The Art of SaaS: How Software Learned to Suck on a Schedule

SaaS was supposed to be liberation. No more shrink-wrapped boxes, no more installation CDs, no more waiting for the next major release to get the feature you needed. Instead, software would live in the cloud, update continuously, and charge you a modest monthly fee. That was the pitch. The reality, roughly a decade in, looks different: you pay more every year for software that does less of what you need and more of what its investors want.

The Price Goes Up, the Value Goes Sideways

SaaS pricing inflation is running at roughly 8.7% year over year, according to industry tracking. That is nearly triple general consumer inflation. And it is not buying you proportionally more. Salesforce raised prices 9% in 2023 and another 6% in 2025; analysts at SaaStr estimated that roughly 72% of Salesforce’s go-forward revenue growth in 2025 came from price increases, not from new customers or expanded usage. The product did not get 72% better. The invoice did.

This is not an outlier. Across the industry, more than half of SaaS revenue growth now comes from raising prices on existing customers rather than winning new ones. When your growth strategy is “charge the people who already depend on you more money,” you have crossed the line from business model to protection racket.

The Bloat Spiral

Meanwhile, the software itself keeps getting heavier. Small and mid-sized businesses now use dozens of different SaaS applications, with some firms spending as much as $9,600 per employee per year. In a 2023 survey, finance leaders reported plans to trim software budgets by 10 to 30 percent after discovering unused licenses and overlapping tools. That is not a market functioning well; that is a market optimized for vendor revenue, not customer outcomes.

The bloat is not accidental. SaaS products are often built to be sold, not used. Features get added because they look good in a demo or a comparison chart, not because anyone asked for them. Each new feature adds complexity, slows down the interface, and creates new failure modes. But it makes the next sales call easier, so it ships.

Klarna’s CEO made headlines by shutting down both Salesforce and Workday, replacing them with an internal AI assistant. Marc Benioff responded by questioning Klarna’s governance. The exchange was revealing: a customer saying “your product is not worth what you charge” and the vendor’s response being, essentially, “you are holding it wrong.”

The SaaS Subscription Trap

The U.S. Department of Justice sued Adobe in 2024 for hiding cancellation fees and making the cancellation process deliberately difficult. BMW tried charging $18 per month to unlock heated seats that were already physically installed in the car. Waves Audio moved to subscription-only pricing and reversed within days after the backlash. These are not edge cases. They are the logical endpoint of a business model that prioritizes recurring revenue over delivering value.

A survey found that 72% of U.S. consumers believe there are too many subscription services. The average American spends $273 monthly on subscriptions. And churn is up 23%, with two-thirds of subscribers canceling at least one service within a six-month period. The market is telling you something.

What Actually Went Wrong

The original SaaS promise was genuine: continuous improvement, lower upfront costs, accessibility from anywhere. The corruption happened when the subscription model met venture capital math. Investors want predictable, growing recurring revenue. That creates three incentives that directly conflict with making good software:

First, never let the customer leave. Hence Adobe’s cancellation labyrinth, hence aggressive annual contracts, hence the slow death of perpetual licenses. JetBrains, to their credit, introduced fallback licenses that let you keep the software after 12 months of payments. They remain an exception.

Second, always add, never subtract. Removing a failed feature means admitting it was wrong. Adding features, regardless of quality, inflates the perceived value of the next tier. The result is software that tries to do everything and does nothing particularly well.

Third, extract more from existing customers. It is cheaper to raise prices on captive users than to win new ones, especially when switching costsThe cost or friction a user faces when moving from one platform to another, including time, money, and effort invested in the original. Also called switching barriers. are high. So prices go up, and the justification is always “added AI features” or “enhanced security” or some other vague improvement that nobody requested.

The Counterargument (and Why It Is Only Half Right)

The steelmanA rhetorical technique where you present the strongest possible version of an opponent's argument before refuting it. The opposite of a straw man. for SaaS is real: it democratized access to enterprise tools. A startup in Lagos can use the same project management software as a Fortune 500 company. Cloud delivery means automatic security patches, no IT department required. These are genuine advantages, and they matter.

But the counterargument mistakes the delivery mechanism for the business model. Cloud-delivered software is great. Subscription-only, price-hiking, cancellation-punishing, feature-bloating software-as-a-hostage is not. The problem is not the cloud. The problem is what happens when the cloud becomes the only option and the vendor has no incentive to respect your autonomy as a customer.

37signals launched ONCE, a line of pay-once, self-hosted tools, as a deliberate alternative. The fact that “you can buy software and own it” is now a radical market position tells you how far the baseline has drifted.

Where This Is Going

Enterprise SaaS spending is projected to reach $576 billion by 2029, according to Forrester. The market is not dying. But it is transforming, and not gently. The February 2026 sell-off erased roughly $1 trillion from software stock valuations. Investors are asking whether AI agents will replace the per-seat model entirely, whether “vibe codingThe practice of using AI to generate code from natural-language descriptions without reviewing or understanding the output, prioritizing speed over quality or security.” will let startups replicate complex platforms at a fraction of the cost, and whether customers will keep paying premium prices for software that a cheaper alternative can match.

Companies are already consolidating: the average enterprise reduced its SaaS stack from 371 applications to 220 between 2023 and 2024, a 40% cut. That is not trimming; that is a reckoning.

The SaaS model is not going to disappear. But the version of it that treats customers as captive revenue streams rather than people who chose your product and could choose to leave? That version deserves to suck. And increasingly, it does.

SaaS was supposed to be liberation. No more shrink-wrapped boxes, no more installation CDs, no more waiting eighteen months for a major release cycle to ship the feature you needed yesterday. Software would live in the cloud, update continuously, and charge a modest monthly fee. That was the pitch, circa 2010. The reality, roughly fifteen years in, requires a more granular examination: customers are paying more annually for software that optimizes for vendor economics rather than user outcomes, and the structural incentives behind this divergence are baked into the model itself.

The Pricing Ratchet: Quantifying the Disconnect

SaaS pricing inflation is running at approximately 8.7% year over year, according to industry benchmarking. For context, general consumer inflation in the same period hovered around 3%. The delta is not explained by proportional improvements in functionality.

Salesforce provides an instructive case study. A 9% price increase in July 2023 was followed by a 6% increase in 2025. Analysts at SaaStr estimated that these increases, combined with reduced discounting, accounted for roughly 25% of total revenue growth over that three-year window. More strikingly, approximately 72% of Salesforce’s go-forward growth in 2025 was attributed to pricing actions rather than new customer acquisition or usage expansion. The product roadmap did not generate 72% of the growth. The pricing team did.

This pattern is not idiosyncratic. Across the SaaS sector, more than 50% of revenue growth now derives from price increases on existing customers rather than net new logos. The industry has quietly shifted from a growth model predicated on winning customers through product superiority to one predicated on extracting more from a captive installed base. In economic terms, this is rent-seeking behavior: the vendor’s market power comes not from continued innovation but from switching costsThe cost or friction a user faces when moving from one platform to another, including time, money, and effort invested in the original. Also called switching barriers. and data lock-in.

Feature Bloat as a Structural Incentive

Small and mid-sized firms now operate dozens of different SaaS applications, with mid-market companies spending $250,000 to $1 million annually on 50 to 70 tools, and some paying as much as $9,600 per employee. Software subscriptions have become the third-largest operating expense for many organizations, trailing only personnel and real estate. In a 2023 survey, CFOs reported plans to trim software budgets by 10 to 30 percent after discovering unused licenses and redundant applications.

The bloat is architecturally rational from the vendor’s perspective. Each feature added to a platform, regardless of adoption rate, serves multiple strategic purposes: it inflates comparison-chart competitiveness, justifies tier differentiation and upselling, and increases switching costs (migrating away from a tool with 200 features is harder than leaving one with 20). The cost of this strategy is borne by users in the form of interface complexity, performance degradation, and cognitive overhead.

The enshittificationA three-phase pattern where platforms first attract users, then exploit them for business customers, then exploit those business customers while degrading all earlier beneficiaries. Coined by Cory Doctorow. framework Cory Doctorow articulated for platforms applies with precision to SaaS: first, the product is good to attract users; then, it degrades the user experience to extract value for business customers; finally, it extracts value from everyone to generate returns for shareholders. The subscription model accelerates this cycle because recurring revenue creates a captured audience that cannot easily exit.

Klarna’s decision to terminate its Salesforce and Workday contracts, replacing them with an internally developed AI assistant, became a minor industry inflection point. CEO Sebastian Siemiatkowski framed the move as seeking “more agility, lower costs, and fewer bloated tools.” Marc Benioff’s public response, questioning Klarna’s data governance practices, illustrated the vendor-side assumption that departure from a major SaaS platform is itself evidence of dysfunction rather than rational evaluation.

The Cancellation Architecture

The U.S. Department of Justice sued Adobe in 2024 for what the complaint described as making cancellation “onerous and complicated” while hiding early termination fees. BMW’s $18/month heated-seat subscription, which monetized hardware already installed in the vehicle, was abandoned by 2023 after sustained public criticism. Waves Audio’s 2023 pivot to subscription-only licensing was reversed within days.

These incidents are not aberrations; they are natural consequences of a business model where customer retention is measured by contract stickiness rather than satisfaction. When 78% of CFOs report being blindsided by hidden fees or price hikes in SaaS contracts, the information asymmetryA situation where one party in a transaction has more or better knowledge than the other, allowing the informed party to gain advantages at the expense of the less informed party. is structural, not incidental.

Consumer response is measurable: 72% of U.S. consumers believe there are too many subscription services available. Churn rates are up 23%, with two-thirds of subscribers canceling at least one service within six months. One in four SaaS users quits within 30 days of signing up. The market signal is unambiguous, but SaaS vendors are structurally disincentivized from hearing it because short-term retention tactics (longer contracts, harder cancellation, bundled pricing) can delay churn even as underlying satisfaction erodes.

The Quality Dimension: AI Accelerates the Existing Problem

Running parallel to the pricing and bloat problems is a software quality crisis that predates AI but is being amplified by it. According to ReversingLabs, 97% of organizations are now using or piloting AI coding assistants, and nearly one-third report that AI generates most of their code. Developers report saving 3.6 hours weekly and shipping 60% more pull requests.

The catch: 65% of organizations acknowledge that AI assistants introduce new security vulnerabilities, and 81% lack complete visibility into AI usage across their development teams. One engineer’s analogy for AI-generated code has circulated widely: “Somebody gifted you a dog. Zero purchase price, daily work for decades.” The velocity gains are real. The maintenance burden they create is also real, and largely unaccounted for in current productivity metrics.

For SaaS customers, this compounds the existing problem. You are paying more for software that ships faster, with more features, built partly by tools that introduce vulnerabilities nobody is tracking. The traditional quality assurance bottleneck, slow but functional, has been replaced by a velocity-first pipeline where architectural decisions that once happened weekly now happen daily.

The SteelmanA rhetorical technique where you present the strongest possible version of an opponent's argument before refuting it. The opposite of a straw man. (and Its Limits)

The case for SaaS remains real: cloud delivery democratized access to enterprise-grade tools. A startup anywhere in the world can deploy the same infrastructure that was once the exclusive province of Fortune 500 IT departments. Automatic security patching eliminates a class of vulnerability that plagued on-premise software for decades. Multi-tenancy creates economies of scale that genuinely reduce costs for small users.

These advantages, however, are features of the delivery mechanism, not the business model. Cloud-delivered software with a perpetual license captures the same benefits without the extraction dynamics. JetBrains’ fallback licensing (perpetual use rights after 12 months of subscription) and 37signals’ ONCE product line (pay-once, self-hosted tools) demonstrate that alternatives exist and find willing buyers. The fact that “you can purchase software and own it” constitutes a differentiated market position in 2026 is itself a diagnostic.

Market Trajectory: Correction, Not Collapse

Enterprise SaaS spending is projected to grow from $318 billion in 2025 to $576 billion by 2029, per Forrester. The market is expanding. But the February 2026 sell-off, which erased approximately $1 trillion from software stock valuations, signals that investors are pricing in structural risk: per-seat pricing becoming obsolete as AI agents replace human users, vibe-coded startups replicating complex platform functionality at marginal cost, and enterprise consolidation reducing the total addressable market for point solutions.

Enterprises reduced their average SaaS stack from 371 applications in 2023 to 220 in 2024, a 40% reduction. 53% of organizations actively consolidated overlapping applications in 2024. This is not a trim; it is an architectural reckoning, and it disproportionately threatens horizontal point-solution vendors with weak switching costs.

The SaaS model is not going to vanish. Vertical-specific vendors controlling proprietary data in regulated industries (healthcare, pharmaceuticals, finance) have structural moats that AI disruption cannot easily breach. But the broad middle market of generic, horizontally positioned, feature-bloated, annually-more-expensive platforms? That segment is facing pressure from below (AI-enabled alternatives), from above (enterprise consolidation), and from within (customers who have simply had enough).

SaaS was a genuine innovation. What it became, for too many vendors, is a mechanism for turning customer dependency into a revenue growth strategy. The market correction currently underway is not punishing the model. It is punishing the abuse of it.