SaaS Metrics for Management Accountants: ARR, NRR, CAC

Management accountants moving into SaaS discover quickly that the P&L is the least interesting page in the pack. The board reads ARR, net revenue retention, CAC payback and gross margin, and expects finance to own those numbers with the same rigour as the statutory ones. That is harder than it sounds. Unlike statutory reporting there is no standard: every SaaS business defines these metrics slightly differently, many define them inconsistently over time, and investors compare them against benchmarks calculated on a basis nobody states.

This guide covers what each metric measures, where the definitional traps are, a single worked example in which every number ties back to the others, how the metrics reconcile to revenue recognised under IFRS 15 and the revised FRS 102, the errors that most often damage a pack’s credibility, and what the skill set is worth in the hiring market. It is written for management accountants, FP&A analysts and the finance leaders who hire them.

ARR and MRR: simpler than they look, harder than they seem

Annual recurring revenue is the contracted, recurring revenue run rate at a point in time. It is not a forecast, not a twelve-month total, and not the same as revenue recognised under the accounting standards (our guide to revenue recognition under IFRS 15 covers the accounting side). Monthly recurring revenue is the same measure expressed monthly: MRR is ARR divided by twelve, and businesses with mainly monthly contracts tend to lead with MRR while those selling annual contracts lead with ARR.

Three questions decide whether your ARR is comparable to anyone else’s.

• What counts as recurring? Subscription fees clearly do. Usage-based revenue, professional services and one-off implementation fees generally do not, though businesses are frequently tempted to include them. If usage revenue is material, report it as a separate line rather than blending it in.
• What about contracts in notice? Some businesses remove churned customers at notification, others at contract end. The difference can move the number by several per cent, and it changes the timing of churn in the ARR bridge.
• How are non-annual contracts annualised? A monthly rolling contract at £500 is £6,000 of ARR on most definitions, but its churn profile is nothing like an annual contract of the same value. Multi-year contracts with stepped pricing raise the opposite question: is ARR today’s price or the average across the term? The defensible answer is the current contracted annual value.

Discounts and free months also need a rule. ARR should normally reflect what the customer is contracted to pay after discount; a three-month free period at the start of an annual contract is usually best shown at the full contracted rate with a note, but whichever rule you choose, apply it to every contract.

The management accountant’s job here is less calculation than consistency and disclosure: define it once, state the definition in the pack, and flag any change. A board that discovers the ARR basis changed quietly loses confidence in every other number. A useful discipline for anyone inheriting SaaS metrics is to spend the first month writing down exactly how each number is currently calculated and agreeing it with the CEO before changing anything. If a definition needs improving, improve it once, restate the comparatives and say so.

Net revenue retention: the metric that matters most

NRR measures what happened to a cohort of customers over twelve months: churn, downgrades and expansion combined, excluding new customers. A business at 100% grew existing accounts exactly enough to offset losses; figures above 110% are generally regarded as strong; below 90% means the business is filling a leaking bucket with new sales.

The distinction that gets confused: gross revenue retention (GRR) excludes expansion and therefore cannot exceed 100%; net retention includes it and can. Reporting one and calling it the other is one of the most common errors in SaaS packs, and investors notice immediately.

The formulas, on a cohort basis:

• NRR = (opening ARR + expansion − contraction − churn) ÷ opening ARR, all measured on customers who existed at the start of the period.
• GRR = (opening ARR − contraction − churn) ÷ opening ARR.

The practical difficulty is cohort discipline. NRR requires tracking the same set of customers across a defined period, which means the customer master data has to be reliable: accounts merged, subsidiaries grouped consistently, contract changes tracked. That data usually lives in the CRM and billing system rather than the ledger, and getting it right is frequently the hardest part of producing the number, as our guide to data quality and the finance function covers.

Churn: logo churn and revenue churn

Churn comes in two forms and they tell different stories. Logo churn is the number of customers lost in the period divided by the number at the start. Gross revenue churn is the ARR lost from cancellations (and, on some definitions, downgrades) divided by opening ARR. A business losing many small customers and keeping its large ones will show high logo churn and low revenue churn, which is a healthy pattern for a business moving upmarket and a worrying one for a business that depends on volume.

Two practical points. First, state whether contraction sits inside revenue churn or is reported separately; the worked example below reports it separately, which makes GRR equal to one minus churn minus contraction. Second, calculate annual churn from the cohort data rather than multiplying a monthly rate by twelve.

CAC and CAC payback

Customer acquisition cost is total sales and marketing spend in a period divided by new customers acquired. CAC payback, the more useful figure, is CAC divided by the gross margin the average new customer generates each month, giving the number of months needed to recover the acquisition cost.

Two definitional traps. What goes into the numerator? Fully loaded sales and marketing including salaries, commission, tooling and an overhead allocation, or just campaign spend? The first is honest and produces an uncomfortable number; the second is common and flatters. And is payback measured on gross margin or revenue? Gross margin is correct and considerably longer, which is why revenue-based payback appears in a lot of investor decks.

A third trap has grown with IFRS 15: if sales commission is capitalised as a cost of obtaining a contract and amortised, the P&L sales and marketing line understates the cost incurred in the period. CAC should use commission incurred, not commission amortised. Under FRS 102 the treatment of contract costs has also been revisited in the periodic review, so confirm your own policy with the financial accountant before you build the calculation.

A payback under twelve months is generally strong for a mid-market SaaS business; beyond twenty-four months the business is effectively funding growth from capital rather than from operations. That is a strategic fact rather than an accounting one, and it belongs in front of the board.

LTV and the LTV:CAC ratio

Lifetime value estimates the gross margin a customer will generate before they leave. The usual simplified formula is average ARR per customer × gross margin % ÷ annual revenue churn rate. Divide LTV by CAC and you have the ratio investors quote; around 3:1 is the commonly cited marker of efficient acquisition.

LTV is the most fragile metric in the set because it divides by churn. A small change in the churn assumption swings the result dramatically, and young businesses with little history often use churn rates that their data cannot support. Show the churn rate you used, show the sensitivity, and prefer GRR-based churn (cancellations plus downgrades) to cancellations alone. The worked example below shows how much that one choice changes the answer.

Burn multiple and the Rule of 40

Burn multiple is net cash burn divided by net new ARR: how many pounds the business consumes to add one pound of recurring revenue. Commonly quoted thresholds treat a multiple below 1x as very efficient, 1x to 2x as acceptable for a growing business, and above 2x as a concern. Net burn should be the actual cash outflow, so it includes capitalised development spend and reflects the working capital benefit of billing annually in advance.

The Rule of 40 adds the growth rate to the profit margin; a combined figure of 40 or more is the widely used marker of a healthy balance between growth and profitability. The rule says nothing about which growth rate or which margin, so state both. Most businesses use ARR or revenue growth with EBITDA margin; investors focused on cash prefer free cash flow margin. The two can differ by several points, as the example shows.

Gross margin, properly calculated

This is the metric SaaS businesses most often get wrong in their own favour. Cost of sales should include hosting and infrastructure, customer support and success, third-party software embedded in the product, the people who deliver implementation and professional services, and the amortisation of capitalised development where it relates to delivering the service. What it should not include is sales, marketing or general R&D.

Businesses that exclude customer success from cost of sales can report gross margins in the eighties that ought to be in the low seventies. That matters because gross margin drives the CAC payback calculation, the LTV estimate and the valuation multiple, so an optimistic margin propagates through everything else in the pack.

The frameworks themselves are set by the IASB for IFRS and by the Financial Reporting Council for UK GAAP, and development cost capitalisation is one of the more judgement-heavy areas: under IFRS (IAS 38) development costs must be capitalised once the criteria are met, while FRS 102 gives a policy choice. That judgement sits with the financial accountant as much as with management accounting: what qualifies, over what life, and whether the amortisation lands above or below the gross margin line. Our comparison of IFRS and UK GAAP in practice covers the wider differences.

A worked example that adds up

The figures below are illustrative, for a fictional UK B2B SaaS business, and are built so that every metric ties back to the same underlying numbers. The point is not the benchmarks; it is to show how the metrics connect to each other and to the P&L.

The ARR bridge

ARR bridge for the year	£000	Customers
Opening ARR	4,000	400
New business	1,200	80
Expansion (upsell, price rises)	800	–
Contraction (downgrades)	(200)	–
Churn (cancellations)	(400)	(48)
Closing ARR	5,400	432

From the bridge: closing MRR is £5.4m ÷ 12 = £450k. Net new ARR is £1.4m, so ARR growth is 35%. NRR is (4,000 + 800 − 200 − 400) ÷ 4,000 = 105%. GRR is (4,000 − 200 − 400) ÷ 4,000 = 85%. Gross revenue churn is 400 ÷ 4,000 = 10%; including contraction it is 15%, which is simply 100% minus GRR. Logo churn is 48 ÷ 400 = 12%, higher than revenue churn, which tells you the business is losing smaller customers.

The P&L behind it

Income statement for the year	£000
Subscription revenue (recognised over time)	4,650
Implementation and professional services	350
Revenue	5,000
Hosting and infrastructure	(400)
Customer support and success	(550)
Embedded third-party software	(150)
Services delivery staff	(250)
Amortisation of capitalised development	(150)
Gross profit (70%)	3,500
Sales and marketing (fully loaded)	(2,000)
Research and development (expensed)	(1,800)
General and administrative	(1,200)
Operating loss	(1,500)

Gross margin is 3,500 ÷ 5,000 = 70%. If customer success were (wrongly) moved below the line, cost of sales would fall to £950k and gross margin would rise to 81%, a flattering eleven points that would then flow into every metric below.

Unit economics

• CAC = £2,000k ÷ 80 new customers = £25,000.
• New customer ARR = £1,200k ÷ 80 = £15,000, or £1,250 a month. At a 70% gross margin that is £875 of monthly gross profit.
• CAC payback on gross margin = £25,000 ÷ £875 = about 29 months. On revenue it would be £25,000 ÷ £1,250 = 20 months, which is the version that tends to appear in decks.
• Blended payback recognises that sales and marketing also drive expansion: £2,000k ÷ ((£1,200k + £800k) × 70%) × 12 = about 17 months. Both figures are valid if labelled; what is not valid is presenting the blended figure as new-logo payback.
• LTV using churn only (10%) = £15,000 × 70% ÷ 10% = £105,000, giving LTV:CAC of 4.2:1. Using GRR-based churn (15%) = £10,500 ÷ 15% = £70,000, giving 2.8:1. The same business moves from comfortably above the 3:1 marker to below it on one definitional choice.

Burn multiple and Rule of 40

Net cash burn starts from the operating loss of £1,500k, adds back £150k of non-cash amortisation, deducts £600k of development spend capitalised in the year (cash out, but not in the P&L), and adds a £300k increase in deferred income from annual billing in advance: 1,500 − 150 + 600 − 300 = £1,650k. Burn multiple = £1,650k ÷ £1,400k net new ARR = about 1.2x.

EBITDA is the operating loss plus amortisation: −£1,350k, a margin of −27%. Rule of 40 on ARR growth and EBITDA margin = 35 − 27 = 8. On free cash flow margin (−£1,650k ÷ £5,000k = −33%) it is 35 − 33 = 2. Either way the business is well short of 40, which is a growth-versus-burn conversation for the board rather than a reporting issue.

Reconciling the metrics to the P&L and to IFRS 15 / FRS 102 revenue

ARR is not revenue, but the relationship between them should be explicable, and finance functions that cannot bridge from one to the other lose credibility on both. In the example, closing ARR is £5.4m while subscription revenue recognised is £4.65m. The difference is timing: ARR is a closing run rate, revenue is earned through the year as ARR grew from £4.0m to £5.4m. The average of opening and closing ARR is £4.7m; the remaining £50k reflects new business weighted towards the second half. A short reconciliation of this kind belongs in the pack every quarter.

Under IFRS 15, and under the revised FRS 102 Section 23, which introduced a comparable five-step model for accounting periods beginning on or after 1 January 2026 (see the ICAEW’s summary of the revenue changes), the points that most affect a SaaS reconciliation are:

• Subscription access is usually satisfied over time, so revenue is recognised evenly across the service period and amounts billed in advance sit in deferred income (a contract liability).
• Implementation fees are often not a distinct performance obligation. Where set-up does not transfer anything to the customer on its own, the fee is recognised over the related service period rather than when the work is done.
• Variable consideration and contract modifications (usage tiers, mid-term upgrades, credits) need consistent rules that match the ARR bridge’s expansion and contraction lines.
• Contract costs, principally sales commission, may be capitalised and amortised, which separates the P&L sales and marketing figure from the cash cost used in CAC.

For UK groups reporting under FRS 102, the transition to the new revenue model is a good moment to align the ARR definitions with the performance obligations the auditors have agreed. A short technical paper, along the lines of our guide to writing technical accounting memos, documents the judgements once for both purposes.

Common errors in SaaS metric reporting

• Calling GRR “retention” without saying which, or presenting NRR as if it were GRR.
• Including services, usage or one-off fees in ARR, which inflates growth and makes the ARR-to-revenue reconciliation impossible.
• Changing a definition without restating comparatives, so that growth reflects the new basis rather than the business.
• Excluding customer success or embedded software from cost of sales, inflating gross margin and every metric derived from it.
• LTV built on a churn rate the data cannot support, typically from a period too short or a cohort too small.
• Burn multiple calculated on the operating loss rather than cash, ignoring capitalised development and deferred income movements.
• Customer counts that do not match the CRM because subsidiaries, resellers and merged accounts are grouped differently in billing and in sales reporting.

Putting them in the pack and reporting to investors

Five practical points for building a SaaS management pack that works.

State every definition, once, in an appendix. It removes an entire category of board argument and it signals rigour.

Report movement, not just position. An ARR bridge (opening, new, expansion, contraction, churn, closing) tells the board considerably more than the closing number, and it is the single most useful slide in most SaaS packs.

Reconcile the metrics to the statutory numbers. Show the ARR-to-revenue bridge and the operating-loss-to-cash-burn bridge each quarter.

Keep the backward and forward halves balanced. The metrics above describe what has happened; the board also needs the forecast and the scenarios, as our guide to board reporting: management accounts versus FP&A sets out. Driver-based planning works particularly well in SaaS because the drivers are the same metrics: new ARR, churn, expansion and headcount.

Give investors the same numbers the board sees. Venture and growth investors will ask for the metrics in their own templates, often monthly. Build one reconciled data set and map it to each investor’s template rather than recalculating for each. Our guide to building an investor-ready reporting pack covers structure, cadence and covenant information, and investor reporting for finance teams covers the working relationship. Expect due diligence to rebuild your ARR from the billing data; if your definitions are written down and your bridges reconcile, that exercise confirms your numbers rather than replacing them.

What this means for hiring

Two implications. SaaS management accounting is a genuine specialism, and a candidate who has produced an ARR bridge and owned an NRR calculation is materially more useful than one who has read about them. Test it directly: ask how they define ARR, what they include in cost of sales, how they reconcile ARR to revenue, and what happened when a definition changed. Our management accountant interview questions cover the wider assessment. Qualification through CIMA, ACCA or ICAEW is the baseline; the sector experience is the differentiator.

And the data dependency is the real constraint. The metrics live in the billing system and the CRM as much as the ledger, so data capability (SQL, a BI tool, comfort with billing platforms) matters more here than in most management accounting roles. As the business grows, the metrics usually move under a Financial Controller or Head of FP&A; our guides to a fractional Financial Controller for SaaS and to hiring a Financial Controller for a technology business cover the level above.

Management accountant and FP&A pay in SaaS businesses

Our house salary benchmarks apply to SaaS roles as they do elsewhere; sector experience tends to decide where a candidate sits within a range rather than moving the range itself. London figures:

Role	London salary	Interim day rate
Management Accountant	£50k–£85k (mid-level £58k–£73k)	£250–£375 London / £225–£325 elsewhere
FP&A Analyst	£48k–£62k	–
FP&A Manager	£62k–£82k	–
Head of FP&A	£80k–£108k	–
Finance Business Partner	£52k–£62k / £62k–£80k / £78k–£100k by seniority	–

Early-stage SaaS businesses often start with a fractional management accountant before a full-time hire. There is no house salary figure for fractional work; as an indicative guide only, one day a week at interim management accountant rates is roughly £11,500–£17,000 a year in London (about 46 working weeks at £250–£375 a day) and £10,000–£15,000 elsewhere. For fuller detail see the management accountant salary guide and the FP&A manager salary guide. Where cover is needed quickly, for a funding round or a reporting backlog, interim management accountant recruitment typically produces candidates within days.

Frequently asked questions

What is the difference between ARR and revenue?

ARR is a point-in-time run rate of contracted recurring revenue. Revenue is what is earned over a period under IFRS 15 or FRS 102. A growing business will usually have closing ARR above the year’s subscription revenue, and the difference should be reconcilable.

Should NRR or GRR go in the board pack?

Both. NRR shows whether existing customers grow; GRR shows how much revenue leaks before expansion. Label them clearly, because confusing the two is one of the errors investors spot first.

Is LTV:CAC still useful?

It is useful as a summary if the inputs are disclosed. Because it divides by churn, show the churn rate used and a sensitivity; for young businesses, CAC payback is often the more reliable measure.

Does the FRS 102 periodic review change SaaS revenue recognition?

For periods beginning on or after 1 January 2026, FRS 102 Section 23 uses a five-step model comparable to IFRS 15. Many SaaS businesses already recognised subscription revenue over time, but implementation fees, contract modifications and multi-element contracts may need fresh judgements, which should be agreed with the auditors and documented.