Build vs Buy: Error Tracking

Every team has, at some point, added a try/catch that posts to a Slack webhook and declared error tracking solved. It works, it took twenty minutes, and it is genuinely the right call for a side project. The trouble starts about six months later.

What the twenty minutes leaves out

• Grouping, or the channel becomes unusable on the first bad deploy.
• Rate limiting, so one loop does not post ten thousand messages.
• Retention and storage, because someone will eventually ask what happened in March.
• A dashboard, because scrolling a Slack channel is not triage.
• Routing rules, because not everything deserves the same destination.
• Retries and escalation, and the on-call process around them.

When building is still right

If error handling is your product, build it. If you have a data-residency constraint no vendor meets, build it. Otherwise the honest comparison is not "a webhook versus a subscription" — it is the third year of maintaining the list above versus not.

See what it costs →