The SaaS Tax Is Compounding: A CFO's Guide to the CapEx Alternative

The Compounding Problem Nobody Models

When a CFO evaluates a SaaS contract, they see the Year 1 number. A CRM at $150/user/month. An ERP at $200/user/month. Manageable. But SaaS pricing doesn't stay flat. It compounds.

Salesforce has raised enterprise prices an average of 9% annually over the past five years. ServiceNow's average contract value has grown 20% YoY. Microsoft 365 E5 has increased 15% since 2023. And that's before the new wave: AI add-on pricing. Salesforce Einstein AI costs $50/user/month on top of existing licenses. Microsoft Copilot adds $30/user/month. These aren't included in the base price — they are new line items that didn't exist two years ago.

For a 200-person organization running a standard enterprise stack (CRM, ERP, HRMS, ITSM, collaboration, analytics), the numbers look like this:

• Year 1 SaaS spend: $1.44M
• Annual compounding rate: 17% (price increases + seat growth + AI add-ons)
• Year 5 SaaS spend: $3.13M/year
• Year 10 SaaS spend: $6.82M/year
• 10-year cumulative SaaS spend: $24.1M

That $1.44M "manageable" number turns into $24.1M over a decade. And at the end of those ten years, you own nothing. Stop paying, and you lose access to your own data workflows, your integrations, your customizations, and your institutional knowledge embedded in the system.

SaaS is not a cost. It is a compounding liability with no residual value.

The CapEx Alternative

Perpetual licensing is not a new concept. It is how enterprise software worked for decades before SaaS. You buy the license once, you own it forever, and you pay an annual maintenance fee (typically 15-20% of the license cost) for updates and support.

The financial structure is fundamentally different:

• CapEx treatment: The license purchase is a capital expenditure. It goes on the balance sheet as an asset and is depreciated over 3-5 years. This improves operating margins and EBITDA compared to pure OpEx SaaS spend.
• Predictable costs: Maintenance fees are contractually capped. There are no per-seat surprise increases, no AI add-on pricing, no mid-contract tier changes.
• Residual value: At the end of the depreciation period, you still have a working system. You own the software. It runs on your infrastructure. There is no vendor kill switch.

The 10-Year Comparison

* Year 1 perpetual cost includes license purchase ($2.4M) + first year maintenance ($0.4M). Subsequent years are maintenance only.

The crossover point happens in Year 2. By Year 3, the perpetual model has already saved $1.6M in cumulative spend. By Year 10, the delta is $17.9M. That is not a rounding error. That is the difference between funding two additional engineering teams or not.

Why SaaS Vendors Don't Want You to See This Math

SaaS vendors have spent twenty years training the market to think in monthly per-seat costs. This framing is deliberately myopic. It hides three mechanisms that drive the compounding:

1. Contractual Price Escalators

Most enterprise SaaS contracts include annual price increase clauses of 5-10%. These are buried in renewal terms and often not negotiable for mid-market buyers. Over 10 years, a 7% annual escalator turns a $100/seat price into $197/seat. The base price nearly doubles without adding a single user.

2. Seat-Based Expansion

SaaS pricing is per-seat, per-month. Every new hire across the organization adds to the bill. For a company growing 10% annually, headcount-driven SaaS expansion is automatic and unavoidable. You cannot hire a new salesperson without adding a CRM seat, a collaboration seat, an ITSM seat, and a security seat.

3. AI Add-On Pricing

This is the newest and most aggressive compounding mechanism. Every major SaaS vendor has introduced AI capabilities as a premium add-on: $30-75/user/month on top of existing licenses. For a 200-person org, Microsoft Copilot alone adds $72K/year. Salesforce Einstein adds $120K/year. These costs did not exist in 2023. They will be unavoidable by 2027.

The Depreciation Advantage

For CFOs, the CapEx treatment of perpetual licenses offers a structural advantage that SaaS cannot match. A $2.4M perpetual license purchase can be capitalized and depreciated over 3-5 years under IAS 38 / ASC 350. This means:

• Balance sheet asset: The license appears as an intangible asset, improving the company's asset base.
• Depreciation deduction: Annual depreciation of $480K-800K (depending on schedule) reduces taxable income.
• Operating margin improvement: The annual SaaS expense disappears from the income statement. Only the depreciation charge and maintenance fee appear as costs.
• Cash flow timing: While Year 1 requires a larger upfront payment, cumulative cash outflows are lower from Year 3 onward.

For companies that care about EBITDA — which is most companies in a debt-financed or PE-backed environment — the shift from OpEx to CapEx is not just a tax optimization. It is a structural improvement to the financial profile of the business.

The Board Conversation

When presenting the CapEx alternative to a board, the framing matters. This is not a technology decision. It is a capital allocation decision. The question is not "Should we switch software?" The question is "Should we continue renting $24M of software over 10 years, or invest $6.2M to own it?"

Framed as a capital allocation question, the answer is straightforward. The perpetual model offers a 74% reduction in 10-year TCO, converts a recurring liability into a depreciable asset, and eliminates vendor pricing power over the organization.

The best time to exit the SaaS compounding curve was five years ago. The second best time is now.

Every quarter you delay, the cumulative cost gap widens. The compounding doesn't wait.