If you squint and think for a moment, it becomes obvious that technical support is deeply broken. One basic indicator is turnover in the industry. One estimate has customer support yearly turnover at 24% (probably less for technical support, but still). The role is not very prestigious, does not pay well, people do not stay for long, and few individuals will choose to invest the amount of effort into the field that would be comparable to other, more professionalised domains. This state of affairs has been linked to compensation, but that is only half of the story – why pay more when you can get by with less? The simple reality is that the role does not scale. It only “scales” in the old-fashioned way, by individuals climbing up organisational ranks.
On the consumer side, the situation is not much better. The human-powered nature of technical support means that capacity is not elastic: either the support organisation is overprovisioned most of the time, or customers have to deal with queues. The customer is presumably already unhappy about their purchase not functioning - and then they get put on hold! Add to that getting bounced from one agent to another, and having to explain the problem from scratch every time. Not all organisations have proper 24/7 support, which means that if the problem happened to occur at an inconvenient time, the customer now has another item parked on their mental TODO list. And last but not least, there’s a shift towards support based on text-based livechat rather than voice.
There is also another hidden cost to all these flaws. Products which could have been easily repaired, instead get thrown away, either because it was too expensive for the manufacturer to provide support for them, or it was too complicated for the user to get them fixed. This is pure waste and has ecological consequences, to the point where governments are offering tax breaks for repair services.
These are all indications of a fundamental problem – there is a dearth of good tooling for technical support. Tooling and automation is what makes other disciplines scale. Consider a ditch digger with a shovel, versus an excavator driver, versus an engineer overseeing a self-guided excavator robot. Sometimes having a human in the loop makes sense, but in technical support there are no good options for automation of repetitive tasks. The amount of work is fixed, and can only be pushed around. The most common approach is to push it onto consumers by hiding your support people behind phone menus, queues, checklists and rat-nests of “online knowledge bases”. The cost gets spread over many customers suffering mild inconveniences. That’s not exactly a winning move in 2018.
Surely we can do much better than that.