In the first two articles of this series, I looked at SaaS economics from the inside: how ARR, ARPU, churn, CAC, LTV and CAC payback work in “classic” SaaS models and how AI and LLM costs reshape the P&L of AI SaaS.

In this third article, I want to switch perspectives. Instead of asking “Is my SaaS/AI SaaS viable?” from the founder’s side, I want to ask a different question: “How does an investor look at this business?” Whether the investor is a VC fund, a growth equity firm or a strategic buyer, the core concerns are similar:

• Is the business solving a real problem?
• Is growth efficient or subsidised?
• Are the margins defensible?
• And in AI SaaS, is there any durable advantage beyond access to the same LLMs everyone else is using?

This article is not meant to be a fundraising checklist. It is a framework to understand how your numbers and your story are read from the other side of the table, especially in 2026, when both skepticism and competition in AI are higher than ever.

1. The Metrics That Still Matter (AI or Not)

Let us start with something simple: the core SaaS metrics investors cared about ten years ago still matter today. AI has not changed that.

1.1 ARR and growth quality

Investors still look at:

• ARR (Annual Recurring Revenue): the size of the recurring revenue base.
• ARR growth: not only the percentage but the quality of that growth.

“Quality” means:

• Is growth coming mostly from new customers, or from expansion within existing accounts?
• Is growth concentrated in a few large customers?
• Is growth heavily dependent on discounting or promotional campaigns?

High ARR with low diversification and weak expansion is not as attractive as slightly lower ARR with strong net revenue retention and a diversified base.

1.2 Net Revenue Retention (NRR) and churn

NRR (Net Revenue Retention) has become a central metric:

• NRR above 100% means your existing customers, as a group, are expanding.
• NRR above 120% in B2B is often considered very strong.

A combination of low gross churn and healthy expansion indicates product-market fit and room for upsell. In AI SaaS, investors will ask:

• Are customers expanding because they see more value, or because they are forced into higher tiers due to usage caps or tokens?
• Is churn starting to increase as the “AI novelty” effect fades?

1.3 CAC, CAC payback and LTV/CAC

CAC, CAC payback and LTV/CAC are still at the heart of the investment case.

• CAC (Customer Acquisition Cost): all-in cost to acquire a paying customer.
• CAC payback: how many months of gross profit it takes to recover that CAC.
• LTV/CAC: a summary ratio of value versus acquisition cost.

For context (rough, not dogmatic):

• CAC payback < 12 months: strong.
• 12–18 months: acceptable, depending on segment and margins.
• 24 months: increasingly hard to defend, unless growth and retention are exceptional.

AI does not change the mathematics. It changes the inputs (gross margin, ARPU, retention) and the volatility of those inputs.  

2. Additional Metrics for 2026: Efficiency and Discipline

Beyond the classics, investors in 2026 often emphasise efficiency metrics: how effectively the company turns spend into growth.

2.1 Burn multiple

The burn multiple relates cash burn to net new ARR: Burn multiple = Net cash burn / Net new ARR A burn multiple around 1–1.5 is often viewed positively in growth or later-stage companies. Ratios above 2–3 may indicate that growth is too expensive or that the company has not adapted to the new funding environment. In AI SaaS, front-loaded investments (infra, team, experimentation) can temporarily increase burn, but investors will want to see:

• A credible path to a lower burn multiple.
• Evidence that the company can dial down growth spend and still maintain a reasonable trajectory.

2.2 Rule of 40 (with nuance)

The Rule of 40 is a simple heuristic: Revenue growth (%) + EBITDA margin (%) ≥ 40 In 2026, many investors treat it more as a reference than as a strict rule, but the spirit remains. For AI SaaS, the nuance is:

• Investors will check if gross margin is truly SaaS-like (>75–80%) or eroded by AI costs.
• A strong Rule of 40 score with weak gross margins may raise questions about sustainability.

3. Stage by Stage: What Investors Expect to See

Expectations vary by stage. A seed-stage founder is not judged by the same metrics as a Series B company. The following table is illustrative, not a universal standard, but it can help frame where you stand. Table 1 – Typical investor focus by stage (SaaS / AI SaaS)

Stage	ARR range (indicative)	Focus metrics & signals
Pre-seed	0 – 100k	Problem clarity, early usage, strong founder insight
Seed	50k – 500k	Early ARR, usage intensity, initial CAC vs ARPU logic
Series A	500k – 3M	Growth rate, NRR, CAC payback, gross margin
Series B	3M – 10M	NRR > 110–120%, efficient CAC, burn multiple, team
Series C+	10M+	Scalability, Rule of 40, market position, defensibility

For AI SaaS, add to each stage:

• Clarity on LLM and infra costs relative to ARPU.
• A credible model and data strategy (beyond “we use the latest GPT”).
• Evidence that AI features drive retention and expansion, not just trials.

4. Reading an AI SaaS P&L Through an Investor’s Lens

Let us imagine an investor reviewing the P&L of an AI SaaS company with:

• ARR: 4M USD
• YoY ARR growth: 70%
• Gross margin: 78%
• LLM and related AI infra costs included in COGS
• CAC payback: 14 months
• Burn multiple: 1.8

4.1 Top line: ARR and growth

A 4M USD ARR business growing at 70% is attractive on the surface. Questions an investor will ask:

• Is this growth coming from net new customers or from expansion?
• How concentrated is ARR across the top 10–20 customers?
• Has growth been consistent, or driven by a few large deals or one-off projects?

In AI SaaS, a further question:

• How much of that ARR is tied to mission-critical workflows vs. “nice to have” experiments with AI?

4.2 Gross margin: is this really SaaS?

A 78% gross margin is decent, but slightly below the 80–85% range of top-tier SaaS. An investor will want to know:

• How much of COGS is LLM/API cost vs. traditional infra and support?
• Does the company have levers to improve that margin (pricing, model mix, usage policies)?
• Is gross margin stable, improving, or eroding over time as usage grows?

If LLM costs are creeping up faster than revenue, that is an early warning sign.

4.3 Sales & marketing: CAC and efficiency

With CAC payback at 14 months: This sits in the “acceptable but needs to be justified” range. In AI SaaS, investors will ask:

• What is the free-to-paid conversion rate?
• What percentage of trial users become retained paying users after 6–12 months?
• Does intensive AI usage correlate with higher retention and ARPU, or just with higher COGS?

4.4 Operating expenses and path to profitability

With a burn multiple of 1.8, the company is burning: 1.8 USD of net cash for every 1 USD of net new ARR. An investor will ask:

• If growth slows from 70% to 40%, can the company reduce burn significantly?
• How flexible are Opex lines (sales & marketing, R&D, G&A)?
• Is there a credible path to breakeven at a certain ARR level?

5. AI-Specific Questions Investors Are Asking in 2026

Beyond generic SaaS metrics, investors looking at AI SaaS have a set of AI-specific questions around risk and defensibility.

5.1 Dependency on LLM providers

• What percentage of COGS is directly tied to a single LLM provider?
• Does the company have multi-model, multi-vendor capability?
• How complex and expensive would it be to switch models or providers?

5.2 Data advantage (or lack thereof)

• Access to proprietary data that improves the product in a meaningful way.
• Mechanisms to safely and legally use that data (consent, anonymisation, governance).
• A path to compound advantage.

5.3 Product and workflow defensibility

• Is the product deeply integrated into customer workflows?
• How hard would it be for a larger platform to replicate the core value?
• Does the company own a critical piece of the workflow or data flow?

6. The Role of Narrative: How Founders Should Tell Their AI SaaS Story

A solid AI SaaS narrative in 2026 usually has three layers.

6.1 Problem, segment, and workflow

• A clear definition of the problem.
• A well-defined segment.
• A concrete description of daily workflows.

6.2 Economic logic and unit economics

• “At steady state, a typical customer pays us X per month, costs us Y to serve (including AI), and stays with us Z months on average.”
• “We recover CAC in N months.”

6.3 Moat and medium-term strategy

• Product differentiation as models commoditise.
• Role of proprietary data and integrations.
• Evolution of pricing and packaging.

7. How Founders Can Prepare for Investor Conversations in AI SaaS

• Build a clear metrics dashboard.
• Run sensitivity scenarios.
• Refine the story.

8. Conclusion: AI SaaS Through an Investor’s Eyes

From an investor’s perspective, the most compelling AI SaaS companies in 2026 are those where:

• SaaS fundamentals are solid.
• AI costs are understood and managed.
• The moat is built around workflows, data and relationships.