Tier 1 SaaS support: the real metric for when to hire vs outsource

Every SaaS founder at Series A or Series B hits the same fork in the road on customer support. The CSM or support engineer who has been doing everything is buried, the founder is getting escalations they should not see, and someone in a leadership meeting says “we need to decide: hire another support engineer or outsource Tier 1.”

This is the wrong frame. The decision is not hire vs outsource. The decision is what kind of tickets you have. Most founders skip that step and pick based on cost, then six months later are paying twice because the first decision was structurally wrong.

The actual deciding metric

The metric that decides hire vs outsource is ticket complexity heterogeneity.

Translated: how much do your tickets vary from one to the next? Not the volume. Not the average difficulty. The variance.

If your tickets are mostly the same shape ticket-to-ticket (password resets, billing questions, “how do I do X in your product”), they are documentable. SOPs work. AI assistants pre-draft responses. The work is repeatable, which means it is outsource-friendly, regardless of volume.

If your tickets vary heavily (custom integrations, data-import edge cases, regulated-industry compliance questions, multi-tenant config bugs), they are not documentable in any useful way. Every ticket requires real product depth. SOPs are a starting point at best. The work is not repeatable, which means it has to live close to the product, which means it has to live in-house.

Volume tells you how many people you need. Variance tells you whether those people can be a team you do not manage.

When to hire in-house

You should hire a support engineer in-house when most of these are true:

• Tickets require reading the customer's actual data to answer
• Escalations regularly become engineering tickets
• A wrong answer would cost the customer real money or create real risk
• Your product is in a regulated space (fintech, healthtech, legaltech)
• Your support engineer is also doing onboarding, implementation, or customer success work

This is the “deep product” profile. The support engineer has to be a near-engineer themselves. You cannot SOP your way through this. You are buying judgment, not throughput.

Cost: $90K to $140K loaded for a mid-level support engineer in the US, or $50K to $80K in Eastern Europe or LatAm. Time to productive: 8 to 12 weeks.

When to outsource to a Pod

You should outsource Tier 1 when most of these are true:

• 70%+ of your tickets fall into 10 to 15 categories that repeat weekly
• A new hire could realistically handle the most common categories after 2 weeks of training
• The founder or a senior CSM is currently the escalation path for things that should not need them
• Volume is growing faster than headcount can keep up
• Your product surface is bounded (one or two main workflows, not 30)

This is the “high-volume, documentable” profile. The work is not less skilled, it is differently structured. A Pod with the right SOPs, AI ticket triage, and a senior Pod Operations Lead (POL) handles this better than a Tier 1 hire would, because the Pod brings the SOP library, the QA layer, and the AI tooling pre-built. You skip the 8-week onboarding curve.

Cost: roughly equivalent to one mid-level US support engineer, but you get the Pod (the POL plus specialists) rather than a single person. The leverage is in the structure, not the per-hour rate.

The mistake most founders make

The mistake is solving for cost first. The cost difference between an in-house hire and a Pod, at the seat level, is real but not gigantic. The structural difference is enormous.

If you hire in-house when you should have outsourced, you over-pay for repeatable work and you bottleneck on a single human. They take vacation, support stops. They quit, support breaks. They have a bad week, your CSAT moves.

If you outsource when you should have hired, you under-staff for complexity. The Pod handles 70% of tickets fine but the 30% that needed real product depth become escalations to your founders, which is exactly the problem you were trying to solve.

The cost question matters. It is the third question, not the first.

How to actually run the diagnostic

Pull your last 200 closed tickets. Categorize them by what kind of response they required:

1. Documented answer (the SOP-or-knowledge-base could have answered)
2. Reasoned answer (required understanding the customer's context, but no engineering)
3. Engineering answer (required reading product code, customer data, or filing a bug)

If category 1 is 60%+, you have outsource-friendly tickets. Pod fits. If category 3 is 30%+, you have hire-in-house tickets. Pod will not help on those. If you are in between, you need both, and the cost-effective sequencing is to outsource Tier 1 first (relieves the in-house person) and then hire the in-house person to focus exclusively on Tier 2 and Tier 3.

This 30-minute analysis decides the next $150K of payroll. Most founders skip it.