Tech debt impact on COGS: when engineering becomes cost of revenue
The least-discussed financial impact of tech debt is the accounting one. As debt accumulates, more engineering work becomes operationally necessary to keep the service running, and accountants reclassify that work from R&D opex into COGS. The shift erodes gross margin in ways that show up in earnings calls long before they show up in engineering reviews.
The 90-Second Answer
Tech debt forces engineering work that is operational rather than developmental, and operational engineering work capitalizes into COGS under GAAP. The result is a quiet 3-8 percentage point gross-margin drag at scale. For a $100M ARR business at 78% gross margin, 4 points of drag is $4M of lost gross profit per year, recurring, attributable to a tech debt position that does not appear anywhere on the income statement.
The Accounting Rule
What goes into COGS for a SaaS business
US GAAP and IFRS converge on the principle that cost of revenue includes the costs directly required to deliver the product or service being recognised as revenue. For a SaaS business this typically includes: hosting infrastructure (the cloud bill), data costs (third-party APIs that the service consumes per customer), customer-success and support labour, payment processing, and the portion of engineering labour required to operate the service. The last category is where tech debt enters the calculation.
The judgement call is which engineering activities count as “required to operate” versus “developing new capability”. The conservative interpretation, which most public-company auditors apply, classifies the following as COGS-eligible: SRE and DevOps work required to keep the service running, customer-specific integration or implementation engineering, support engineering escalation handling, and any platform or infrastructure work whose absence would degrade service availability. The development-of-new-capability work (greenfield product features, exploratory R&D) sits in operating expenses.
The practical effect is that engineering labour is allocated across COGS and opex categories on every accounting period, with the allocation based on activity tracking, surveys, or fixed-ratio assumptions. The auditor reviews the methodology, not the per-engineer activity, but the methodology itself produces the ratio that matters for the gross-margin calculation.
The Tech Debt Connection
How accumulated debt migrates engineering into COGS
Tech debt creates more operational engineering work without creating more developmental work. The ratio shifts. Every quarter a team spends 30% of capacity firefighting accumulated issues is a quarter where 30% of engineering payroll is, under conservative GAAP interpretation, COGS-eligible. The work was supposed to be R&D opex, but its actual character (keeping the service running) is operational, and the auditor's allocation should reflect the actual character.
The accounting effect is rarely instantaneous because allocation methodologies use surveys or ratios that lag the actual activity by months or quarters. The lag is itself a problem because it masks the underlying deterioration: a CFO who reviews gross margin quarterly may see the margin compress over four to six quarters before the underlying ratio shift is identified, by which point the tech debt position is several quarters deeper than when it could have been remediated cheapest.
The strongest signal that tech debt is migrating into COGS is the trend of the engineering-cost-as-percentage-of-revenue ratio held against the gross margin trend. If engineering costs are growing as a percentage of revenue at the same time as gross margin is compressing, the most likely explanation is that engineering labour is reclassifying into COGS at the margin. CFOs who watch these two metrics together catch the deterioration months earlier than CFOs who track them separately.
The Tipping-Point Arithmetic
When the COGS impact becomes material
The COGS impact becomes financially material when the company scales past a revenue level where a few percentage points of gross margin translates into eight-figure annual gross profit dollars. The tipping point varies by gross margin level but is broadly visible at $50-100M ARR.
The 4pp midpoint at $100M ARR ($4M annual lost gross profit) is roughly the dollar size of a senior engineering manager's full annual budget. A debt remediation that recovers 2pp is the financial equivalent of half a department's annual spend.
The Public-Company View
How analysts read the gross-margin trend
For public SaaS companies, gross margin is one of the most-watched metrics on the income statement. Analysts compare against the company's own historical trend, against the public-comp peer group, and against the company's own management guidance. A gross margin trajectory that deviates from the peer group or from prior guidance triggers analyst questions and, in extreme cases, multiple compression at the stock-price level.
Tech debt is rarely named explicitly in the analyst conversation, but its effects are visible. A SaaS company whose hosting costs are growing faster than revenue is often a company carrying significant tech debt in the architecture (the per-customer cost is not being amortised efficiently). A SaaS company whose support engineering line is growing faster than revenue is often carrying product debt that manifests as support load. The analyst infers from the operational signals what the engineering organisation knows from the inside.
For pre-IPO companies preparing for the public listing, the gross-margin trajectory is one of the most scrutinised aspects of the S-1 financial story. A company with a deteriorating gross margin in the two years before IPO typically prices below the equivalent peer; a company with an improving trajectory typically prices above. Tech debt remediation that improves gross margin meaningfully in the pre-IPO window is a value-creation lever that founders sometimes recognise only late, when the IPO underwriter raises the question. See the late-stage framing page for the pre-IPO timing-specific treatment.
What Reverses It
Engineering investments that recover gross margin
The COGS-side gross-margin recovery is driven by engineering investments that reduce the operational labour required per unit of revenue. The four highest-leverage categories: infrastructure automation that reduces SRE load, observability investment that reduces incident-response load, multi-tenancy architecture that reduces per-customer integration work, and customer-success tooling that reduces support escalation rates.
Each of these investments is a structural debt remediation pitched as a gross-margin recovery. The CFO who hears “invest in observability” will deflect; the CFO who hears “invest $1.4M in observability to recover 1.8 percentage points of gross margin within four quarters” will engage. The translation from engineering activity to gross-margin recovery is the same translation pattern used in the CFO pitch page, applied specifically to the COGS-side of the income statement rather than the opex-side.
The timeline for gross-margin recovery is slower than the timeline for engineering velocity recovery. Velocity moves in weeks; gross margin moves in quarters. The CFO conversation needs to accommodate this asymmetry; promising a one-quarter gross-margin recovery from a tech debt initiative is over-promising and damages credibility for the next ask. The honest commitment is 2-3 quarters before initial measurable improvement, 4-8 quarters to peak recovery.
Cross-Reference
COGS sits inside the business-impact stack
The COGS impact is tightly coupled to the SaaS gross-margin page (the income-statement view of the same effect) and to the burn-rate impact (the cash-flow view). For the tax-vs-accounting distinction with IRC §174 R&D capitalization, see the CapEx vs OpEx page. The CFO pitch includes COGS framing as one of the most persuasive translations.
For the engineering-practitioner view of which architectural patterns most efficiently reduce per-customer operational load (multi-tenancy, infrastructure-as-code, observability stacks), see the sister site technicaldebtcost.com.
Field Notes
Frequently asked questions
Does engineering payroll count as COGS?+
Some of it does. Engineers whose work is required to deliver the service (DevOps, SRE, support engineering, customer-specific integration work) typically capitalize into COGS under GAAP. Engineers working on new product capabilities sit in R&D opex. The line moves toward COGS as the company matures and as the service becomes more operational.
Why does tech debt accelerate the COGS shift?+
Tech debt increases the proportion of engineering work that is operational rather than developmental: keeping the system running, debugging issues, performing migrations, handling escalations. This operational work is more clearly COGS-classified than greenfield development, so the worse the debt, the more engineering payroll migrates from opex to COGS in the accounting treatment.
What does this mean for gross margin?+
Direct hit. SaaS gross margins are benchmarked at 70-85% for top-quartile companies. Each percentage point of engineering payroll that migrates from opex to COGS reduces gross margin by approximately that percentage point (after accounting for the revenue base). A 3-5 percentage point drag from tech debt is common and material at scale.
Is the COGS shift visible in financial statements?+
Yes, in the gross margin trend line. Most CFOs notice the deterioration over 6-8 quarters before they trace it to tech debt. Companies that scale rapidly often see gross margin compress 200-400 basis points as the engineering organisation grows faster than the operational efficiency improvements.
Can tech debt remediation reverse the COGS shift?+
Yes, with effort. The shift back from COGS to opex requires structural improvement: automation that reduces operational engineering load, observability that reduces incident-response load, and platform investment that reduces per-customer integration work. None of these is instant; expect 4-8 quarters for measurable gross-margin recovery.
How does this differ from R&D capitalization under IRC §174?+
IRC §174 (post-TCJA) requires capitalization-and-amortization of R&D expenses for tax purposes. COGS classification is a GAAP financial-reporting question, separate from the tax treatment. The two interact but answer different questions: GAAP COGS affects reported gross margin; §174 capitalization affects cash taxes. See /capex-vs-opex for the §174 treatment.
Adjacent Reading