Laurent Lemaire, co-founder of SimpleBackups, described it better than most: “Support is taking a lot of our time, and the main problem is that it is randomly breaking our days.”
Not the volume. The randomness.
The problem with unpredictable support
Deep work requires uninterrupted time. Building a feature, writing a proposal, reviewing a design — these all require you to hold a lot of context in your head at once. An interruption does not just pause that work. It destroys the context. Getting back to the same level of focus takes another twenty minutes.
Support emails do not arrive on schedule. A customer in a different timezone sends a question at 3am that is sitting in your inbox when you open your laptop. A billing issue lands mid-afternoon when you were deep in a codebase. A frustrated user replies to a thread you had mentally closed two weeks ago.
“The unpredictability of support requests prevents us from being efficient and productive and often we’ll end up having to stop what we’re doing to reply,” as Lemaire put it.
What the context switching overhead actually costs
One Hacker News commenter described it directly: “The hardest is the context switching overhead. One minute you’re coding or marketing, and then you have to switch into doing two or three hours of customer support.”
Research from ClearFeed found that agents managing support across multiple channels lose 6.8 hours per week just to context switching. For a founder handling support solo, the number is worse — because there is no separation between support mode and building mode. Every email that demands attention is a tax on everything else.
One founder on Indie Hackers described the end state: “80% of my time went towards support.” They eventually abandoned the product.
The trap of treating support as urgent by default
Not every customer email is urgent. A feature request does not need a response in the next hour. A general question can wait until your designated support window. But when everything arrives in the same inbox with no classification, everything feels equally urgent — and your brain treats it that way.
The result is that low-priority emails interrupt high-priority work. A spam-adjacent inquiry that could be dismissed in thirty seconds derails a two-hour coding session.
What structured support looks like
The teams that solve this do not necessarily spend less time on support. They spend that time intentionally.
AI triage means every email is classified before a human sees it — category, urgency, sentiment. A BLOCKER from an enterprise client gets flagged immediately. A general feature request gets routed to the product channel and picked up when the team has bandwidth. Nothing gets lost, but not everything demands immediate attention.
When support emails arrive in Slack or Discord with context already attached, your team can make fast decisions: who handles it, how urgent it is, and when. That is not the same as being interrupted by an unclassified email in a shared inbox at the worst possible moment.
The randomness does not have to go away entirely. But it can stop breaking your day every time.