Mistakes to Avoid

BAD: Proposing SLOs without pricing infrastructure cost. At a 2023 debrief for an Airtable competitor, a candidate proposed 99.99% for a document collaboration feature. When the engineering lead asked “what does that cost,” the candidate said “I can calculate that.” The role went to someone who had the calculation ready: $23,000/month additional, 2.3% of burn rate. Specificity signals readiness.

GOOD: Arriving with three costed options and a recommendation that names the business trade-off. “I recommend 99.9% at $8K. The incremental $14K for 99.95% buys us customer segments that generate $2K monthly. Break-even at 7 months if we don’t grow. I’d revisit after Series B.”

BAD: Treating SLO violations as purely technical incidents. In a 2024 loop, a candidate described an incident response without mentioning revenue impact or customer communication. The hiring manager, former SRE at Netflix, wrote: “Doesn’t understand the job.”

GOOD: Framing every incident in business terms. “The 14-minute outage affected checkout for 12% of users. We estimated $47,000 in abandoned carts. I proposed accepting the error budget depletion rather than emergency patching, because the alternative was a 6-hour maintenance window during Black Friday prep.” This is the language of someone who sets SLOs with stakeholders, not in isolation.

BAD: Negotiating equity without understanding liquidation preferences. A candidate for a 2022 startup role demanded 0.5% equity without asking about preference stack. The offer was 1x non-participating. He could have asked for 0.3% and 1x participating, or acceleration on change of control. The gap marked him as inexperienced with startup compensation.

GOOD: “I’d like to understand the preference stack and model my effective ownership at 2x and 4x liquidation multiples. My target is $X effective value at Series C, not nominal percentage. Can we discuss acceleration on acquisition?”

---

FAQ

Should I mention my FAANG SLO experience in startup interviews?

Mention it once, then pivot. In a 2023 debrief for a Notion competitor, the candidate opened with “At Google, we set SLOs for…” The interviewer, former founding engineer, stopped taking notes. The hired candidate said: “I managed SLOs for systems with 100M users. Here’s how I’d adapt that to your stage,” then proposed a 90-day measurement plan. The signal is adaptability, not pedigree.

How do I answer “What’s your target SLO?” when they haven’t defined the product?

You don’t. In a 2024 seed-stage loop, a candidate answered “99.95%” to an undefined product. The founder later said: “He’d break us.” The passing candidate said: “I’d need to understand what customer success looks like, what failure costs per incident, and your current architecture constraints. Can you walk me through your last customer-impacting incident?” The question signaled partnership. The 99.95% signaled rigidity.

What if the startup refuses to commit to any SLO?

Negotiate for commitment to a decision timeline, not the SLO itself. In a 2023 debrief, a candidate accepted “we’ll revisit after launch” without a date. The SLO was never revisited; the product failed. The candidate who passed on a different role insisted on: “SLO proposal within 30 days of launch, reviewed in weekly engineering standup, CEO decides in writing.” That structure, not the number, showed SRE maturity. She got $170,000 base, 0.18% equity, and a signing bonus of $30,000.amazon.com/dp/B0GWWJQ2S3).