Skip to content 
If you’ve spent any time on tech Twitter, LinkedIn, or in a VC’s Substack this year, you’ve heard the funeral march. “SaaS is dead.” “AI agents are eating software.” “Why would anyone pay for Salesforce when you can vibe-code a CRM in a weekend?”
It’s a great story. It gets clicks. Satya Nadella said “SaaS is dead” on a podcast back in December 2024 and the line has been recycled in a thousand hot takes since. In February of this year, the narrative got teeth: roughly $285 billion in software market cap evaporated in a single trading session. Salesforce stock alone is down about 30% year to date. ServiceNow dropped 7%. Intuit dropped 11%. The “SaaSpocalypse” had a name and a body count.
Here’s the problem with the funeral. Nobody checked the pulse.
While the market was busy writing Salesforce’s obituary, Salesforce was busy writing checks. Big ones. Since the start of this calendar year, Salesforce has signed or closed deals for Momentum, Cimulate, Convergence, Doti, Contentful, m3ter, and most recently Fin (the company formerly known as Intercom) for $3.6 billion. That’s not the spending pattern of a company that thinks its core business is going away. That’s the spending pattern of a company buying up every piece it needs to own the next layer of enterprise software before anyone else gets there first.
If you run a nonprofit, a public sector agency, or really any organization that depends on Salesforce to run operations, this matters to you directly. Let’s get into why.
The Narrative vs. The Numbers
The “SaaS is dead” argument sounds intuitive. AI can write code now. If a five-person team can spin up a working app over a weekend, why would any company pay six or seven figures a year for licensed software?
It’s a fair question. It’s also missing the point, and the actual financial results say so.
Salesforce closed fiscal year 2026 with $41.5 billion in revenue, up 10% year over year. Then Q1 of fiscal 2027 came in at $11.1 billion, up 13%. Agentforce, the company’s flagship AI agent platform, crossed $1 billion in annual recurring revenue across more than 18,500 customers. Salesforce liked those numbers enough to raise full-year guidance to a range of $45.9 to $46.2 billion.
That’s not a company in decline. That’s a company whose AI bet is starting to pay off, while the stock price gets punished for a narrative that hasn’t caught up to the balance sheet yet.
This disconnect between sentiment and fundamentals isn’t unique to Salesforce. IDC has been blunt about it: SaaS isn’t dying, it’s metamorphosing. The traditional per-seat license is getting replaced by consumption and outcome-based pricing, but the platforms underneath aren’t disappearing. They’re becoming something else: the system of record that AI agents need in order to act at all. An agent that can draft an email is a parlor trick. An agent that can draft the email, log it against the right account, trigger the next step in a workflow, and respect your org’s data governance rules is a business tool. You don’t get the second one without a platform underneath it.
Forbes ran a useful framing on this earlier in the year, scoring major SaaS players across data moat, governance, and platform shift. The conclusion: the “SaaSpocalypse” was real as a stock market event and wrong as a business thesis. Long live SaaS, just not the SaaS of 2019.
What Salesforce Is Actually Building
So what is Salesforce buying, and why does it matter for nonprofits and public sector organizations specifically?
Look at the pattern in this year’s acquisitions and a clear shape emerges. Salesforce isn’t randomly bolt-on shopping. It’s filling in every gap between “an AI agent exists” and “an AI agent can be trusted to run your operations.”
Momentum (closed March 2026) gives Agentforce and Slack the ability to pull insight out of unstructured conversations happening on Zoom and Google Meet, not just structured CRM data. For an org running donor calls or constituent meetings, that’s the difference between an agent that only knows what got typed into a record and one that actually heard the conversation.
Cimulate (closed March 2026) strengthens Agentforce Commerce with better product discovery and shopper intent modeling. If your org runs any kind of e-commerce, merchandise, or donation storefront, this is the layer that makes a storefront agent actually useful instead of a glorified search bar.
Doti, an agentic enterprise search company, is being folded in to strengthen unified search and knowledge discovery across the platform, with Slack as the conversational front end. Translate that for a nonprofit: staff stop hunting across five systems to answer “what’s our relationship with this funder,” and an agent surfaces it.
Contentful, the headless CMS acquisition that closed Q1 of this fiscal year, fills a hole Salesforce has had forever: a real content layer. Salesforce has talked about deeper Agentforce integration so content gets dynamically assembled by AI rather than manually built page by page. For comms-heavy nonprofits running constant campaign content, that’s a direct hit on workload.
Fin, the $3.6 billion acquisition of the company formerly known as Intercom, is the biggest of the year and the clearest statement of intent. Fin’s AI agent resolves customer queries end-to-end across chat, email, WhatsApp, SMS, phone, and Slack. Salesforce isn’t buying a feature here. It’s buying a fully built, already-proven customer agent and bolting it directly onto Agentforce so customers of every size can deploy autonomous agents without building them from scratch.
And then there’s m3ter, a SaaS utility pricing platform, which tells you something quieter but just as important: Salesforce is also retooling its own pricing infrastructure for a consumption-based, agent-driven world. That’s the company preparing for the same pricing shift IDC has been forecasting industry-wide.
None of this reads like a company hedging against irrelevance. It reads like a company assembling the full stack: data ingestion, search, content, commerce, customer service automation, and the billing infrastructure to charge for all of it on a usage basis instead of a seat basis.
The Counterargument, Taken Seriously
To be fair to the skeptics, not everyone making the “SaaS is dead” case is just chasing clicks. There’s a real version of this argument worth sitting with.
Klarna is the case study everyone points to. The company says it replaced more than 1,200 SaaS tools with a combination of large language models and a knowledge graph, built in-house. It’s a real example of a company tearing out licensed software and rebuilding it with AI. Shopify’s CEO told his own team that AI use is now a prerequisite for every hire, and he’s been candid that for a shopper who starts in an AI chat instead of a search bar, the traditional storefront disappears. Those aren’t hypotheticals. They’re operating decisions at real companies, made in public.
IDC’s own forecasting backs up part of the skeptics’ case too: by 2028, pure seat-based pricing is expected to be effectively obsolete, with the large majority of vendors rebuilding their pricing around consumption or outcomes instead of headcount. If your mental model of Salesforce is “they charge you per user license,” that model is already going away, and Salesforce knows it. The m3ter acquisition we mentioned above is Salesforce buying the infrastructure to make that exact pricing transition themselves.
So the honest version of the argument isn’t “SaaS is dead.” It’s “the SaaS business model you grew up with is dead, and the companies that don’t see that coming will get run over.” Salesforce clearly sees it coming. That’s the entire point of this post. The acquisitions aren’t defensive. They’re Salesforce building the next pricing model, the next interface, and the next governance layer before the disruption arrives instead of after.
Klarna pulled off the build-it-yourself approach because Klarna is a fintech company with an enormous internal engineering org and a CEO willing to bet the operating model of the entire company on it. Most nonprofits and public sector agencies are not Klarna, and have no interest in becoming a software engineering shop. For everyone else, the realistic path is riding the platform that’s already doing this work for you.
Why “Bellwether” Is the Right Word
A bellwether isn’t the company that’s loudest about AI. It’s the company whose moves tell you where the whole market is actually headed, because its scale forces it to solve the hard problems first.
Salesforce has direct experience here. Within weeks of launching Agentforce, researchers found a critical vulnerability. Most companies would have quietly patched it and moved on. Salesforce built Agent Fabric and Trusted Agent Identity, governance layers specifically designed to let AI agents act on enterprise data without creating an open door. That’s not a sexy press release. It’s the unglamorous infrastructure work that separates a platform you can actually deploy agents on from a demo that falls apart in production.
This is the structural advantage that the “vibe-code a CRM over the weekend” crowd consistently underestimates. A homemade agent built on a weekend sprint doesn’t have a governance layer. It doesn’t have a data moat built from over a decade of structured customer relationship data. It doesn’t have regulatory standing. Salesforce holds approved vendor status with the US Senate for Marketing Cloud. No AI agent, however clever, can code its way into that kind of trust relationship. Trust like that gets built over years, through audits, compliance reviews, and track record. It can’t be prompted into existence.
That’s also precisely why Belmar has built our practice the way we have. We’re not a generalist AI consultancy chasing whatever model is trending this quarter. We’re Salesforce-exclusive, and we have been for over fifteen years, because we bet a long time ago that the platform itself, not any single feature on top of it, would be the durable asset. The acquisitions this year are proof we bet right.
What This Means If You Run a Nonprofit or Public Sector Org
If you’re evaluating whether to double down on your Salesforce investment or hedge toward some AI-native alternative, here’s the honest read: the alternative doesn’t exist yet at the scale or trust level your organization needs, and the platform you already have is getting more capable by the month, often without you having to do anything.
The practical risk isn’t that Salesforce becomes irrelevant. It’s that your org keeps treating Agentforce and the surrounding ecosystem as a future project instead of a current capability. Every acquisition we covered above is showing up in your org’s roadmap whether you’ve turned it on or not. The orgs that get ahead of this will be running agents on donor reengagement, constituent case management, and content production well before their peers figure out what Agentforce even is.
That’s exactly the gap we help close. Hypercare exists so you have a team watching this evolve in real time and translating it into something your staff actually uses, instead of a feature buried in a release note nobody reads.
The Bottom Line
SaaS isn’t dead. It’s consolidating, and Salesforce is doing the consolidating. Every acquisition this year points toward the same destination: a platform that doesn’t just store your data but acts on it, with the governance and trust infrastructure that makes acting on it safe. The companies writing the “SaaS is dead” headlines aren’t wrong that the model is changing. They’re wrong about who’s left standing when it does.
Salesforce just spent the better part of a year proving which side of that line it’s on. Worth paying attention to if your organization runs on the platform, and worth ignoring the noise if you’ve been wondering whether now is the time to walk away from it.
It isn’t.
