Revenue Quality and Sustainability (3/5)

Quality of Earnings analysis typically starts with EBITDA, but the real work begins one line higher. Revenue is the foundation. If the revenue line is overstated, non-recurring, or economically hollow, no amount of cost normalization will produce a reliable earnings figure.

This article examines how to assess revenue quality in a transaction context. The question is deceptively simple: is the reported revenue repeatable, collectible, and economically real? Answering it requires looking beyond the top line.

What revenue quality means in practice

Revenue quality is about sustainability and economic substance. A business might report €50 million of revenue, but that number alone tells you nothing about whether the buyer can expect to collect €50 million next year.

Three dimensions define revenue quality:

Recurrence. How much of the revenue is contractually committed versus dependent on winning new business each year? A subscription business with 3-year contracts has fundamentally different visibility than a project-based business that starts each year at zero.

Collectibility. Revenue recognized in the accounts is not the same as cash collected. If a company books revenue but struggles to collect it, the economic value is lower than the headline suggests.

Economic substance. Does the revenue reflect genuine value delivered to customers, or does it include items that inflate the number without creating real economic activity? Related-party transactions, pass-through arrangements, and aggressive recognition policies can all distort the picture.

When I evaluate revenue quality, I am really asking: if I bought this business today, how confident am I that this revenue continues at this level and converts to cash?

The recurrence spectrum

Not all recurring revenue is created equal. Understanding where a business sits on the recurrence spectrum is one of the first things I assess.

At one end are pure subscription models with multi-year contracts. A SaaS company with 3 to 5-year agreements, invoiced annually in advance, has exceptional revenue visibility. If 94% of revenue comes from such contracts, the buyer knows almost exactly what next year looks like.

At the other end are purely transactional businesses. A distributor that takes orders daily has no contractual visibility. Every month starts from zero. The revenue might be stable historically, but there is no contractual lock-in.

Between these extremes sit most businesses. They have some contracted revenue, some repeat customers without contracts, and some genuinely new business. The analysis requires decomposing the revenue base to understand the mix.

Here is a framework I use:

For a buyer, contracted recurring revenue is valued at face value. New business carries more risk and may be discounted in the valuation or require earn-out protection.

Decomposing the revenue base

To assess recurrence properly, you need to decompose the revenue by customer, contract, and cohort. This analysis reveals patterns that the aggregate number hides.

Customer concentration

Start with the basics: what percentage of revenue comes from the top 5, top 10, and top 20 customers? High concentration creates risk. If one customer represents 25% of revenue, the business has a material single-point-of-failure.

But concentration alone does not tell the full story. A top customer on a 5-year contract with no termination rights is different from a top customer buying on spot terms. The analysis must combine concentration with contract quality.

In my experience, customer concentration becomes a significant due diligence finding when: the top customer exceeds 15-20% of revenue, there are termination provisions that could be exercised, or there are known relationship issues with key accounts.

Contract roll-off analysis

For contracted revenue, build a schedule showing when each material contract expires. This reveals whether the business faces a "cliff" where significant revenue is up for renewal simultaneously.

A business might have 90% recurring revenue, but if 40% of that renews in the next 12 months with uncertain outcomes, the forward visibility is weaker than it appears. The analysis should show: contract expiry dates, renewal history for each customer, any termination-for-convenience clauses, and current relationship status.

Cohort analysis

Cohort analysis tracks how customers acquired in each year behave over time. It answers questions like: what percentage of customers acquired in 2021 are still active? How does their spending change year-over-year?

This analysis reveals true churn rates and customer lifetime patterns. A business might claim 95% retention, but cohort analysis might show that the 5% lost each year are systematically the larger customers. The average masks a problematic pattern.

The calculation itself is straightforward but requires clean data. For each vintage year, track the number of customers and their revenue from acquisition through the present. The patterns that emerge are often the most important finding in the entire revenue analysis.

Churn and retention metrics

Churn is the silent killer of revenue quality. A business with 10% annual churn needs to replace that revenue just to stay flat. At a 30% gross margin, that 10% churn might require 30% more selling effort to maintain the same EBITDA.

Gross vs. net retention

There are two ways to measure retention, and the distinction matters:

Gross retention measures what percentage of prior-year revenue from existing customers remains, excluding any expansion. If you had €10 million from a customer last year and €9 million this year, gross retention is 90%.

Net retention (or net revenue retention, NRR) includes expansion. If you had €10 million from a customer last year, lost €1 million from churn, but gained €2 million from upselling, net retention is 110%.

High-quality subscription businesses often report net retention above 100%, meaning they grow without acquiring any new customers. This is the gold standard for revenue quality.

For the TS analysis, I always ask for both metrics. Strong net retention can mask high gross churn. A business with 80% gross retention and 110% net retention is heavily dependent on expansion sales to existing customers, which is a different risk profile than a business with 95% gross retention.

Churn by segment

Aggregate churn numbers can be misleading. Always segment the analysis by customer type, contract value, and product line.

I have seen businesses where overall churn looked acceptable at 8%, but enterprise customers (60% of revenue) had 3% churn while SMB customers (40% of revenue) had 15% churn. The blended number hid a significant quality gap in part of the business.

Similarly, churn patterns often differ by contract type. Customers on month-to-month terms churn faster than those on annual contracts. Multi-year contracts by definition cannot churn until expiry, so the true churn rate only becomes visible at renewal.

Revenue recognition issues

Revenue recognition determines when revenue appears in the accounts. The rules are set by accounting standards, but companies have latitude in how they apply them. This latitude creates room for aggressive or conservative positioning.

Common recognition timing issues

Cut-off adjustments. Revenue should be recognized when earned, but the definition of "earned" can be stretched. A company might ship goods on December 31 and recognize revenue, even though the customer has not accepted delivery. Cut-off testing verifies that revenue in the period relates to economic activity in that period.

Deferred revenue release. For subscription businesses, cash collected in advance is recorded as deferred revenue and released over the service period. If a company changes its release methodology, it can accelerate revenue recognition without any change in underlying performance.

Percentage of completion. For long-term contracts, revenue is often recognized based on progress toward completion. The estimates underlying this calculation are inherently judgmental. If management becomes more optimistic about project progress near year-end, revenue increases without any additional cash.

Bill-and-hold arrangements. Some companies recognize revenue when goods are ready for shipment even if the customer has not taken delivery. The accounting may be technically correct, but it pulls revenue forward from when cash will actually be collected.

When reviewing revenue recognition, I look for year-end spikes that suggest pull-forward, changes in recognition methodology, and significant estimates that could be manipulated.

The December deferral pattern

A pattern I have seen multiple times: management historically deferred a portion of December revenue to January to smooth the P&L, but in the sale year, they stopped deferring.

The result is a one-time bump in reported revenue that makes the reference year look stronger than trend.

This is often disclosed as a "change in accounting estimate" and may be technically defensible, but it creates a comparability problem.

The adjustment required is to normalize December recognition across all years to see the true trend.

Non-recurring revenue

Just as costs can be non-recurring, so can revenue. The TS analysis must identify revenue that appeared in the historical period but will not repeat.

One-time project revenue

Professional services businesses often have a mix of recurring support revenue and one-time implementation projects. A large implementation project can spike revenue in a given year without creating any ongoing benefit.

The question is: what is the normalized level of project revenue?

If a company typically wins €2-3 million of projects annually but won €5 million in the reference year due to one large engagement, the excess should be flagged.

Customer-specific revenue at risk

Sometimes revenue is recurring in form but at risk for specific reasons. A major customer might be up for contract renewal with a known competitive threat. Another might be going through financial distress. A third might have signaled dissatisfaction.

These situations do not always result in QoE adjustments, but they should be disclosed as risks. The buyer needs to understand what portion of revenue has identifiable threats to continuation.

Unusual items in "other revenue"

Just as with costs, the "other revenue" line often contains items that deserve scrutiny. Grants, insurance recoveries, litigation settlements, asset sale gains, and foreign exchange impacts can all flow through revenue lines depending on how the company structures its P&L.

I always request a breakdown of any revenue line that is not core operating revenue. The question is whether these items are recurring and operating in nature.

Revenue bridge analysis

One of the most useful tools for understanding revenue quality is the revenue bridge. This walks from prior year revenue to current year revenue, decomposing the change into its components.

A well-constructed revenue bridge shows:

• Prior year revenue

• Churn (customers lost)

• Contraction (existing customers spending less)

• Price increases/decreases

• Expansion (existing customers spending more)

• New business

• Current year revenue

This decomposition reveals the engine of growth. Is the business growing because it is winning new customers, or because it is expanding within existing accounts? Is growth coming from price or volume? Is churn being masked by new wins?

For example, consider two businesses both showing 10% revenue growth:

Business A: Lost 5% to churn, gained 15% from new business. Net growth 10%.

Business B: Lost 2% to churn, gained 5% from expansion, gained 7% from new business. Net growth 10%.

Both grew 10%, but Business B has higher-quality growth. Lower churn, meaningful expansion, and new business. Business A is running hard just to stay in place.

The revenue bridge also reveals pricing dynamics. If a business shows flat volume but 5% revenue growth, pricing power exists. If volume is up 8% but revenue up 5%, there is price erosion that may accelerate.

Working with limited data

In practice, you do not always get clean cohort data, perfect customer-level revenue, or multi-year contract schedules. The analysis must adapt to available information.

When data is limited, I rely on proxy measures:

DSO trends. Rising days sales outstanding can signal collection problems that indicate revenue quality issues. If revenue is growing but DSO is expanding faster, something is wrong.

Bad debt provisions. The level and trend of bad debt expense relative to revenue indicates collectibility. A spike in provisions suggests the company is booking revenue it cannot collect.

Deferred revenue movements. For subscription businesses, the relationship between cash collections and deferred revenue reveals recognition patterns. If revenue grows faster than deferred revenue and cash, recognition may be aggressive.

Customer count trends. Even without detailed cohort data, tracking total customer count over time reveals net acquisition. If customer count is flat but revenue is up significantly, either ARPU is expanding (good) or recognition is aggressive (concerning).

The goal is triangulation. No single metric tells the full story, but multiple metrics pointing in the same direction create confidence.

Revenue quality adjustments

Revenue quality issues can result in QoE adjustments, working capital adjustments, or risk disclosures depending on their nature.

QoE adjustments to revenue

Some revenue quality findings translate directly to EBITDA adjustments:

Non-recurring revenue removal. If €500k of revenue came from a one-time project that will not repeat, both revenue and the associated margin are removed from adjusted EBITDA.

Cut-off corrections. If revenue was pulled forward from the following period, it should be pushed back to present a normalized picture.

Related-party revenue at non-market terms. If intercompany revenue is priced above or below market, an adjustment normalizes to market terms.

Working capital implications

Some revenue issues affect working capital rather than EBITDA:

Accelerated recognition without cash. If revenue is recognized before cash is collected, receivables are inflated. This increases the working capital requirement.

Deferred revenue timing. If the company has changed how it recognizes deferred revenue, this affects both the P&L and the balance sheet. The working capital target must reflect a normalized deferred revenue balance.

Risk disclosures without adjustment

Not every finding becomes an adjustment. Some are disclosed as risks:

Customer concentration. A finding that 25% of revenue comes from one customer with a contract expiring in 18 months is a risk disclosure, not an adjustment.

Competitive threats. Knowledge that a major customer is tendering the contract is a risk factor that may affect valuation or deal structure, but it does not change historical EBITDA.

Churn trends. If churn is accelerating, this is relevant to the buyer's view of forward performance but does not typically adjust historical earnings.

What buyers and investment committees focus on

From an IC perspective, revenue quality drives several key decisions.

Valuation multiple

Businesses with high revenue quality command premium multiples. A SaaS company with 95% gross retention, 115% net retention, and 80% contracted recurring revenue might trade at 12-15x EBITDA. A project-based business with no contracted revenue and high customer concentration might trade at 6-8x.

The revenue quality analysis directly informs where on this spectrum the business belongs.

Debt capacity

Lenders care about revenue stability because it determines the reliability of debt service. Businesses with contracted recurring revenue can support higher leverage than businesses with volatile, project-based revenue. The revenue quality analysis feeds into the financing structure.

Deal protections

Revenue-related risks often translate into deal protections:

Customer-specific escrows. If a material contract is at risk, part of the purchase price might be held in escrow pending renewal.

Earn-outs. If revenue quality is uncertain, part of the consideration might be contingent on maintaining revenue levels post-close.

Representations and warranties. The buyer may require specific representations about customer relationships, contract status, and revenue recognition policies.

Integration priorities

Revenue quality findings inform post-close priorities. If churn is high, customer success investment becomes urgent. If concentration is the issue, diversification becomes a strategic priority. If recognition is aggressive, the buyer may need to reset revenue reporting to sustainable levels.

Illustrative example: a B2B services platform

To make this concrete, consider a B2B services platform with reported revenue of €25 million in the reference year.

The initial analysis reveals:

• 70% of revenue from subscription fees, 30% from professional services

• Top 5 customers represent 45% of revenue

• Contracts typically 12-24 months

• Management claims 92% revenue retention

Deeper analysis uncovers:

Customer concentration risk. The largest customer (18% of revenue) has a contract expiring in 6 months. Discussions with management reveal the relationship is under strain due to service issues.

True churn is higher than reported. Management's 92% retention figure includes expansion. Gross retention is actually 85%. The difference is being masked by successful upselling to remaining customers.

One-time project revenue. €1.2 million of professional services revenue came from a single implementation project that completed in the reference year. Normal annual project revenue is €600-800k.

Recognition timing. The company changed its methodology for recognizing setup fees from upfront to over the contract term. In the transition year, this created a one-time catch-up of €300k.

The findings translate to:

Adjusted EBITDA is €500k lower than reported. The valuation discussion also incorporates the customer concentration risk, potentially through an escrow on the at-risk revenue or a reduced multiple.

Closing thoughts

Revenue quality analysis requires patience and skepticism. The headline number is rarely the full story. The work involves decomposing revenue into its components, understanding the contractual foundation, testing recognition policies, and assessing the sustainability of each revenue stream.

The goal is not to find problems for their own sake. It is to understand what the buyer is actually acquiring. A business with €20 million of rock-solid recurring revenue may be worth more than a business with €30 million of volatile, concentrated, questionably-recognized revenue.

When I finish a revenue quality analysis, I want to be able to say: here is what this business will reliably generate, here are the risks to that number, and here is how confident we should be in the forward trajectory. That clarity is what enables informed decisions on price, structure, and post-close priorities.

This is Part 3 of a 5-part series on Quality of Earnings. Part 4 will cover cost structure analysis and Part 5 will cover how QoE findings shape deal outcomes.