The Stack Tax
Every SaaS subscription is a tax on your runway. Your Stack Bleed Rate tells you when to evict.
Quick Answer
The Stack Tax is the cumulative monthly cost of every SaaS subscription you carry. The Stack Bleed Rate formula tells you when it's too much. Stack Bleed Rate = (monthly SaaS spend ÷ headcount) ÷ MRR per employee. Below 5% is healthy. 5–12% is the watch zone. Above 12% means you're paying to grow slower — every dollar of new revenue is partially consumed before it can compound. As of mid-2026, most $1M–$50M businesses I audit run between 8% and 22%.
Coined by Justin McKelvey, fractional CTO · Updated June 2026
Often Confused With
The Stack Bleed Rate is not burn rate (rate at which a company depletes cash reserves), not churn rate (rate at which customers cancel subscriptions), and not net cash burn (operating expenses minus revenue). The Stack Bleed Rate is specifically SaaS subscription cost per employee, expressed as a percentage of revenue per employee. It measures the recurring tax your software stack imposes on your unit economics — not your company's overall cash health.
What is the Stack Tax?
The Stack Tax is the literal sum of every SaaS line item on your monthly card statement. Accounting tools. Project management. CRM. Email. Analytics. Scheduling. The long tail of $9–$49/month tools nobody remembers signing up for. All of it.
It's a tax in the literal sense: a recurring cost on every dollar of revenue, automatically deducted before any of it reaches the bottom line. Unlike payroll, it doesn't scale with output. Unlike rent, it doesn't shrink when you renegotiate. It just compounds quietly.
Most operators don't know their Stack Tax to the dollar. They know it feels high. The Stack Bleed Rate formula turns that feeling into a number you can act on — usually a number that's worse than they hoped and easier to fix than they feared.
The Formula
What is the Stack Bleed Rate formula?
Stack Bleed Rate =
(monthly SaaS spend ÷ headcount)
÷ MRR per employee
What is a healthy Stack Bleed Rate?
| Range | Status | Action |
|---|---|---|
| Below 5% | Healthy | No action — keep an eye on additions. |
| 5%–12% | Watch zone | Quarterly audit. Justify each tool. |
| 12%–20% | Paying to grow slower | Cut, consolidate, or replace. Start now. |
| Above 20% | Structurally limited | Biggest single line item first. Don't try to fix it all at once. |
How do you calculate your Stack Bleed Rate? (Worked examples)
Example 1 · Healthy
A SaaS company spending $4,000/month on tools, 8 employees, $50,000 MRR per employee.
(4000 ÷ 8) ÷ 50000 = 1%
Example 2 · Watch zone
A services business spending $3,500/month on tools, 5 employees, $8,000 MRR per employee.
(3500 ÷ 5) ÷ 8000 = 8.75%
Example 3 · Paying to grow slower
An early-stage agency spending $5,000/month on tools, 2 employees, $8,000 MRR per employee.
(5000 ÷ 2) ÷ 8000 = 31.25%
How do you lower your Stack Bleed Rate?
Three levers. Pull them in this order:
-
Lever 1
Cut what doesn't justify itself
Quarterly audit. List every subscription. Write one sentence justifying each tool's value. Anything you can't justify in one sentence, cancel within the week. No "but we might need it." This typically removes 15–30% of the Stack Tax with zero functional loss. The first one is the hardest. The next twelve are easy.
-
Lever 2
Consolidate overlapping tools
Pick the smallest survivable set. If you have a project management tool, a separate task tool, a separate doc tool, and a separate calendar tool — most teams under 50 can collapse three of those into one. Each consolidation kills both the cost and the integration debt.
-
Lever 3
Replace SaaS with AI or self-built
The biggest swing as of 2026. A custom Claude project replaces a $99/month writing assistant. A simple Rails or Next.js app replaces a $200/month no-code internal tool. A self-hosted scheduling page replaces $30/month per seat. I built a personal example — replaced $2,200/year of SaaS with one Rails app costing $15/month. The math compounds: each replacement saves the subscription cost AND removes a vendor relationship to manage. The replacement doesn't have to be elegant. It has to be yours.
Frequently Asked Questions
Stack Tax FAQ
- What is the Stack Tax?
- The Stack Tax is the cumulative monthly cost of every SaaS subscription a business carries — accounting tools, project management, CRM, email, analytics, calendar scheduling, and the long tail of $9-$49/month tools nobody remembers signing up for. Coined by Justin McKelvey, fractional CTO, 2026. It's a tax in the literal sense: a recurring cost on every dollar of revenue, automatically deducted before any of it reaches the bottom line.
- What is the Stack Bleed Rate formula?
- The Stack Bleed Rate formula: (monthly SaaS spend ÷ headcount) ÷ MRR per employee = Stack Bleed Rate %. Worked example: a company spending $4,000/month on SaaS with 8 employees and $50,000 MRR per employee has a Stack Bleed Rate of (4000 ÷ 8) ÷ 50000 = 1% (healthy). A company spending $5,000/month with 2 employees and $8,000 MRR per employee has a Stack Bleed Rate of (5000 ÷ 2) ÷ 8000 = 31.25% — paying to grow slower.
- What's a healthy Stack Bleed Rate?
- Below 5% is healthy. 5–12% is the watch zone — survivable but worth auditing. Above 12% means you're paying to grow slower: every dollar of new MRR is partially consumed by stack costs before it can be reinvested into hiring or product. Above 20% is critical — the tax is structurally limiting growth. These thresholds hold for most $1M–$50M businesses; very early-stage companies and pure-infrastructure businesses have different math.
- How do I lower my Stack Bleed Rate?
- Three levers. (1) Cut subscriptions — quarterly audit, justify every tool in one sentence, kill anything that can't justify itself. (2) Raise revenue per employee — usually means pricing changes or moving up-market, not headcount cuts. (3) Replace SaaS with self-hosted or AI-built alternatives — particularly for low-complexity tools that have become commodity (newsletter, scheduling, simple CRM). The third lever has the biggest swing as of 2026.
- Can AI tools replace SaaS subscriptions?
- Many of them, yes — for most $1M–$50M businesses. A custom Claude project replaces a $99/month writing assistant. A simple Rails or Next.js app replaces a $200/month no-code internal tool. A self-hosted scheduling page replaces $30/month per seat. I built a personal example — replaced $2,200/year of SaaS with one Rails app costing $15/month. The math compounds: each replacement saves the subscription cost AND removes a vendor relationship to manage.
- When should I cut a SaaS subscription?
- When you can't justify it in one sentence. When the original use case has changed. When team adoption is below 30%. When the tool's job-to-be-done is now handled by another tool in your stack (consolidation). When the tool ships a price hike that's larger than the value it provides. And — critically — when your Stack Bleed Rate is above 12% and you have to choose where to cut.
- Does the Stack Tax include AI subscriptions?
- Yes. Claude Pro, ChatGPT Team, Cursor, GitHub Copilot, Perplexity Pro — all count. The Stack Bleed Rate doesn't care what category a tool is in; the math is the same. AI tools tend to have very high ROI per dollar in 2026, so they often survive a Stack Tax audit while older subscriptions get cut.
- What if my Stack Bleed Rate is over 20%?
- You're structurally constrained on growth. Every new hire makes the math worse (more tools per person). Every new customer pays partially into your stack, not into your runway. Fix in this order: (1) cut what's clearly unused this week, (2) consolidate overlapping tools next month, (3) replace one expensive SaaS tool with an AI or self-built equivalent this quarter. Don't try to fix everything at once — pick the biggest single line item and start there.
Related reading
Why I replaced $2,200/year of SaaS with a single Rails app
The case study — applying Lever 3 to my own business.
AI for Business Owners — strategy and workflows for $1M–$50M operators
The broader operating playbook the Stack Tax fits into.
The Operator's Ladder — 5 archetypes of business owners on the AI curve
Where you are on the ladder shapes which Stack Tax lever to pull first.
Related frameworks.
The other named frameworks in the McKelvey AI Operating System. Each names a specific failure pattern operators hit at a specific stage.
Vibe Debt
The technical debt every vibe-coded MVP carries the moment it meets a real user.
The Crash Cart
The triage protocol when a vibe-coded production app starts breaking and you have hours, not days.
The Operator's Ladder
Five archetypes of business owners on the AI adoption curve. Skip a rung and you fall.
The Founder's Build Order
Three-stage decision matrix: vibe-code the prototype, hire a dev for v1, fractional CTO before scale. Never reverse it.
All five frameworks: the McKelvey AI Operating System · Every defined term: the Vibe Coding Glossary
Want help cutting your Stack Tax?
Thirty minutes. We'll walk through your stack, calculate your Stack Bleed Rate live, and identify the biggest single line item to cut, consolidate, or replace.
Book a strategy call