Lesson 5: Technology goes wrong, it's how humans respond that matters
It doesn’t matter who you are or what the technology is, sometimes it goes wrong.
Technology has advanced to the point that we have created amazing services and products but they have become complex systems, reliant on many individual elements working together. Even the mighty AWS has experienced some very public failures, one most recently following the company’s re:Invent conference with an outage in its US-East 2 region.
After 10 years in the hosting industry, I can confidently say that I have never worked anywhere that hasn’t had an unexpected outage.
Faced with this inevitability, there are two key things that infrastructure hosting providers need to ensure. Firstly, that they do everything they can to make sure any outages or failures happen as far apart as possible. Secondly, when outages do occur, they need to handle them in the correct way. And that isn’t by giving customers access to a chatbot that will offer them a series of predefined options to choose from, only to be directed to a knowledge base article on the website.
Or, and I’m sure we’ve all been in this position when calling our water, gas, electricity, internet supplier, being passed from pillar to post, explaining each time who you are and why you’re calling, with no one willing to take responsibility for the issue that you’re experiencing.
It’s perhaps one of the most excruciatingly frustrating feelings in the world.
Now imagine that same experience when you’ve got a server outage and your company website has gone down or you’re unable to process payments. It’s no longer a case of inconvenience but potentially business-ending.
And yet, that kind of detached customer support has somehow become the norm. With the recent launch of ChatGPT - OpenAI’s new artificial intelligence chatbot - and the hype it’s generating about the advance of AI-based tools for things like customer service, it’s certainly not a trend that is slowing.
That’s not to say that chatbots don’t have their place. They absolutely do, but they cannot replace humans when it comes to supporting something as important as infrastructure.
What people want when something goes wrong with their infrastructure is someone on the end of the phone, or WhatsApp, or Telegram, or any other platform they might choose, who can talk though the problem and reassure them that something is being done to rectify it.
Ultimately, people want communication, and to feel heard, to know that their issue is being resolved and that they are being kept informed on progress.
That level of support - to get straight through to someone who can solve your problem - will cost more money. For the hosting provider, it’s more resource intensive both from a headcount perspective and because they train staff to empower them to support customers with their issues directly - even if that’s just knowing who can help them with their issue, opening up that channel and facilitating communication between parties.
I truly believe that as technology has become more complex, rather than making humans redundant, it has placed an even greater spotlight on the importance of our role when something inevitably goes wrong.
I’d love to open up the discussion in the comments. Would you pay more for better service? What experiences have you had with support provided by chatbot vs a human in the world of infrastructure hosting?