Dysfunctional

Here I will appear to be griping a bit so bear with me. A few problems as I see with the current way of working in my group (around 200 folks) are drafted below. Here I have noted only the problems which I have encountered in more than one environments/ organizations/ over time. So these are certainly not related only to my current work. But when I was somewhat junior, I could ignore, sweep these problems under the carpet. In fact, I can do that even now. But, I think, if I could make a difference in one or more of these systemic problem areas, that would really be a high value impact. The idea here is to look out for information. Any pointers to case studies/ books would help. Would a management diploma/ course help? TIA. I have just put these problems in my current project’s context.

In the project, we don’t have much maneuvering space because we are dealing with business critical systems. As some systems are live in production, keeping those working is a priority (as it should be), and whatever small movements we can allow there don’t really have much chance of becoming a culture.

A few people who are long associated with the systems have a sort of institutional knowledge and are not replaceable. And they end up occupying more space without adding much value. And they don’t help us take the direction where we want to go. But because we cannot replace these, they dictate and constrain the paths a lot. A people problem. I recently read Bob Iger’s The Ride of a lifetime (listened to audiobook) and the situation seems- remotely- relatable to his pulling the plug on the series Twin Peaks- but not sure.

A side effect of KRAs/ OKRs is that these get optimized for. This is a general observation in performance evaluation and management. Any measure against which you evaluate someone’s performance will be gamed/ optimized for. This is like the cobra effect/ perverse incentive. So for example, if you place big importance on story points, people would game the idea of story points (they would deliberately estimate that an item would take 8 story points instead of realistic 5). Same goes for what did you do today, you should be more visible to the top management kind of metrics. I think this placing of importance on something doesn’t even have to be explicit for that measure to get optimized for. I don’t think there is an easy solution to this. (I should finish Andy Grove’s High Output Management soon.) The current management- from where I see it- is capable from domain, process, and technical sides but they are pressed for time and limited by some organization level constraints. Aren’t we always, though? This could also be due to a large group size. For a large group, the normal distribution would mandate focusing on getting predictable output (six sigma) from x percent of people and having them fall in line on some measures/ follow some processes. Continuing with the story point example, I don’t think for a large group, you can work without agile. But to get a large group follow some common processes like agile/ measures like story points and velocity will give rise to certain practices which are not so productive or valuable- for example, you may end up with too many standards, meetings, committees, and complacency behaviour patters- we already have 30 story points work in this sprint. Even minimal standards/ processes can result in sub-optmail results and outcomes for which these were created. For example, we have 70% code coverage with no critical SonarQube issues may become a sort of excuse where OWASP vulnerabilities get ignored. Again, as I said, I don’t think there is an esay solution to this. As a workaround I have seen more processes/ standards/ best practices getting added with the side effect that none of these are followed whole-heartedly. But a start-up kind of culture- devoid of standards/ processes is not a solution. It has its own problems.

Another aspect is networks of people. As you grow, your networks grow (as they should). And people feel more comfortable working with people they know, have known. As a result, new ideas have to be visibly good immediately, and new people have to be really smart to get something done and make their presence felt. It is so everywhere in the industry. A good management should strike a balance between old and new people and ideas. How do we go about it? Any books I can read on this? From my perspective and somewhat limited understanding, a particular area of focus for someone who is new to the team should be to emit information better. The management, when they want some information, find it easy to pull it from the structures they have been comfortable with for some time (teams, people, dashboards). This could be an area of disruption for new folks. It is well known that emitting information is better than pulling it because ‘it is right there in front of you if and when you need it’. For example, a grafana dashboard, or an array of display monitors on the work floor which continuously display build statuses, production traffics, subsystem statuses, hotspots, etc.

Another observation is that some people are more comfortable in projecting appearance of activity than in actual activity. I am not certain yet if this is a people (individual) problem or wider culture problem. But, as I see it, some problems are not getting solved. There is some activity about solving these problems, some noise and then the problems get deferred and some organization change happens, people move on and there are new problems to solve or to appear to solve.

Again, even though I have used present tense here, I have seen these problems more than once in many orgs- not just current. And I understand that in the end any problem is a people problem (Weinberg’s three laws of consulting).

How do we address these?