrated and without options to do something tangible that will advance things. All you can have are short slots where you can send a small and hasty message to someone, that hardly will advance things. Probably it just creates more hassle and work in other parts of the organization. This clueless power-eater theater will drain people and diminish their input.
Context switching in this hard-tempo meeting mix is rough for individuals who want to be more than numb actors in a play. Sadly, many of us are just used to running this game daily. Collecting thoughts and getting to the flow state is important to actually do something, drive things, and get meaningful outcomes. Context switching in the brain just takes time before the thoughts start to flow again. Messing the flow with constant interruptions and info load will back off the person’s capabilities like the original Ethernet’s performance. What we need is unspoiled time without pointless distractions. This might feel like wasting resources, but many people forget or deny that idle-looking thinking is probably one of the most important tasks. At least someone at the company must do the thinking, form reasonable thoughts, and be the manager who makes other workers’ jobs clear and productive enough. Without that, the company won’t succeed.
Processes and AI
Engineers need information and feedback to do their jobs. Therefore, it’s important to make all information easily available. That could be agile boards, chat, wikis, ITSM, CRM, or any type of documentation. The problem is that usually needle in the haystack is hard to find. AI tools can help here making information access quick and easy. But we must also ingest the proper information to the systems to be available to others. AI tools can help here too by collecting and organizing data thoroughly and systematically. The obstacle will be the resisting mindset and hard-locked operational models, both at the company and individual levels. IT systems are so complex and distributed that humans alone can’t manage them properly anymore. Automation is needed, and more specifically acceptance of automated operations and trust for them. Automation evolution calls for tools to mature and offer more usable features, so that they deserve the user’s trust and users can build productive operations around them. Also, internal development is needed to match tools and the company’s operations together into a functional compound.
That brings us to the processes which are essentially people and their actions. AI-enabled tools can provide a far better view of the whole IT landscape and what is the situation at different levels of infrastructure and services. Tools can spot anomalies and raise the flag with the context to augment the operator’s understanding of what’s happening, where, and why. Issues can be quickly triaged and moved to the right person to be solved. This means a much shorter meantime to innocence and less blame game between different parties. Vague support tickets like “network is not working” can lead to quick identification of what is not actually working, and then resolution is far closer. Best AI tools can recommend solutions and even fix them automatically if you just dare to use them.
Tooling for the win
Ultimately, it’s about making decisions based on real information, analytics of the IT systems and services which now have a direct link to customer satisfaction and company success. If we can see a wide and deep snapshot of our infra and services, we can make educated decisions that have the intended focus, scope, timing, and impact. That’s a huge win, a time and money saver, and a quality improvement. Digital twin is the tool for simulating changes and predicting the future. We can now spot emerging problems in advance and fix them before they affect services or users. Change management got a whole lot easier if we could see the correct initial data and the state of the infra before the changes. Then we can simulate the impacts of planned change or model different scenarios. This might be an even more important feature than automating operations.
Digital experience has been an emerging topic in recent years thanks to more advanced software solutions. Monitoring services with DEM gives deep visibility to infra stack and service chains end to end. E.g. SASE clients usually have built-in DEM to measure and report clients view of infrastructure and services. We need to understand that the network is just an enabler for services and data transport. We operate at underlay and basement in the IT stack anyway, which means there are a lot of layers on top of network layers. Visibility to them helps to understand how services work dependently or independently of the network components. Multi-level measurements also reveal where the problems might be from the end user perspective. Users are now championed as they should because that’s what services are for. DEM also brings together different IT groups and even different providers. It would be a great help if we could indicate what part of the service chain is failing and who is responsible for that.
Look forward
If we really want to improve productivity, we must invest in new strategic tools and capabilities. This is so far a bit of an uncharted journey, a leaf of faith for many until pioneers show the way, commoditization happens, and we all get confidence to implement own operational models. Tools are here, and we should start utilizing them and getting experience. When you get an impression of what tools can offer, you can start developing operational models around them. Tools will evolve as time goes by, so are you and your organization.
All that it needs is money and effort, forward-thinking and brave investment for something you can’t truly know. If you dare to take that step, it will pay off. First, you need to find, evaluate, and buy smarter products or select the provider who uses these tools. Then engineers need time to test, develop, and deploy these systems. Same time operational models and mindsets must change. This is not a sprint, but more like a constant evolution path without a clear ultimate goal. Managers must understand that. Results will be beneficial for both end users and IT staff as well, which will be a great overall success for IT managers, and eventually for the whole company. AIOps is like Zero Trust in security, a new way to operate the same old infrastructure blocks, but with smarter and more effective ways. This is how it always should have been.