If you work with platforms and infrastructure, you know the struggle to get long-term investment approved. People outside your domain don’t understand the benefit of the work, or they do understand but somehow other investments take priority, and you can only motivate short-term optimization when things start going wrong.

The issue isn’t that what you propose is bad, it’s that infrastructure investment gets treated like an expense instead of what it actually is, an investment with measurable returns. Once you understand where infrastructure spending lands in the financials and how leadership thinks about it, you can reframe your proposals in a way that actually gets them approved.

Why This Keeps Happening

Infrastructure investment tends to be a tough sell for four reasons:

1. It’s invisible in the financials Your €300K production infrastructure investment sits in the same COGS bucket as routine AWS bills and maintenance work, so there’s no line item saying “Infrastructure Investment with Clear ROI,” it just looks like costs went up.

2. Benefits are delayed and hard to quantify You optimize production architecture and pages load 200ms faster, you upgrade production databases and avoid future capacity issues, you improve production monitoring and catch issues before customers do. All real benefits, but hard to quantify in euros upfront.

3. Prevention isn’t celebrated If you spend €500K to prevent a €5M production outage, leadership just sees €500K in costs because there’s no outage to point to, and nobody gets promoted for preventing problems that never happen.

4. Deferring is easier than approving Approving means taking responsibility now and defending it in quarterly reviews, while deferring kicks the problem down the road and makes it someone else’s problem later. The executive who defers doesn’t get blamed for anything (yet).

The companies that consistently invest in production infrastructure have better margins long-term, scale more efficiently, deliver better customer experiences, and handle growth without constant firefighting. Building great infrastructure is part of the work, making the business case compelling enough to get approved is the other part.

Where Infrastructure Spending Goes (And Why That Matters)

The core issue is that infrastructure investment flows into Cost of Goods Sold (COGS), which is the direct cost of delivering your service to customers. When COGS goes up, your gross margin goes down, and that’s the number leadership cares about most.

What Goes Into COGS for SaaS?

PRODUCTION costs (serving customers today): • Production cloud hosting (AWS, Azure, GCP) • DevOps teams keeping production systems running • Production database costs and optimization • Production monitoring, security, and backups • Customer support and technical support teams

NOT COGS (these go to R&D/Engineering OpEx): • Dev, test, and staging environments • Build systems and CI/CD pipelines • Internal developer tooling • Code refactoring for developer productivity

Your gross margin shows up on the income statement like this:

​
Revenue €10,000,000
- COGS (€2,000,000) ← Your production infra costs are here
═══════════════════════════════════════════
= Gross Profit €8,000,000
Gross Margin 80%
​

That 80% gross margin number matters because it determines your company valuation (higher margins equal higher multiples), it shows whether your business can scale profitably, and it’s what you have left to invest in sales, marketing, R&D, and everything else. Investors compare companies on this metric, and every percentage point counts.

So when you propose an infrastructure investment that increases COGS from €2M to €2.3M, leadership sees gross margin dropping from 80% to 77%, which looks like the business is getting worse.

What they don’t see is the future cost savings (next year’s COGS drops to €1.5M giving you an 85% margin), improved reliability (fewer outages mean less revenue lost), better production performance (faster page loads mean higher conversion), reduced incident response costs (less time firefighting production issues), and competitive advantage (better customer-facing performance than competitors).

Two Strategies That Work

Strategy 1: Separate “Investment” from “Run Rate”

The problem: Your €300K investment gets lumped in with routine operational costs, so it just looks like expenses went up.

The fix: Request that your finance partner break down COGS into two categories in your reporting, recurring run-rate costs and one-time investment.

How to do it:

Reason with your finance partner how you can break down COGS to help leads see that operational efficiency isn't declining but instead that temporary investments will improve baseline over time. Here is an example you can use:

COGS Breakdown:

• Recurring/Run-rate COGS: €1,700,000 (17% of revenue)
• Infrastructure Investment: €300,000 (3% of revenue, one-time)
  ​Total COGS: €2,000,000 (20% of revenue)

→ Normalized Gross Margin (without investment): 83%​
→ Current Gross Margin (with investment): 80%

Leadership sees that your baseline efficiency is actually 83%, not 80%, so the 3-point dip is temporary and margins will bounce back (or improve) once the investment pays off.

Show the trajectory, not just the snapshot:

Gross Margin Over Time:

• Q1: 80% baseline
• Q2: 77% investment period ← You are here
• Q3: 78%
• Q4: 82%
• Q1 (next year): 84%
• Q2 (next year): 85% new baseline

Comparison:​
Without investment: flat at 80%
With investment: trending to 85%

Your message: “Yes, margins dip this quarter, but we’re investing in a structurally better cost base that gives us 5 additional margin points long-term.”

Strategy 2: Frame It as “Cost of NOT Investing”

The problem: Your proposal says “we need €500K” but doesn’t quantify what happens if we don’t spend it.

The fix: Reframe your proposal to lead with the risk and cost of inaction, then show the investment as the solution.

Before and after:

❌ Instead of this: “We need €500K to modernize our production database infrastructure”

✓ Say this: “Our current production database architecture puts €12M in annual revenue at risk due to scalability constraints. We’ve had 2 incidents in the past 18 months that cost €200K each. For a €500K investment, we eliminate this risk and reduce ongoing production hosting costs by €200K annually. Payback in 2.5 quarters.”

Use this template for your next proposal:

Here's the version without emojis:

INFRASTRUCTURE INVESTMENT PROPOSAL:

CURRENT STATE:

• COGS: €2M/year (20% of revenue)
• Production infrastructure costs: €800K/year
• Production performance: 99.5% uptime, 2-second page loads
• RISK: [Quantify the specific risk in euros - this is critical]

INVESTMENT REQUIRED:

• One-time: €400K (implementation and migration)
• Ongoing: +€50K/year (additional production monitoring/maintenance)

EXPECTED OUTCOMES (Year 1):

Production/COGS Impact:

• Infrastructure costs: €600K/year (25% reduction in production hosting)
• Performance: 99.9% uptime, 500ms page loads (4x faster for customers)
• Incident response: 60% reduction in production incidents
• Revenue protection: Eliminates scalability risk on €12M annual revenue

FINANCIAL IMPACT:

Year 1:​
Investment: (€400K)
COGS savings: €200K
​Net impact: (€200K)

Year 2+:​
Annual COGS savings: €200K
Ongoing maintenance: (€50K)
​Net annual benefit: €150K/year

→ Payback period: 2 years (on COGS savings alone)
→ 3-year NPV: €250K

RISK MITIGATION:

• Eliminates risk of €5M production outage (occurred twice in 18 months)
• Removes scaling bottleneck for 3x customer growth
• Reduces on-call burden 60% (improves retention, reduces hiring costs)
• Improves customer experience (page load times = conversion/retention)

You’re speaking their language with risk, payback period, and NPV (Net Present Value equals the total value of future cash flows in today’s money), and you’ve quantified the cost of inaction, not just the cost of action.

Three more examples of rephrasing:

Example 1: Production Database Optimization

❌ Before: “We need to optimize our production database queries”

✓ After: “Our checkout flow currently takes 3 seconds due to database performance. Industry data shows every 100ms delay costs us 1% in conversion. Optimizing production queries to 500ms equals a 2.5% conversion lift, which equals €400K additional annual revenue. Investment: €150K. Payback: 4.5 months.”

Example 2: Production Monitoring & Observability

❌ Before: “We need better production monitoring tools”

✓ After: “Last quarter, we had 4 hours of production downtime that cost €200K in lost revenue. Better monitoring would have reduced MTTR (Mean Time To Recovery) from 4 hours to 30 minutes, saving €175K. This happens roughly 3 times per year. Investment: €100K/year. ROI: 5x.”

Example 3: Production Architecture Modernization

❌ Before: “We need to modernize our production architecture to reduce technical debt”

✓ After: “Our current production architecture requires 40% of DevOps time on firefighting versus proactive work. This costs us €400K/year in incident response and creates €2M in outage risk. Modernizing costs €300K but reduces incident response by 60%, which equals €240K/year savings. Payback: 15 months.”

What challenges have you faced getting infrastructure investment approved? I’d love to hear your stories and what’s worked for you.