Is SaaS Dead? What AI Is Really Changing in 2026

Short answer: SaaS is alive and still growing. Gartner forecast that worldwide end-user spending on SaaS would reach $299.1 billion in 2025, up 19.2% from 2024, and in July 2026 it forecast total IT spending of $6.37 trillion for 2026, up 14.2%. What AI is changing is the shape of SaaS: who does the work inside the software, what the price is attached to, and what a buyer expects to receive for the money.

For agencies and service businesses, that shift is an opening. The software your clients rely on stays the foundation. AI makes it practical to sell the result that software produces (a booked appointment, an answered call, a paid invoice) as the product itself. That model is called Results as a Service (RaaS).

Where the "SaaS is dead" idea came from

In December 2024, Microsoft CEO Satya Nadella told the BG2 podcast that business applications are largely databases wrapped in business logic, and that this logic would increasingly move into AI agents (Cloud Wars transcription). Headlines compressed that into "SaaS is dead." Nadella himself was describing where the work happens: inside agents rather than screens and forms.

Read closely, the prediction is about the interface and the labor. The database underneath becomes more important, because an agent needs somewhere to read a customer's history, write down what happened and check what should happen next. That is exactly what a system of record does.

What the spending data shows

Sequoia's estimate is that businesses spend roughly six dollars on services for every dollar they spend on software. That ratio explains the excitement: when software can perform part of a service, the addressable market grows well beyond the software budget.

Three things AI is actually changing

1. The seat is losing ground as the unit of price

When a person does the work, charging per user makes sense: more users, more work, more value. When an agent does part of the work, the number of log-ins says less about the value delivered. Kyle Poyar's 2025 State of B2B Monetization survey of more than 240 software companies found seat-based pricing as the primary model fell from 21% to 15% in a year, while hybrid models rose from 27% to 41%. Outcome-based pricing was the primary model for 5% of companies, and 25% expected to use it by 2028 (Growth Unhinged).

2. Software companies are charging per outcome

Several large vendors now price their AI agents by the result. Fin (formerly Intercom) lists $0.99 per outcome such as a resolved conversation, HubSpot moved its Customer Agent to $0.50 per resolved conversation in April 2026, and Zendesk charges per automated resolution. The outcome-based pricing guide has the full table with sources and dates.

The pattern matters more than any single price: buyers are getting used to paying for the job done.

3. Services are becoming software

Foundation Capital called this "service as software" in 2024 and sized it as a $4.6 trillion opportunity (Foundation Capital). Work that used to need a person on the clock, such as answering the phone, qualifying a lead or chasing an unpaid invoice, can now be carried out by software and delivered as a finished task.

What stays valuable in SaaS

AI agents depend on the same foundations good SaaS has always provided:

• A system of record. Contacts, conversations, deals, appointments and payments in one place, so every action has context.
• Connections to real channels. Phone numbers, text messaging registration, email domains, calendars and payment accounts are where customers actually reach a business.
• Permissions and history. Who can see what, who changed what and what a customer agreed to.
• Workflows a person can read. A team can inspect, test and adjust what the automation does.

The stronger these foundations are, the more an agent can do on a business's behalf with confidence.

What this means for agencies

Many agencies already resell software under their own brand: the client gets a login, the agency gets a monthly subscription. That works well for clients who open the software every day and run it themselves.

AI adds a second offer. The agency operates the system for the client and charges for an agreed result. The client buys what they wanted in the first place (more appointments, faster replies, more reviews) and the agency keeps the machine running.

Here is how the two offers look for the same dental practice:

Both offers can run on the same platform, and an agency can sell both: software for clients who want to run it, results for clients who want it done. RaaS vs SaaS compares the two models in detail.

A practical next step

1. Choose one result your best clients already measure. Booked appointments, answered calls, reviews collected or invoices paid are good starting points.
2. Write down how it is counted. Name the record that proves it, the time window and how duplicates are handled.
3. Build the delivery system in one client account. Connect the phone number, calendar and follow-up workflow, then test it with your own contact details.
4. Price a pilot. A base fee plus a per-result fee keeps both sides comfortable while you learn the real numbers.
5. Report from the records every month. Share the list of results, each linked to its contact and date.

How to switch from SaaS to RaaS turns these steps into a 30-day plan, and Results as a Service examples lists offers you can adapt.

Sources

• Gartner, public cloud end-user spending forecast, November 19, 2024.
• Gartner, worldwide IT spending forecast, July 27, 2026.
• Cloud Wars, coverage of Satya Nadella's BG2 remarks, May 12, 2025.
• Growth Unhinged, 2025 State of B2B Monetization, June 4, 2025.
• Sequoia Capital, Services: The New Software, March 5, 2026.
• Andreessen Horowitz, Good News: AI Will Eat Application Software, March 2, 2026.
• Foundation Capital, AI leads a service as software paradigm shift, 2024.