By Shanoor Seervai
It’s difficult to diagnose that something is failing; it’s even more difficult to be the person to call it.
One of the lessons Bryan shared with us is that communication is central to overcoming challenges, but is particularly constrained when it comes to tech.
I wonder how a blame-free culture could help ease channels of communication. Etsy has “blameless postmortems,” because the company wants to “view mistakes, errors, slips, lapses, etc. with a perspective of learning,” so that everyone can learn about what went wrong without fear of punishment or retribution. But in an environment where retribution is rife, it becomes more difficult to express one’s concerns: Brian told us about being alone in his sense that something was wrong, but not voicing these concerns. If retribution were less of a concern, perhaps Bryan would have been bolder in identifying what was wrong with healthcare.gov. As he recounted, saying something could be wrong led to his being shut out from high-level conversations.
This brings me to a second, bigger question (and lesson), which is that deciding who makes decisions, i.e. whose decisions count, is not trivial. For example, if it took President Obama months to decide that health exchanges would be called marketplaces, who was the person who decided (or didn’t decide) that that decision required presidential sign-off? If lines of communication are blurred, it means that there is uncertainty about who is taking decisions—and so decisions get punted up the pipeline until they arrive at the desk of the chief executive.
Maybe that decision about language and messaging was important enough for the President to weigh in. The issue at hand, however, is that the decision languished as long as it did—maybe the urgency wasn’t conveyed to the President; maybe it wasn’t made a priority. Regardless, greater transparency about who is making decisions and why could help avoid such pitfalls.
This points me to the third lesson, which is that unless management has a better idea about how things actually get done, operational crises are inevitable. As Clay Shirky points out, as long as talking to the “people who understood the technology became demeaning”, there will be no clear bridge between the big strategy and ideas conversations and the nuts and bolts of how those become reality. Bryan observed that the people managing things are oblivious to their mechanics—the “magical” changing of a name, as observed by managers, was actually the consequence of several hours of hard work on the part of technologists changing strings of code.
This does not mean that every manager must also be an expert coder (or vice versa). But intelligent, competent people need to step up to the plate not only when it comes to what they know, but also what they need to learn about. Managers who attempted to understand the intricacies of building a website would make more effective managerial decisions. Technologists who are allowed into the room for decision-making conversations may actually direct management to make smarter decisions. Whether or not a team is flat or hierarchical, greater communication between silo-ed rules could help prevent an IT failure.