Three Paths, Very Different Bills

Every support team eventually hits the same fork in the road. You're growing, tickets are piling up, and you need to decide: hire in-house, outsource, or automate with AI.

Each option has a price tag that looks simple on the surface. None of them actually are.

In-House: The Known Quantity

Hiring a full-time support agent in the US costs between $45,000 and $65,000 in base salary. But salary is maybe 60-70% of the real cost. Add employer-side payroll taxes (7.65% for FICA alone), health insurance ($6,000-$12,000/year for a company contribution), equipment, software licenses, and management overhead.

A realistic fully-loaded cost for one in-house agent: $65,000-$90,000/year.

One agent handles roughly 40-60 tickets per day, or about 10,000-15,000 per year. That puts your cost per ticket somewhere between $4.30 and $9.00.

The upside is real though. In-house agents understand your product deeply. They build institutional knowledge. They catch patterns that nobody else sees. For complex, high-value interactions, there's no substitute.

Outsourcing: Cheaper Per Hour, Tricky Per Outcome

BPO pricing varies wildly by region:

Those numbers look attractive compared to in-house. A Philippines-based agent at $10/hr costs roughly $19,200/year. That's a quarter of the US in-house cost.

But the hidden costs are significant. You'll spend 2-4 weeks training each agent. Turnover at BPOs runs 30-45% annually, so you're retraining constantly. Quality assurance requires dedicated staff. You'll need a point of contact managing the relationship. Communication gaps cause escalations.

Realistic outsourcing cost per ticket, including overhead: $2.50-$6.00.

AI: The Wildcard That's Getting Less Wild

AI support pricing has settled into two models: per-seat add-ons and per-resolution charges.

Per-resolution is easier to compare. Intercom's Fin charges $0.99 per resolution. Zendesk charges $1.50 on committed volume ($2.00 pay-as-you-go). Supp charges $0.20 per classification, $0.30 with an executed action.

At 1,000 tickets per month where AI resolves 70%:

The remaining 30% still needs humans. That's 300 tickets going to either in-house or outsourced agents.

The Breakeven Math

Here's where it gets interesting. Let's model a team handling 3,000 tickets/month.

A pure in-house approach needs roughly 3-4 agents at that volume. Cost: $16,000-$30,000/month fully loaded.

Pure outsourcing (Philippines): 3-4 agents at roughly $1,600/month each, plus management overhead. Cost: $6,000-$9,000/month.

AI handling 70% (2,100 tickets) plus 1 in-house agent for the rest:

With Supp ($0.20): $420 AI cost + $5,400 for one agent = $5,820/month. With Intercom Fin ($0.99): $2,079 + seat fees + $5,400 = $7,600+/month. With Zendesk ($1.50): $3,150 + agent fees + $5,400 = $9,100+/month.

The hybrid model wins at almost every volume level. AI handles the repetitive stuff (password resets, order status, basic how-to questions). Humans handle the rest.

Where Each Option Actually Makes Sense

In-house works best when your product is complex, your support interactions require deep product knowledge, your ticket volume is under 2,000/month, and your customers expect white-glove service. The quality advantage is worth the premium.

Outsourcing makes sense for high-volume, relatively standardized support. If you're an e-commerce company handling 10,000+ tickets about shipping and returns, a trained BPO team can handle that efficiently.

AI works best for the predictable slice of your support queue. Password resets, order tracking, account questions, basic troubleshooting. These follow patterns that classification and automation handle well.

Most teams over 1,000 tickets/month should be running a hybrid. Let AI handle the 60-70% of tickets that follow known patterns. Route the rest to humans who can actually think.

The companies paying $9 per ticket for an in-house agent to answer "where's my order?" are burning money. The companies outsourcing everything to save a few dollars are burning trust.

Pick your battles. Automate what's automatable. Put humans where humans matter.