Solo SaaS support: SLAs that don't kill you

Solo micro-SaaS economics only work if support doesn't consume the founder. The instinct is to promise fast response times to compete with bigger competitors; the math is that 24/7 Slack-grade support is incompatible with shipping the product. Here's the realistic SLA shape and the deflection tactics that make it sustainable.

The SLAs that customers accept

For micro-SaaS in the $10-50/month band:

• Median first response: 24 hours business days. Not 2 hours. Customers at this price point don't expect instant.
• Resolution time for non-urgent issues: 3-5 business days. Acceptable when the customer knows what to expect.
• Urgent / "production down" SLA: 4 hours business days. Stripe-down or auth-broken counts; "I can't find a setting" doesn't.
• Weekends: 48-72 hours response. Set the expectation in the help docs explicitly.

Publish these. Customers calibrate to what you tell them. Founders who don't publish SLAs get held to the customer's imagined SLAs, which are always faster than reality.

The deflection tactics that actually work

1. A self-serve help centre with 20-30 articles. 60-80% of support questions are repeat questions. Documenting answers once deflects them forever. The articles need real screenshots and concrete examples, not generic prose.

2. An in-app help search. A search bar in the product that surfaces help articles. Cuts ticket volume 30-40% on its own.

3. A "did this help?" thumbs in every help article. The articles with low thumbs-up rates are the ones to rewrite. Customers tell you what's broken in docs faster than you'd find it yourself.

4. A status page. Atlassian Statuspage free tier or BetterStack. When something's broken, customers check the status page before emailing. Drops "is it down?" tickets to near zero during incidents.

5. A clear template for bug reports. Email auto-responder with a structured template ("what were you doing? what happened? what did you expect? include screenshot if possible"). Customers who get the auto-responder usually provide enough info to triage without a back-and-forth.

What kills solo support

• Trying to respond instantly. The team of one cannot match the responsiveness of a team of twenty. Set the SLA realistically and meet it consistently; that beats inconsistent fast responses.
• Saying yes to every feature request via support. Have a separate channel ("feature requests welcome at [email]"); don't promise builds during a support ticket.
• Personal-touch support for free-tier users. Free users at scale will consume more support than they generate revenue. Help-centre-only for free tier; live support for paid only.
• Refunding without policy. Codify a clear refund policy and stick to it. Bespoke refund decisions on every ticket consume founder energy.

What to do today

1. Publish your SLAs in a public help-centre article. Be specific.
2. Audit your last 30 support tickets. Which 5 questions repeat? Write a help article for each one.
3. Set up the auto-responder template.
4. Add a status page (~30 minutes setup).
5. Defer everything in the ticket queue that's not actually urgent, with the SLA email.