SaaS (Software as a Service): from the founding of Salesforce (1999) to the subscription-saturation era of 2026

SaaS —Software as a Service— is the business model in which software is delivered as a service accessible over the internet on a subscription basis instead of being sold as a licensed product for local installation. The user pays (typically monthly or annually) for access to the application; the provider maintains the infrastructure, the updates, the security, and the support.

This sounds obvious today because almost all modern enterprise software follows this model. But twenty years ago it was a radical idea that was publicly ridiculed. The transformation of how companies buy and use software is one of the most significant industrial stories of the past two decades, and understanding its context helps you make informed decisions in 2026 — especially at a moment when subscription saturation has become a real operational problem for many companies.

The origin: Salesforce, 1999, and the "End of Software" campaign

Modern SaaS has a founding case with a precise date. Marc Benioff, a former Oracle vice president, founded Salesforce.com on March 8, 1999 in San Francisco with a public manifesto: "The End of Software."

The idea was radical at the time. In 1999, buying enterprise software meant:

• Upfront outlays of hundreds of thousands to millions of dollars for licenses.
• Your own infrastructure to host it (servers, IT staff).
• Implementations that took months or years.
• Costly ongoing maintenance.
• Complicated upgrades that required full-blown projects.
• Periodic replacement that also required projects.

Salesforce proposed something different: the customer would access the CRM application via a web browser, paying for monthly usage, without installing anything locally. Benioff's company took care of hosting, maintenance, updates, and security. The customer only had to worry about using the product.

The early marketing campaigns were provocative: Benioff staged fake protests outside Siebel Systems conferences (then the leader in enterprise CRM) with demonstrators holding "End of Software" signs. It was guerrilla marketing designed to force the conversation. It worked. Salesforce grew incredibly fast; it went public in June 2004; by 2010 it was already a multi-billion-dollar company.

The term "SaaS" itself took hold around 2001-2002. Before that, the model was typically called ASP (Application Service Provider), an older term from the '90s. SaaS replaced ASP because of different technical and economic implications — ASPs typically managed pre-existing software; SaaS companies build web-native applications from the ground up.

The consolidation: the 2000s and 2010s

Once the viability of the SaaS model was established, a wave of companies built on it emerged:

NetSuite (1998, even earlier than Salesforce, founded by Evan Goldberg with backing from Oracle's Larry Ellison). Cloud ERP. Acquired by Oracle in 2016.

WebEx (1995, web conferencing). Acquired by Cisco in 2007.

Concur (SaaS travel expense management). Acquired by SAP in 2014.

Workday (2005, HR and finance). IPO 2012.

ServiceNow (2003, IT service management). IPO 2012.

Veeva (2007, vertical for the pharmaceutical industry). IPO 2013.

Atlassian (2002, developer tools). IPO 2015.

Slack (2013, communication). IPO 2019, bought by Salesforce in 2021 for $27.7B.

Zoom (2011, video conferencing). IPO 2019. Explosive growth during the pandemic.

HubSpot (2006, marketing automation, covered in previous pieces).

Shopify (2006, e-commerce, covered in marketplace).

Adobe executed probably the most significant SaaS transition by an established company: in May 2013 it discontinued the sale of boxed software (Creative Suite 6) and moved everything to Creative Cloud on a subscription basis. It was a controversial decision —many professionals criticized the loss of the ownership model— but financially transformative. Adobe went from release-dependent revenue fluctuations to predictable recurring revenue streams. The stock rose consistently over the following years.

Microsoft made a similar transition with Office 365 (launched in 2011, a brand that later evolved into Microsoft 365). The shift from boxed Office to subscription Office transformed Microsoft's economics.

By the mid-2010s, SaaS was the dominant model for new enterprise software. Companies launching software with a perpetual-license model were increasingly the exception.

The distinctive economics of SaaS

The SaaS model has economics distinct from traditional software, with its own specific vocabulary and metrics:

ARR (Annual Recurring Revenue) and MRR (Monthly Recurring Revenue). Recurring subscription revenue, annualized or monthly. Central metrics that SaaS companies report quarterly. The model is highly predictable.

ACV (Annual Contract Value). The value of a new annual contract. Distinguishes between new sales, expansion, and renewal.

CAC (Customer Acquisition Cost). The cost of acquiring a new customer. Covered in CAC as a diagnostic.

LTV (Lifetime Value). The total value a customer will generate over the course of their relationship with the company. Covered in LTV.

Payback period. How long it takes a new customer to repay the CAC invested. Healthy SaaS: < 18 months ideally; > 24 months is problematic.

Churn rate. The percentage of customers (or of revenue) lost in a period. Typical B2B SMB SaaS: 5-10% annually; enterprise SaaS: 2-5% annually; B2C SaaS: can be 30-50% annually.

Net Revenue Retention (NRR). How much the revenue from existing cohorts grows (or shrinks) through upsells, downgrades, and churn. NRR > 100% indicates that existing customers generate more revenue each year through expansion. The best B2B SaaS companies reach NRR of 130-150%; "decent" starts at 100-110%.

Magic Number. A growth-efficiency metric: (new ARR current quarter − ARR previous quarter) / sales and marketing cost of the previous quarter. > 1.0 suggests efficient growth; < 0.75 suggests problems.

Rule of 40. A heuristic that combines growth and profitability: growth rate (%) + profit margin (%) ≥ 40 indicates a healthy SaaS company.

Gross Margin. SaaS gross margin tends to be high (70-85%) because, after serving the first customer, serving the Nth has a low marginal cost.

Knowing these metrics is not an academic exercise. Well-run SaaS companies are evaluated internally and by investors with this logic. A pricing, retention, or expansion decision that seems reasonable on its own can destroy the model if you don't calculate its effect on these aggregate metrics.

Types of SaaS

The SaaS model spans very different categories:

Horizontal SaaS. Serves multiple industries with the same functionality. CRM (Salesforce, HubSpot), email marketing (Mailchimp), project management (Asana, Monday, ClickUp), communication (Slack, Microsoft Teams).

Vertical SaaS. Specialized for a single industry. Veeva (pharmaceutical), Procore (construction), Toast (restaurants), ServiceTitan (HVAC and trades), Mindbody (fitness and wellness). Typically higher ticket sizes than horizontal SaaS because of the deep specialization.

Self-serve SaaS / PLG (Product-Led Growth). The product is tried/purchased without sales involvement. Free tiers or trials. Notion, Figma, Slack at launch. A growing model since 2015.

Enterprise SaaS. Consultative sales with long cycles, large contracts, customizations. Workday, ServiceNow, Salesforce enterprise.

B2C SaaS. Subscriptions for consumers. Spotify, Netflix, Dropbox for individual users, 1Password, Notion personal.

Verticalized B2C. Creator platforms (Patreon, Substack), education (MasterClass, Coursera Plus), wellness (Calm, Headspace).

API as a Service / Infrastructure as a Service. Twilio, Stripe, the OpenAI API. Technical software that other companies integrate. A SaaS model, but with different economics and a different customer.

The 2026 reality: subscription saturation

Something that has changed significantly since roughly 2022: companies have discovered that the aggregate cost of their SaaS stack is enormous and often out of control. Available data:

SaaS sprawl is a quantified problem. A typical mid-sized company (200-500 employees) uses, on average, 100-200 distinct SaaS applications, according to studies by Productiv, Zylo, and Vendr. Large companies can have 500-1,000+. Most of those subscriptions are never audited.

Significant annual cost. For a mid-sized company, SaaS spend can represent 10-20% of the total operating budget. It's a spending category that has grown faster than the rest of the budget for a decade.

Typically low utilization. Many paid seats go unused. Consistent studies show that between 30% and 50% of contracted SaaS seats are underused (access without real usage) or unused (paid seats that nobody uses).

Functional duplication. Companies with 3-4 tools that do roughly the same thing (several project management tools, several communication tools, several storage tools) without anyone having consolidated them.

Impact of the 2022-2024 SaaS winter. The crash in the valuations of public SaaS companies (Cloudflare, Twilio, Snowflake, and Datadog lost 50-70% of their value in 2022-2023, recovering partially in 2024-2026) reflects the fact that SaaS's "golden age of growth at any price" is over. The market shifted to valuing profitability and efficiency.

Systematic SaaS audits. In response, a category of tools has emerged (Productiv, Zylo, Vendr, Cledara, BetterCloud, Torii) specifically to manage a company's SaaS portfolio. They identify duplicate apps, unused seats, and subscriptions that renew automatically without being used. They can produce savings of 20-40% of total SaaS spend.

Growing consolidation. Companies like Microsoft, Google, and Salesforce offer increasingly broad suites in an attempt to replace multiple point SaaS products. The argument "use the Microsoft 365 you already have instead of Notion + Slack + ClickUp + Zoom + Asana" is gaining traction in buying decisions.

For a company in 2026, the discipline of managing the SaaS stack has gone from optional to fundamental. Without periodic audits, costs accumulate invisibly.

How to choose and manage SaaS well

The practices that distinguish companies with a healthy SaaS stack from saturated ones:

A complete initial audit. Before adding any new SaaS, create an exhaustive list of what you already have, what it solves, who uses it, and what it costs. Most companies are surprised by what they find.

A clear procurement policy. Who approves new subscriptions, and on what criteria. Without a policy, anyone on the team can add tools that pile up.

Centralized billing. A single point that sees all SaaS billing. If it's spread across personal cards and departments, there is no real visibility.

SSO (Single Sign-On) where it applies. For companies with many employees, SSO reduces onboarding friction and enables quick revocation when someone leaves. Some vendors charge significantly more for SSO; the debate over the "SSO tax" is ongoing.

Renegotiating contracts at renewal. Don't accept automatic renewal without reviewing. Most SaaS vendors are open to negotiation, especially for companies with significant business. Cledara, Vendr, and similar tools can help with negotiation.

Regular removal of unused seats. A monthly or quarterly audit of who uses what. Removing access for people who no longer use it reduces cost.

Periodic evaluation of duplication. Do you have Asana, Monday, ClickUp, and Notion managing projects differently across different teams? Consolidating can save thousands a month and reduce coordination friction.

Vendor lock-in awareness. Before committing critical data to a SaaS, verify export policies, open formats, and migration options. Companies that don't think about lock-in discover it when it's too late.

Compliance and privacy. GDPR, HIPAA in healthcare, SOC 2 for enterprise — verify that the SaaS complies before uploading sensitive data.

Independent backups. Don't assume the SaaS will back up your data adequately. For critical data, periodic independent backup is prudent.

Common mistakes in using SaaS

Paying for a higher plan without needing the features. Many companies choose the more expensive plan "just in case" without auditing whether they actually use its distinctive features. Downgrading is usually easy; so is upgrading when needed.

Not canceling free trials. Trials that automatically convert to paid if you don't actively cancel. They rack up surprise bills.

Approving SaaS out of one employee's enthusiasm. "This tool is great" without evaluating real need, fit with the current stack, or three-year cost. It leads to tools adopted by one person that the team never uses.

Tool-hopping. Switching SaaS every few months for novelty's sake. Migration and learning costs outweigh the benefits in most cases.

Not reading the terms of service. Particularly the section on data ownership, retention, and deletion. Some SaaS vendors retain data for months after cancellation.

Ignoring discounted annual contracts. Many SaaS vendors offer a 10-20% discount for annual payment vs. monthly. For tools you'll use long term, it pays off.

Not documenting what each SaaS is used for. When someone asks "Why are we subscribed to this tool?" and nobody knows, there's a problem.

Invisible lock-in. Building processes that depend on a specific SaaS feature without thinking about what happens if the vendor raises the price or discontinues the feature.

Not connecting to business metrics. Paying for SaaS without knowing whether it's generating proportional value. Audit the ROI of each tool category against results.

The 2026 trend: AI built into SaaS

Something that is changing the landscape: every SaaS has incorporated generative AI capabilities since 2023, and this is transforming both pricing and usefulness:

Premium pricing for AI features. Notion AI, ChatGPT built into Microsoft Copilot, Salesforce Einstein, HubSpot Breeze AI. Providers charge extra for AI capabilities, frequently as an add-on to their base plan.

Productivity acceleration. For many tasks (writing copy, generating images, processing data, drafting emails), SaaS with well-integrated AI produces significant time savings.

Questioning per-seat pricing models. If AI does the work of several humans, does it make sense to charge per seat? Some vendors are experimenting with usage-based or output-based pricing.

Concentration toward large vendors with their own AI. Vendors with preferential access to models (Microsoft with OpenAI, Google with Gemini) have an advantage over small vendors that depend on third-party APIs.

Saturation with superficial "AI features." Many SaaS products added AI so quickly that the features are more demo than real utility. Auditing whether AI features genuinely improve the workflow is a necessary discipline.

SaaS and creative operations

For a brand or agency, the SaaS stack operationally defines how work gets done: which tools for which function, what integration between tools, what dependencies, what recurring costs. Without conscious management, the stack grows and coordination breaks down.

That coordination is the discipline of creative operations: approval workflows can integrate control over which SaaS gets approved, content production benefits from tools that reduce friction, and creative KPIs must justify the ROI of the tools the creative team uses.

Polimake itself is an example of SaaS applied to creative operations: Studio, Studio, and Media are SaaS products designed to coordinate creative work. The choice of these tools (vs. Drive folders + Slack + email + spreadsheets) reflects a decision about what level of investment in infrastructure the creative workflow deserves.

---

If you lead technology, operations, or procurement and you arrived here looking for an answer about SaaS, the most useful thing you can take from this article is probably the combination of three ideas: SaaS is not a neutral category; it has distinctive economics and dynamics (ARR, MRR, churn, and NRR are metrics that matter), subscription saturation is a real operational problem in 2026 (most companies pay more for unused SaaS than they think), and the discipline of managing the SaaS stack is now a basic part of operations, not optional. Companies that periodically audit their SaaS portfolio and eliminate duplication find significant savings without hurting productivity.

To round this out, the cloud covers the broader technical context where SaaS is a specific category, CMS covers a specific vertical of SaaS for content management, and marketplace covers an adjacent platform model.

Quick references

• The cloud — the broader technical context.
• CMS — a specific SaaS vertical.
• Marketplace — an adjacent platform model.
• How much a CMS costs — the real pricing of one type of SaaS.
• LTV — a central metric in SaaS economics.
• CAC as a diagnostic — the other foundational metric.