You know what's weird? Every developer tool prices differently—Stripe charges per transaction, Twilio charges per message, AWS charges per compute hour. But support software? Everyone charges per seat.
Zendesk, Freshdesk, Intercom, Help Scout—they all want $50-150/month for each person who can look at a support ticket. And we've just... accepted this?
Let me show you why this model is probably costing you way more than you think.
The Hidden Tax on Modern Dev Teams
Here's the thing about modern SaaS companies: support isn't a department anymore. It's everyone.
Who needs to see support tickets?
- Engineers debugging reported bugs
- PMs tracking feature requests
- Founders responding to VIP customers
- DevOps checking if outages were reported
- Sales answering pre-purchase questions
Under per-seat pricing, each of these people needs a $79/month seat. Even if the PM touches 5 tickets a month. Even if the engineer just needs to see the stack trace someone pasted.
// What your costs look like
const team = {
supportAgents: 3, // These actually need it
engineers: 4, // Need occasional access
productManagers: 2, // Review feature requests
founders: 2, // VIP escalations
};
const seatsNeeded = Object.values(team).reduce((a, b) => a + b, 0);
const monthlyBill = seatsNeeded * 79;
console.log(`${seatsNeeded} seats × $79 = $${monthlyBill}/month`);
// 11 seats × $79 = $869/month
That's $869/month so your engineers can occasionally check if a bug was already reported.
The Workarounds Everyone Uses (And Why They're Worse)
When seats are expensive, teams get creative:
# The shared login approach (don't do this)
export ZENDESK_LOGIN=support@company.com
export ZENDESK_PASSWORD=hunter2
# "Just use the team account"
Shared logins - No audit trail, security nightmare, probably violates your SOC2 compliance.
Screenshot relay - Engineer: "Can you forward me that ticket?" Support: takes screenshot, pastes in Slack - Now you have context in three different places.
The human API - Support becomes a proxy layer between your team and customer feedback. Every escalation adds latency.
All of this operational overhead doesn't show up on your SaaS bill. But it's real.
Let's Talk Numbers
I'll run through three scenarios and you can see which one looks like your company:
Scenario A: Small SaaS (500 tickets/month)
Per-seat model:
3 agents + 5 occasional users = 8 seats
8 × $79 = $632/month
Cost per ticket: $1.26
Per-ticket model:
$29/month for 1,000 tickets
All 8 users included
Cost per ticket: $0.06
Delta: -$603/month (-95%)
Scenario B: Growing E-commerce (3,000 tickets/month)
Per-seat model:
5 agents + 8 occasional users = 13 seats
13 × $79 = $1,027/month
Cost per ticket: $0.34
Per-ticket model:
$99/month for 10,000 tickets
All 13 users included
Cost per ticket: $0.03
Delta: -$928/month (-90%)
Scenario C: Agency with 20 clients (8,000 tickets/month)
Per-seat model:
4 staff + 15 client contacts = 19 seats
19 × $79 = $1,501/month
Cost per ticket: $0.19
Per-ticket model:
$99/month for 10,000 tickets
All 19 users included
Cost per ticket: $0.01
Delta: -$1,402/month (-93%)
The pattern is clear: the more people who need access, the worse per-seat pricing gets.
The AI Elephant in the Room
Here's where it gets really interesting. AI is changing support economics fast.
Let's say you implement an AI that handles 40% of your tickets automatically:
// Per-seat economics with AI
const ticketsBefore = 1000;
const ticketsAfterAI = ticketsBefore * 0.6; // AI handles 40%
const seatsNeeded = 8; // Still need humans for oversight
const perSeatCost = 8 * 79; // $632/month - UNCHANGED
// Per-ticket economics with AI
const perTicketCost = 29; // Base tier still covers 1000
const actualCost = ticketsAfterAI < 1000 ? 29 : 29 + (ticketsAfterAI - 1000) * 0.02;
// Cost drops as AI handles more!
With per-seat pricing, your AI investment saves you nothing on software costs. You still need the same seats for human oversight and edge cases.
With per-ticket pricing, your costs actually decrease as AI handles more volume. The efficiency gains translate directly to savings.
When Per-Seat Actually Makes Sense
I'm not saying per-seat is always wrong. It can work if:
- You have 1-2 users total - At that scale, $79/seat might be cheaper than a $99 base fee
- Extremely high volume, tiny team - 100,000 tickets handled by 5 people
- You need specific enterprise features - SSO, advanced workflows only available in legacy vendors
But honestly? Scenario 1 and 2 are pretty rare. Most companies have more people who should have support access than currently do.
The Operational Cost Iceberg
Here's what most cost comparisons miss—the hidden overhead:
What you see:
┌──────────────┐
│ Software bill │ ← $632/month
└──────────────┘
─────────────────────────────────────────
What you don't:
┌─────────────────────────┐
│ Escalation delays │ ← Waiting for access
│ Context handoffs │ ← Explaining issues
│ Slack threads about │ ← tickets
│ tickets not in │ ← the ticket system
│ Knowledge silos │ ← Engineers miss patterns
│ Customer frustration │ ← "Let me check with..."
└─────────────────────────┘
Studies show cross-functional tickets take 2-3x longer when access is restricted. If your average ticket costs $15 in labor and you add 30 minutes of coordination overhead, that's $7.50 per ticket in hidden costs.
TL;DR
Per-seat pricing charges for capacity, not usage. You pay the same whether someone handles 500 tickets or 5.
Modern dev teams have 3-4x more people who need support access than traditional "agents"
The workarounds (shared logins, screenshot relay, human proxies) add operational overhead that doesn't show on invoices
AI makes it worse—your per-seat costs stay flat while AI handles the easy tickets
Per-ticket pricing aligns with how dev tools should work: pay for what you use, not who might use it
Run the math for your team. Count everyone who should have access, not just who currently does.
Originally published at Dispatch Tickets. We're building an API-first ticketing system with per-ticket pricing—unlimited users included. Check it out if the per-seat tax is hitting your team.
Top comments (0)