Todd Gardner

The argument: value and the ability to raise capital depend on whether the books hold together, not on advanced reporting.

• An AI product's cloud costs must be split between production support and R&D; overstated cost of goods sold lowers valuation and the odds of raising capital.
• A separate MRR report is acceptable only if it reconciles to revenue.

Two portfolio companies fell behind on cash reconciliations; the financials were restated, the CFOs were fired, and both companies failed in part because confidence in the accounting had collapsed.

• Another client deferred revenue on some contracts but not others.
• Three term sheets arrived, but every deal fell apart as diligence stalled and buyers tried to decipher the profit and loss statement.

Presented as a 2026 update, the list reads as a recurring diagnostic rather than a one-off observation.

Read original

Their reasoning runs from GAAP treatment of costs required to satisfy the promised service to the distortion caused by growth.

• These costs behave more like sales commissions than ongoing hosting; in modern SaaS, much of the expense comes from infrastructure such as initial data loads, dedicated environments, and pre-purchased cloud capacity.
• If left uncapitalized, rapid customer growth pushes reported gross margin below its long-term average and distorts unit economics such as the CAC ratio and the ratio of lifetime value to CAC.
• The approach has two caveats: present a cash flow statement alongside the profit and loss statement, because this treatment understates cash burn, and use it only when growth is depressing margins—not to temporarily improve the numbers of a low-growth business.

A young, growing SaaS business with negative gross margin caused by up-front implementation costs cannot raise capital until it fixes the accounting—even though Gardner says the negative margin is a timing difference masking strong unit economics.

The tension is explicit: Gardner generally opposes capitalizing software development but favors capitalizing implementation and onboarding in this case.

Read original

The buyer will run a discounted cash flow exercise, so the seller should run it first.

• The framework: 36 months of a SaaS profit and loss statement, with months across columns; a terminal value based on the run-rate operating profit in month 36, not the full third year; conservative retention; and little or no assumed new sales.
• Gardner suggests a minimum 20% discount rate for businesses above $10 million in ARR and 25% below, arguing that the rate should reflect the risk in the assumptions rather than the overall risk of the business.
• Retention has the greatest effect on the outcome; gross margin also matters in a revenue multiple.

Challenge a buyer who assumes 15% churn when the historical average is 5%, and ask why they chose that discount rate. Gardner says the ARR multiple ranges were tested with several buyers and proved, on average, surprisingly predictive.

Read original

The Silicon Valley Bank comparison makes the case: its loans more than covered deposits, but withdrawals came faster than the portfolio could be liquidated, and the bank failed anyway.

• Expected losses come from two sources: lending too much relative to ARR, such as debt above roughly one times ARR, and specific companies facing an immediate AI threat.
• The base case is that most SaaS customers are unlikely to churn soon and a shrinking SaaS company can still service its debt even when its equity is worth little.
• Exposure also depends on the instrument: stock in a lender can become worthless even with sound credit, while a traditional fund holder is backed by the loan assets.

Gardner describes funding two versions of SaaS Capital.

• The first had an asset-liability mismatch and failed during the 2008 financial crisis despite having sound loans.
• The next was built on funds investors cannot withdraw, which Gardner says protects everyone from the panic that causes the most harm.

The post expresses confidence in the credit while identifying pockets where AI disruption poses an immediate risk. Gardner says they are not smart enough to identify every such case.

Read original

Free code has existed for decades, yet open source represents only about 12% of the application software market by revenue because proprietary vendors provide implementation, training, support, hosting, stability, and accountability.

• An AI-native startup can replicate Salesforce's code cheaply but still has to build the rest of the company.
• Salesforce's existing code base is a sunk cost, not a moat; future spending is what matters.
• Gardner separates three threats: AI-built replacements, direct obsolescence of analytical tools where AI tools already do things better, and the broader possibility that current software becomes irrelevant as the enterprise turns into a centralized data lake that agents read from and write to.

Finance professionals told Gardner that Excel plus Claude is better than any FP&A platform on the market. They repeat the claim without endorsing it, saying they do not know if it is true, and use it to illustrate the risk of immediate obsolescence.

Read original

It answers the practical question a finance lead or founder asks first: what specifically goes wrong in SaaS books, and what does it cost?

Gardner writes that they have reviewed the financial statements of over a thousand SaaS companies and lists the five problems they see most: poorly defined gross margin, weak bank reconciliations, income statements that do not connect to the balance sheet and cash flow statement, a mix of cash and accrual accounting, and MRR reporting that stands apart from the income statement.

• Each problem comes with a consequence.
• Two portfolio companies fell behind on reconciling cash; the financials were restated, the CFOs were fired, and both companies failed in part because confidence in the accounting was gone.
• Another client deferred revenue on some contracts but not others; three term sheets arrived, and every deal fell apart as diligence stalled while buyers tried to decipher the profit and loss statement.
• On gross margin, Gardner argues that an AI product's cloud costs must be split between what directly supports production and what belongs in R&D, because overstated cost of goods sold lowers a company's valuation and its odds of raising capital.
• They also note that the cash flow statement is the most underused report in SaaS, and that a separate MRR report is acceptable only if it reconciles to revenue.

For brands planning a campaign

For a financial reporting and valuation-readiness brief, open with this diagnostic list. A workshop can audit the five areas, then tackle the two fixes with the clearest payoff: allocating costs between COGS and R&D, and reconciling MRR to the income statement.

Gardner opens by calling themself a person who hates capitalizing costs, then argues for doing it in one specific case: implementation and onboarding.

• GAAP defines the costs necessary to satisfy the promised service; if that work is required to get a customer running, it belongs in COGS, even though it behaves more like a sales commission than a hosting bill.
• In modern SaaS, a large share of that cost is infrastructure, such as initial data loads, provisioning dedicated environments, and pre-purchased cloud capacity.
• Capitalizing those costs gives a truer picture of performance.
• Without it, rapid customer growth pushes gross margins below their long-term average and distorts unit economics like the CAC ratio and the ratio of lifetime value to CAC.
• The concrete case is a young, growing SaaS business with a negative gross margin from up-front implementation costs that cannot raise capital unless the accounting is fixed.
• Gardner adds two caveats: present a cash flow statement alongside the P&L, because this treatment understates cash burn, and use it only when growth is depressing margins, not as a temporary improvement for a low-growth business.

For brands planning a campaign

This is the accounting decision most likely to come up in a valuation-readiness engagement, and the caveats make it defensible. Present it as a change that improves accuracy, not one that flatters the numbers, and pair it with the cash flow statement so a board or lender sees the full picture.

Gardner reports that 35% of SaaS companies shrank in 2025, based on data from more than 2,000 companies.

• The post separates what is new from what is not.
• New: AI is depressing prices and capital for mid-sized SaaS buyers, even where disruption is years away.
• Not new: buyers value retention above everything.
• They can be wrong about cost savings, integrations, and cross-selling and still close a good deal, but overestimating retention kills the acquisition.
• Gardner's framework is a discounted cash flow model in Excel covering 36 months of the SaaS P&L, with months across columns, a terminal value based on month 36's run-rate operating profit rather than the full third year, conservative retention, and little or no new sales.
• They suggest a minimum 20% discount rate above $10 million in ARR and 25% below it, and argue the rate should reflect the risk in the assumptions, not the overall risk of the business.
• They advise challenging a buyer's assumed churn rate when it contradicts the historical average, and asking why the discount rate is what it is.
• Gardner says the ARR multiple ranges were tested with several buyers and were surprisingly predictive.

For brands planning a campaign

This is a ready-made exercise for a modeling workshop: each participant builds the 36-month model, then reruns it with the buyer's retention and discount-rate assumptions to see how far the valuation moves. It also shows Gardner is willing to challenge a buyer's numbers.

Gardner writes that they helped start the private credit push into SaaS, and argues the current problem is not credit quality but a loss of confidence combined with an asset-liability mismatch.

• They compare it with Silicon Valley Bank: the loans covered the deposits, but withdrawals were faster than the portfolio could be liquidated, and the bank failed anyway.
• The post distinguishes owning stock in a lender, where value can be wiped out even with sound credit, from investing in a traditional fund, where the loans themselves back the position.
• Gardner expects losses to come from two places: lending too much relative to ARR, such as debt above one times ARR, and specific companies facing an immediate AI threat.
• The counterweight is that most SaaS customers are unlikely to churn soon, there is no evidence yet of that, and a shrinking SaaS company can still service its debt even when its equity is worth little.
• They also describe funding two versions of SaaS Capital: one with an asset-liability mismatch failed in the 2008 crisis despite sound loans, and the next was built on funds investors cannot withdraw.

For brands planning a campaign

Useful context for a company raising or refinancing: it explains who is lending against ARR, what terms lenders watch, and why retention matters to a lender as much as to an equity buyer. The distinction between lender stock and fund exposure also clearly explains the difference for an investor audience.

Gardner sets up the claim that customers or startups can now replicate a vendor's code cheaply and undercut it, then compares this with open source.

• Free code has existed for years, yet open source represents only around 12% of the application software market by revenue because proprietary vendors supply what free projects cannot: implementation, training, support, hosting, stability, and accountability.
• An AI-native startup can copy Salesforce's code cheaply but still has to build the rest of the company.
• Salesforce's existing code base is a sunk cost rather than a moat, so future spending is what matters.
• Gardner then names two different threats: direct obsolescence, where generic tools already beat specialist software, illustrated by finance professionals claiming that Excel plus Claude outperforms any FP&A platform—a claim Gardner repeats without endorsing—and software irrelevance, where the enterprise becomes a centralized data lake that agents read from and write to, making function-specific applications less central.

For brands planning a campaign

A useful framing for positioning against AI-built alternatives: the durable advantage is the service layer around the product, not the code itself. It also gives a company a credible way to acknowledge AI risk without conceding the near-term market.

The failure cases are concrete.

• Two portfolio companies fell behind on reconciling cash and had to restate financials, and a client that deferred revenue on only some contracts lost three term sheets as diligence stalled.
• Gardner's central point is that basic accounting integrity, not sophisticated metrics, is where value and fundability are decided.

Gardner works through each red flag, explaining its underlying mechanism and consequences for investors or buyers. They draw on diligence and portfolio experience rather than a survey.

Cash accounting gets a qualified pass: it makes SaaS revenue noisy but can be manageable and clean. The problem is mixing methods across contracts.

For brands planning a campaign

For a SaaS company preparing for diligence or a raise, this is core workshop material: a finance and engineering session on mapping cloud costs to COGS versus R&D, tying the three statements together, and reconciling MRR to revenue. A brand sponsoring a CFO-facing workshop can frame it as a pre-diligence reporting audit built on Gardner's checklist.

Examples: Source 1

Left uncapitalized, those costs depress reported gross margin as growth accelerates, distorting CAC and LTV-to-CAC ratios and making a young SaaS business look less viable than its unit economics warrant. The post includes a concrete case: a growing company with negative gross margin because of front-loaded implementation costs, which Gardner says will struggle to raise capital if the accounting is not fixed first.

Gardner builds the case from GAAP treatment of costs required to deliver the promised service, then explains how growth distorts ratio metrics and closes with board-level caveats.

Gardner generally opposes capitalizing software development, with two caveats: keep cash burn visible in a cash-flow statement, and reserve the treatment for growth-driven margin depression rather than a temporary boost for a low-growth business.

For brands planning a campaign

Useful for a finance-oriented workshop segment aimed at SaaS operators scaling implementation teams or weighing a raise. A brand can host a session on when capitalization clarifies performance and when it obscures cash, positioning the topic as decision support rather than accounting theory.

Examples: Source 1

The answer is a discounted cash flow exercise the buyer will run, so the seller should run it first: 36 months of SaaS P&L, a terminal value based on month-36 run-rate operating profit, conservative retention assumptions, and little or no assumed new sales. Gardner gives practical thresholds — around a 20% discount rate for businesses above $10 million ARR and 25% below — and stresses that retention is the assumption that most changes the outcome, with gross margin also mattering in a revenue multiple.

Gardner built the model in ChatGPT using specified prompts and tested the resulting ARR multiples against several buyers, reporting that the multiples were, on average, surprisingly predictive.

For brands planning a campaign

This is a ready-made valuation-readiness workshop: walk a finance team through the assumptions a buyer will challenge — churn, discount rate, gross margin — before it enters a process.

• The fit is strongest for a company considering a sale or raise.
• Gardner's framing, build your own model and challenge the buyer's assumptions, gives the session a natural structure.

Examples: Source 1

They see losses coming from two places: lending too much relative to ARR, such as over roughly 1x ARR, and specific companies facing an immediate AI threat.

• Their base case is that most SaaS customers will not churn quickly, so a shrinking SaaS company can still service its debt, and lenders that underwrote retention and repayment capacity will take few losses.
• Exposure depends on the instrument: equity in a lender can lose everything even when the loans are sound, while a traditional fund holder is backed by the loan assets.

Gardner reasons by analogy from SVB and from two versions of SaaS Capital they funded, one of which failed in 2008 despite sound loans because of an asset-liability mismatch.

For brands planning a campaign

Useful for reaching credit, lending, and SaaS-finance audiences — a panel or workshop on retention metrics as the core underwriting signal, and on why equity in a lender differs from fund exposure. A brand can use this as evidence that Gardner can translate capital-markets risk into operating decisions for SaaS finance leaders.

Examples: Source 1

Free code has existed for decades, yet open source is only about 12% of application software revenue, because proprietary vendors bundle implementation, training, support, hosting, stability, and accountability.

• An AI-native startup can replicate Salesforce's code cheaply, in Gardner's argument, but still has to build the rest of the company, and legacy vendors' existing code is a sunk cost rather than a moat.
• They separate three threats: AI-built replacement, direct obsolescence of some analytical tools — the Excel-plus-Claude example — and the broader possibility that current software becomes irrelevant as the enterprise shifts toward a centralized data layer that agents read and write.

Gardner uses the open-source market-share comparison to test the replacement thesis, then distinguishes the timing and severity of the three threat types.

For brands planning a campaign

The strongest hook for an AI-and-software audience: a workshop or editorial session on which categories face near-term obsolescence versus long-run irrelevance, and what that means for product and investment decisions. Brands selling to SaaS operators or enterprise software buyers can use Gardner's framing to structure a panel on AI risk without resorting to hype.

Examples: Source 1

They describe one version as having an asset-liability mismatch and failing during the 2008 financial crisis, and the next as built on funds investors cannot withdraw.

View example

View example

The revenue-quality post splits an identical $8M top line into compounding repeat revenue and ad-funded one-time volume.

• The AI post lists five ways founders misread their own numbers without a CFO.
• Both use concrete operating situations to show why a number that looks fine can mislead a buyer.

Souza's frame is direct-to-consumer and ecommerce — SKU margins, repeat-customer rates, BFCM volume, fractional CFO support — while Gardner's is SaaS and the capital side: statements that must tie together, implementation costs, debt and valuation models for buyers and lenders.

Read original · Original 2

Jenkins' posts connect AI-native products and internal automation to revenue multiples, while their pricing talk asks how well a company understands the value gap on each account — a question close to Gardner's insistence that metrics reconcile to what was actually sold and used.

Jenkins writes from the operating seat of a software CEO, focused on org design, product monetization and pricing conversations, and cites the same kind of company data Gardner uses. Gardner writes from the finance and diligence seat, focused on reporting integrity, accounting treatment, and the credit and valuation questions a buyer or lender will raise.

Read original · Original 2

Nasraty's firm's anniversary post describes the same problem Gardner opens with — owners making important decisions without timely, accurate financial information — and the same remedy: clean systems and clear reporting.

Nasraty runs a consulting practice serving business owners broadly, so their writing centers on cloud-based accounting systems, real-time visibility, and CFO support for founders across industries. Gardner stays closer to SaaS specifics — MRR reconciliation, COGS definitions, ARR multiples, and private-credit underwriting.

Read original · Original 2

Souza offers a CFO’s checks for distinguishing revenue that compounds from revenue that merely covers the month: new versus repeat customers, margin by SKU, and one-time versus high-LTV cohorts.

• They also warn against dropping a P&L into AI and treating the output as a verdict, listing five failure modes, including hallucinated conclusions, missing ecommerce context, and misread cash versus profit.
• For a workshop on judging which revenue is worth defending and using AI only as an assistant under CFO oversight, Souza’s posts speak more directly to the task than Gardner’s.

Souza focuses on DTC and ecommerce margins and cash flow.

• The posts do not cover SaaS-specific reporting, including mapping cloud costs between COGS and R&D, tying a separate MRR report to income-statement revenue, or building a SaaS valuation model around retention and discount-rate assumptions.
• SaaS finance teams looking for diligence mechanics and multiple-setting should ask for those topics to be covered.

Examples: Source 1 · Source 2

Jenkins writes as a CEO about the shift from growth-at-all-costs to capital efficiency to a premium on AI, arguing that traditional SaaS KPIs now earn single-digit revenue multiples while AI-native companies command more.

• They describe continuously tracking three things per account: what was sold, what the customer is entitled to, and what they actually use.
• The example is a customer running $250 million in annual API calls who asked for a discount.
• For a brief on aligning pricing and packaging with delivered value, this is closer to the decision than a review of reporting mechanics.

Jenkins writes from a CEO and platform perspective, offering strategic framing rather than the accounting audit detail Gardner provides: bank reconciliations, cash versus accrual mix, three-statement integration, and MRR reconciliation. Teams tackling inaccurate or unintegrated financials should ask for step-by-step accounting guidance and concrete red flags to test their books against.

Examples: Source 1 · Source 2

Nasraty’s posts describe building financial clarity as a service: helping a founder outgrow a legacy accounting system, implementing cloud-based reporting, and giving owners real-time visibility and confidence in their numbers.

• One post links competitive defensibility to financial discipline, noting that as AI lowers the cost of building products, reliable accounting data and forward-looking CFO insight help determine whether an advantage is translating into value.
• This fits engagements that include reporting infrastructure and an ongoing advisory relationship.

The posts focus more on general financial visibility and system implementation than on SaaS-specific valuation modeling. Teams seeking a modeling workshop anchored in SaaS unit economics should ask for more on valuation multiples, retention assumptions, diligence-readiness checks, and the SaaS reporting pitfalls Gardner covers.

Examples: Source 1 · Source 2