How to Build a SaaS Financial Model: ARR, Cohorts, and Unit Economics

SaaS financial models look fundamentally different from traditional 3-statement models because recurring revenue, customer cohorts, and churn dynamics replace one-time transactions and linear growth. Understanding ARR builds, cohort retention curves, and unit economics (CAC, LTV, payback) is essential for forecasting cash, valuing the business, and communicating with investors.

Unlike manufacturing or retail companies, SaaS businesses do not recognise all revenue when the cash hits the bank. A customer who pays £12,000 upfront for a 12-month subscription generates £12,000 in cash immediately but only £1,000 in recognized monthly revenue. This mismatch between cash and revenue, combined with the science of churn and expansion, makes SaaS financial modelling a specialized discipline.

This guide walks you through the architecture of a production SaaS model: ARR rolls, cohort-level retention curves, customer-level unit economics, revenue recognition schedules, and cash flow reconciliation. We'll use a worked example of a Series B SaaS company ($5M ARR at the start of Year 1) to make each concept concrete.

What Makes SaaS Models Different

SaaS businesses operate on three core dynamics that reshape financial modelling:

1. Recurring Revenue: Revenue is predictable and recurring, not one-time. This is the company's greatest strength (high forward visibility) and its biggest risk (if churn accelerates, revenue collapses).
2. Customer Cohorts: Each cohort of customers acquired in a given month or quarter has its own retention curve and expansion rate. Blending all cohorts into a single churn number masks cohort deterioration.
3. Cash vs. Revenue Timing: Multi-year or annual upfront payments create a large deferred revenue liability on the balance sheet. Recognized revenue is much lower than cash received, and the two diverge in early growth stages.

These dynamics demand a different model architecture. Rather than building a single income statement, you build cohort-level retention tables, then roll them up into annual ARR, which flows into revenue recognition.

ARR Architecture: New Logos, Expansion, Contraction, and Churn

ARR (Annual Recurring Revenue) is the annualized value of all active subscriptions at a point in time. It flows as follows:

Each component has distinct drivers:

Worked example for a Series B SaaS company over Year 1:

Year 1 ARR Roll-Forward:

Starting ARR (from Year 0)                        £5,000,000
+ New Logos (100 customers × £50k ACV)            £5,000,000
+ Expansion (10% of existing £5M base)              £500,000
- Contraction (2% of existing base)               (£100,000)
- Churn (8% of existing base)                     (£400,000)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Ending ARR                                       £9,500,000

NRR Calculation:
(£5M + £500k - £100k - £400k) / £5M = 110%

This Year 1 build shows 110% Net Revenue Retention (NRR), a healthy expansion rate that outpaces churn—the hallmark of product-market fit.

Cohort Analysis: Retention Curves and Unit Economics by Acquisition Month

A cohort is a group of customers acquired in the same month or quarter. Cohort analysis reveals whether early cohorts are more valuable, whether retention is deteriorating, and where CAC payback breaks down.

Why Cohort Analysis Matters: If you blend all 100 new Year 1 customers into a single churn rate, you may hide the fact that January cohort customers have 75% 12-month retention but December cohort customers have only 45%. This signals declining product-market fit or a sales quality problem. Cohort analysis makes this visible.

Building a Retention Curve

For each month's cohort, track the percentage of customers retained month-by-month:

Each row represents one cohort's journey. The columns show retention at each month following acquisition. A downward slope to the right indicates churn. The slope steepens if product-market fit is declining.

Rolling Up Cohorts to Annual Retention and NRR

For forecasting purposes, you collapse each cohort's curve into an annual retention rate, then build forward by simulating new cohorts:

Year 1 Cohort Revenue Build:

Cohort      | Customers | ACV      | Month 0 ARR | 12M Retention | Year-End ARR
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Jan 2025    | 8         | £50,000  | £400,000    | 70%           | £280,000
Feb 2025    | 8         | £50,000  | £400,000    | 62%           | £248,000
Mar 2025    | 9         | £50,000  | £450,000    | 55%           | £247,500
Apr 2025    | 9         | £50,000  | £450,000    | 52%           | £234,000
May 2025    | 9         | £50,000  | £450,000    | 48%           | £216,000
Jun 2025    | 10        | £50,000  | £500,000    | 45%           | £225,000
Jul 2025    | 10        | £50,000  | £500,000    | 42%           | £210,000
Aug 2025    | 10        | £50,000  | £500,000    | 40%           | £200,000
Sep 2025    | 11        | £50,000  | £550,000    | 38%           | £209,000
Oct 2025    | 11        | £50,000  | £550,000    | 35%           | £192,500
Nov 2025    | 11        | £50,000  | £550,000    | 32%           | £176,000
Dec 2025    | 13        | £50,000  | £650,000    | 28%           | £182,000
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Total Year-End ARR (all cohorts)                            £2,920,000

Note: This is the ARR from new cohorts only. Add to the existing base of £5M.

In this example, even though 118 new customers were acquired (£5.9M in new ARR), 12-month retention across cohorts averages only ~50%, and expansion is not modelled. The company must acquire many customers simply to offset churn.

Unit Economics: CAC, LTV, Payback Period, and the Magic Number

Unit economics are the first questions investors ask because they determine whether the business can grow profitably.

Customer Acquisition Cost (CAC)

CAC is the blended cost to acquire one customer:

CAC Calculation:

Total S&M Spend (Year 1)                          £2,500,000
Divided by New Customers Acquired                 100
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
CAC (per customer)                                £25,000

Note: Some models allocate only variable S&M (commissions, ads),
others include full S&M (salaries, marketing). Be consistent.

Lifetime Value (LTV)

LTV is the total profit a customer generates over their lifetime:

LTV Calculation (Simplified):

Annual Revenue Per Customer (ARPU)                £50,000
Multiplied by Gross Margin %                      × 80%
Gross Profit Per Customer Per Year                £40,000
Divided by Annual Churn Rate (%)                  ÷ 8%
(Equivalent to 12.5-year average lifetime)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
LTV (simplified)                                  £500,000

Note: This assumes churn is constant and does not account
for expansion, discounting, or acquisition cohort quality.
More rigorous models discount cash flows and use
cohort-specific retention curves.

A more rigorous formula accounts for cohort retention curves and discounting:

LTV Calculation (Detailed):

Month   | Customers Retained | Monthly Revenue | Discounted CF (5% rate)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
M1      | 92 customers       | £4,600          | £4,386
M2      | 85 customers       | £4,250          | £3,858
M3      | 80 customers       | £4,000          | £3,461
...     | ...                | ...             | ...
M60     | 42 customers       | £2,100          | £165
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Sum of Discounted Cash Flows (before COGS and S&M)  £165,000
Multiplied by Gross Margin                           × 80%
LTV (conservative)                                   £132,000

The detailed method is more conservative because it discounts future cash and accounts for actual cohort retention.

CAC Payback Period

CAC payback is the number of months to recover the acquisition cost:

CAC Payback Calculation:

Customer Acquisition Cost                         £25,000
Divided by Monthly Gross Profit Per Customer      ÷ £3,333
(£50k ARPU × 80% gross margin ÷ 12 months)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
CAC Payback Period (months)                       7.5 months

Note: A payback period under 12 months is considered excellent.
Above 18 months, the unit economics become questionable unless
LTV is exceptionally high or churn is declining rapidly.

Benchmark: CAC payback of <12 months is venture-scale; 12–18 months is acceptable; >18 months signals unit economics problems.

The Magic Number (Sales Efficiency)

The Magic Number shows how much ARR growth you get per pound of S&M spend:

Magic Number Calculation:

Year 1 Ending ARR                                 £9,500,000
Minus Year 0 Ending ARR                           £5,000,000
Net ARR Growth                                    £4,500,000

Divided by Year 1 S&M Spend                       ÷ £2,500,000
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Magic Number                                      1.8x

Interpretation: For every pound spent on S&M,
the company generated £1.80 in ARR growth.

Benchmark: >0.75x is good; >1.0x is excellent; >1.5x is exceptional.

LTV:CAC Ratio

The LTV:CAC ratio tells investors how capital-efficient the business is:

LTV:CAC Ratio:

Lifetime Value (conservative, from detailed model)   £132,000
Divided by Customer Acquisition Cost                 ÷ £25,000
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
LTV:CAC Ratio                                        5.3:1

Interpretation: For every pound spent to acquire a customer,
the company generates £5.30 in lifetime profit (gross margin basis).

Benchmark: 3:1 or higher is healthy; 5:1+ is excellent;
below 1.5:1 signals broken unit economics.

SaaS Benchmarks: What Investors Expect

Here are the key metrics and ranges investors use to evaluate SaaS companies:

Revenue Recognition: ARR vs. Recognized Revenue

This is where cash and P&L diverge in SaaS. A customer who pays £12,000 upfront for 12 months generates:

• Cash Flow Statement: +£12,000 inflow (when payment is received).
• Balance Sheet: +£12,000 deferred revenue liability (cash received but not yet earned).
• Income Statement: +£1,000 revenue recognized in Month 1, +£1,000 in Month 2, etc.

The timing mismatch is the source of confusion in early-stage SaaS models.

Building a Deferred Revenue Schedule

Deferred Revenue Schedule (Year 1):

Month    | New Contracts | Payment Terms    | Cash In      | Deferred Rev (BS) | Recognized Rev (P&L)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Jan 2025 | £400k         | 12-month annual  | £400,000     | £400,000          | £33,333
Feb 2025 | £400k         | 12-month annual  | £400,000     | £766,667          | £33,333
Mar 2025 | £450k         | 12-month annual  | £450,000     | £1,183,334        | £33,333
...
Dec 2025 | £650k         | 12-month annual  | £650,000     | £650,000          | £33,333
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Year 1 Total                                  | £5,900,000   | £3,716,667 (bal)  | £400,000 
                                             |              | (Deferred Rev     | (Jan contracts only,
                                             |              | declining across  | partial recognition)
                                             |              | months)

Each contract is a separate line. As it matures, the deferred revenue decreases and recognized revenue increases. The Year 1 "Recognized Revenue" line on the P&L is the sum of all monthly recognitions from all active and new contracts.

Proof of Concept: Cash vs. Revenue

In Year 1:

• Cash collected: £5,900,000
• Revenue recognized: ~£400,000 (only from the Jan cohort's 12-month contract, amortized as it's earned; new cohorts add progressively)
• Deferred revenue on balance sheet (end of Year 1): ~£3.7M

This is why early-stage SaaS companies look "cash-rich" but have low profitability on the P&L. The cash is a liability (deferred revenue), not profit, because the company must deliver the service over the next 11 months.

Cost Structure: S&M, R&D, and G&A as Percentages of Revenue

SaaS cost structure should be modelled as a percentage of recognized revenue (not cash or ARR):

Early-stage companies often spend more than they recognize in revenue because they are investing for future growth. As the company scales and deferred revenue recognizes in future periods, OpEx as a % of revenue declines.

Worked Example: Cost Structure for a Series B SaaS

Year 1 Recognized Revenue: £400k (conservative, from Jan cohort only) Proforma operating costs:

Year 1 Operating Expense Forecast:

Recognized Revenue (P&L)                         £400,000

Cost of Goods Sold (hosting, support)            (£80,000)  (20% of revenue)
Gross Profit                                      £320,000   (80% margin)

Operating Expenses:
  Sales & Marketing (70% of revenue)            (£280,000)
  Research & Development (20% of revenue)        (£80,000)
  General & Administrative (15% of revenue)      (£60,000)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Total OpEx                                      (£420,000)  (105% of revenue)
Operating Income (EBIT)                        (£100,000)  (-25% margin)

Note: Negative operating margin is expected in early SaaS growth.
This company is investing £500k in customer acquisition and product
to capture £5.9M in cash flow and £5M in ARR growth.

This is typical for a high-growth SaaS company: negative GAAP profitability but strong cash flow and unit economics.

Cash vs. P&L: Why Deferred Revenue Is Your Secret Weapon

The cash flow statement reconciles these divergences:

Year 1 Cash Flow Statement (SaaS):

Operating Activities:
  Net Income (P&L loss)                        (£100,000)
  Add back: Depreciation & amortization         £20,000
  Change in Deferred Revenue                  +£3,700,000  (KEY LINE: cash from
  Change in AR (negative, bad)                   (£50,000)   future revenue)
  Change in AP (positive, good)                 +£100,000
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Cash from Operations                           +£3,670,000

Investing Activities:
  CapEx (servers, equipment)                     (£200,000)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Free Cash Flow                                 +£3,470,000

Interpretation: The company burned £100k on the P&L but generated
£3.47M in free cash flow because customers paid upfront.
This is why SaaS companies can spend aggressively on growth
without immediate profitability.

The "Change in Deferred Revenue" line is the most important line in a SaaS cash flow. It transforms a loss into strong cash generation.

Sensitivity Analysis: What Happens If Churn Changes?

SaaS businesses are sensitive to churn. A 1% shift in monthly churn rate can destroy unit economics. Here's a sensitivity table for Year 2 ARR:

Even small changes in churn dramatically shift long-term ARR. This is why cohort-level retention analysis is essential: it allows you to forecast scenarios where churn improves (via product improvements) or deteriorates (due to competitive pressure).

Putting It Together: The Series B Worked Example

Let's model a Series B SaaS company with the following characteristics:

Assumptions:

• Starting ARR (Year 0): £5,000,000
• Average Contract Value (ACV): £50,000
• Target Year 1 customers: 100 new logos
• Gross margin: 80% (COGS is cloud hosting, ~20% of revenue)
• Monthly churn: 7% (annual retention: ~41%)
• Expansion rate: 10% of existing customer base per year
• Contraction: 2% of base
• S&M spend: £2,500,000 (blended CAC: £25,000)
• R&D spend: £500,000
• G&A spend: £300,000
• Tax rate: 0% (loss-making, no tax)

Year 1 Forecast:

━━━ ARR BUILD ━━━
Starting ARR                                    £5,000,000
New Logo ARR (100 × £50k)                       £5,000,000
Expansion (10% of £5M base)                       £500,000
Contraction (2% of £5M base)                     (£100,000)
Churn (7% monthly = ~41% annual of £5M base)    (£2,050,000)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Ending ARR (Year 1)                             £8,350,000
NRR: (£5M + £500k - £100k - £2.05M) / £5M = 70%

━━━ CASH FLOW (SIMPLIFIED) ━━━
Cash collected from customers (upfront annual)  £5,900,000
(New logos: £5M; expansion: £500k - contraction losses)
Cash paid for COGS (20% of collected)          (£1,180,000)
Cash paid for OpEx (S&M + R&D + G&A)           (£3,300,000)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Free Cash Flow (Year 1)                          £420,000

━━━ P&L (GAAP, REVENUE RECOGNIZED) ━━━
Recognized Revenue (amortized deferred rev)     £1,200,000
(Blended from all cohorts' partial-year coverage)
Cost of Goods Sold (20%)                         (£240,000)
Gross Profit                                      £960,000
S&M Expense                                    (£2,500,000)
R&D Expense                                      (£500,000)
G&A Expense                                      (£300,000)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Operating Income (EBIT)                       (£2,340,000)  (-195% margin)
Net Income (no interest, no tax)              (£2,340,000)

━━━ UNIT ECONOMICS ━━━
CAC (blended)                                    £25,000
CAC Payback (£25k ÷ (£50k × 80% ÷ 12 months)) 7.5 months
Magic Number ((£8.35M - £5M) ÷ £2.5M S&M)      1.34x
LTV (simplified: £50k × 80% ÷ 7% monthly churn) £571,429
LTV:CAC Ratio (£571k ÷ £25k)                    22.9:1
Gross Margin                                     80%
Rule of 40 (71% growth + (-195% margin))        NEGATIVE

Interpretation:
  • Strong unit economics (LTV:CAC 22.9:1, CAC payback 7.5 months)
  • High growth rate (71% ARR growth)
  • Negative Rule of 40 because the company is in heavy investment mode
  • NRR of 70% is concerning (expansion not offsetting churn; this is
    a red flag that suggests weak expansion or high contraction)

This Series B company is cash-flow positive and has strong unit economics, but negative GAAP profitability and weak NRR. This is a typical profile for a high-growth SaaS company; the question is whether NRR will improve as the product matures and expansion accelerates.

Common SaaS Modelling Mistakes

1. Blending all churn into a single rate: This hides cohort deterioration. Always model cohorts separately.
2. Forgetting deferred revenue: Forecasting only GAAP revenue without tracking deferred revenue and cash means you miss the company's liquidity and growth capital.
3. Confusing ARR with revenue: ARR is the annualized subscription base at a point in time; recognized revenue is the amortized amount earned. They are not the same.
4. Not accounting for expansion separately: Expansion (upsells) should be a distinct line item in the ARR roll-forward because it reflects product quality and customer satisfaction.
5. Including S&M salaries in CAC: CAC should typically include variable costs (commissions, ad spend, tools); fixed costs (salaries) should be in S&M OpEx. Be explicit about the definition.
6. Assuming linear churn: Churn is rarely linear. Most cohorts have steep early churn in months 1–3 (initial fit), then flatten. Use cohort curves, not averages.
7. Ignoring tax and interest: Early-stage SaaS models often assume 0% tax because the company is loss-making. But as the company scales, this changes. Also, if you have debt, interest expense compounds the profitability challenge.