Software LBOs: The Metrics Private Equity Actually Underwrites
Michael King, PE Investment Manager · 10 min read ·
- Software is underwritten on the durability of recurring revenue, not current profit — a business that is loss-making on a reported basis can still clear 6-8x revenue if that revenue renews
- Net revenue retention is the number that carries the multiple: a median private SaaS business sits around 106%, best-in-class above 120%, and anything below 100% means the installed base is shrinking before a single new customer is won
- The Rule of 40 is the screen, not the average — the median software company scores roughly 25 and misses it; the top quartile clears about 43
- The deferred revenue haircut still bites under IFRS 3, even though US GAAP largely removed it for deals closing after December 2022 — a live distinction for a London-based buyout
Software Is Priced on Revenue Because the Revenue Renews
Private equity evaluates a software company on the durability of its recurring revenue, not on its current profit. A business that loses money on a reported basis can still command 6-8x its revenue — a multiple no industrial would ever see — because a signed subscription that renews at 110% each year is closer to an annuity than to a sales pipeline. The whole exercise is an argument about how much of next year's revenue is already contracted, and how much of it survives.
This is why the standard LBO screen breaks down on software. An industrials buyout is underwritten on EBITDA, a multiple, and a debt schedule that delevers off free cash flow. A software buyout of a high-growth, sub-scale asset may have almost no EBITDA to lever — the margin is being spent on sales and marketing to acquire customers whose lifetime value lands in years three to five. Thoma Bravo, Vista and, in London, Hg built their franchises on exactly this asset: predictable, contracted, high-gross-margin revenue that supports leverage and multiples the reported profit alone cannot justify. Technology has run at roughly a third of global buyout deal value in recent years, so this is not a niche — it is the centre of gravity of the industry the audience is trying to enter.
If the revenue is the asset, the first question is how much of it stays.
Net Revenue Retention: The Number That Carries the Multiple
Net revenue retention (NRR) measures how much revenue a cohort of existing customers generates this year versus last, after churn, downgrades, and — critically — upsell and price increases. It is expressed as a percentage of the starting base and excludes any revenue from new logos. An NRR of 120% means the installed base grew a fifth on its own, before the sales team won a single new account. An NRR of 95% means the company is running down an escalator: it has to sell hard just to stand still.
The reason NRR dominates a software diligence is that it compounds. A business with 120% NRR doubles its contracted base roughly every four years with zero new customers; a business at 100% is entirely dependent on new-logo acquisition, which is expensive and cyclical. Two companies with identical revenue today and a 20-point gap in NRR are not the same asset, and a sponsor will pay a materially higher entry multiple for the first. It is the single line most likely to move the price.
Retention tells you the base holds. The next question is whether the company is allowed to keep spending to grow it.
The Rule of 40 Is a Screen, and Most Software Fails It
The Rule of 40 states that a healthy software business's revenue growth rate plus its profit margin — usually free cash flow or EBITDA margin — should sum to at least 40. A company growing 30% at a 10% margin scores 40 and passes; one growing 15% at a 5% margin scores 20 and does not. The logic is that the market will forgive burn in exchange for growth, and forgive slow growth in exchange for cash, but not tolerate the absence of both.
The instructive fact is that most software misses the Rule of 40. Treating 40 as the average, rather than the threshold that separates the top quartile from the pack, is the tell of someone who has read the metric on a blog rather than run it across a portfolio. For a sponsor, the score is a triage tool: a company well above 40 has optionality — it can throttle growth to harvest cash, or spend cash to buy growth. A company at 20 has neither lever and is one bad quarter from a covenant conversation.
Assume the metrics support the thesis. The moment the deal closes, the accounting introduces a problem that has nothing to do with the business.
The Deferred Revenue Haircut: Where the Revenue Vanishes
Software is sold in advance. A customer pays for an annual contract on day one, and the vendor recognises that cash as revenue over the twelve months it delivers the service; the unearned portion sits on the balance sheet as deferred revenue, a liability. In an acquisition, that liability has historically been written down to its fair value — defined as the cost of fulfilling the remaining obligation plus a normal margin, not the price the customer actually paid. For a software business with near-zero marginal cost of delivery, that fair value is a fraction of the booked amount, so the acquired deferred revenue is slashed. The revenue attached to it is never recognised by the buyer. It simply disappears.
The effect is that a freshly acquired software company reports a revenue air-pocket in its first post-close year — GAAP revenue drops even though the underlying subscriptions are unchanged and cash is still arriving. A candidate who models the target's stated revenue straight into the buyer's year-one accounts, with no haircut, has made a real error, and it is a favourite trap in a software-focused case study.
The accounting distortion is temporary. The reason software supports the leverage it does is not.
Why Leverage Works Off ARR, Not Just EBITDA
A conventional LBO lends against EBITDA, typically 4-6x for an industrial. A high-growth software business that is deliberately loss-making has little EBITDA to lend against, yet these deals still carry meaningful debt. The resolution is that lenders underwrite the recurring revenue directly. An annual recurring revenue (ARR) facility sizes the loan as a multiple of contracted ARR rather than of profit, on the logic that a book of renewing subscriptions with high gross retention is more predictable collateral than the cyclical earnings of a manufacturer.
That predictability is the entire case. A subscription book retaining north of 90% gross is a more bankable cash stream than an industrial earnings line that swings with the cycle, which is why software clears both higher entry multiples and, structured through ARR debt, higher effective leverage than its current profitability would ever allow. The covenant then flips from a leverage ratio to an ARR-and-retention test: breach the retention floor and the facility resizes, because the collateral has shrunk. It is the same instinct as the metric set — price and lend against the durability of the revenue, not the accident of this year's margin.
What an Interviewer Is Actually Testing
A software case study or technical question is checking whether a candidate understands why the sector breaks the standard playbook. The reflex to reach for an EV/EBITDA multiple on a business with no EBITDA, or to model stated revenue through a close with no deferred-revenue adjustment, signals someone applying an industrials template to an asset that does not fit it. The stronger answer connects the metrics to the thesis: retention justifies the multiple, the Rule of 40 tests whether the growth-margin trade is real, ARR supports the debt, and the deferred revenue haircut is an accounting artefact — sharper under IFRS — not a deterioration in the business.
The same discipline that reads a software deal reads any deal: separate the durable economics from the accounting presentation, and price the part that renews. That is the habit a sponsor is hiring for, and it is visible in a candidate's first answer or absent from it.
Take Your Preparation Further
Software sits on top of the mechanics every buyout shares, so read it alongside the pieces that build the structure underneath. Start with the beginner's LBO model guide for the entry-leverage-exit framework a software deal still runs on, and quality of earnings and EBITDA add-backs for how a diligence team pressure-tests the profit number — the exact discipline software forces onto revenue. For the accounting mechanics behind the haircut, see purchase price allocation and goodwill; for where retention and the multiple actually turn into a return, the value-creation bridge and what makes a good LBO candidate.
For a model you can pick apart line by line, download the LBO Model Template, and for the valuation methods that set the multiples on either side of the hold, the free Valuation Methods Cheat Sheet.
Ready for personalised feedback? Book a 1-on-1 mentoring session with an experienced IB/PE professional.
Frequently asked questions
Why do software companies trade at higher multiples than other businesses?
Because their revenue renews. A subscription book with high gross retention and net revenue retention above 100% behaves like an annuity — most of next year's revenue is already contracted before the sales team wins a single new customer. That predictability, combined with high gross margins and low incremental cost of delivery, is what supports enterprise-value-to-revenue multiples of 6-8x or more, far above what a cyclical industrial earnings line would command. The market is paying for durability of revenue, not for this year's reported profit.
What is net revenue retention and what is a good number?
Net revenue retention (NRR) measures how much revenue a cohort of existing customers produces this year versus last, after churn and downgrades but including upsell and price increases, expressed as a percentage of the starting base and excluding new logos. The median for private SaaS in 2025 is roughly 106%; best-in-class businesses clear 120%, and larger companies above $100M of ARR tend to sit around 115%. Anything below 100% means the installed base is shrinking on its own, which forces the company to rely entirely on expensive new-customer acquisition just to grow.
What is the Rule of 40 in software?
The Rule of 40 states that a software company's revenue growth rate plus its profit margin — usually free cash flow or EBITDA margin — should sum to at least 40. The idea is that investors will forgive cash burn in exchange for fast growth, and forgive slow growth in exchange for strong cash generation, but not the absence of both. It is a threshold, not an average: the median private software company in 2025 scores about 25 and misses it, while the top quartile clears roughly 43. Sponsors use it as a triage screen rather than a valuation input.
What is the deferred revenue haircut in an acquisition?
Software is paid for in advance, so unearned amounts sit on the balance sheet as deferred revenue. Historically, when a company was acquired, that liability was written down to fair value — the cost of fulfilling the remaining obligation plus a normal margin — which for a low-marginal-cost software business is far below the amount the customer actually paid. The revenue attached to the written-down portion is never recognised by the buyer, creating a year-one revenue air-pocket even though the underlying subscriptions are unchanged. US GAAP largely removed this under ASU 2021-08 for deals closing after December 2022, but IFRS 3 still requires fair value, so the haircut remains live for most UK and European targets.
How is leverage on a software LBO different?
A conventional LBO lends against EBITDA, but a high-growth software business is often deliberately loss-making and has little EBITDA to lever. Lenders instead underwrite the recurring revenue directly through an ARR facility, sizing the loan as a multiple of contracted annual recurring revenue on the logic that a renewing subscription book with high gross retention is more predictable collateral than cyclical industrial earnings. The covenants shift accordingly — from a leverage ratio to ARR and retention tests — because if retention falls, the collateral shrinks and the facility resizes.