Bain says agentic AI could reshape SaaS, not replace it

Agentic AI is not killing SaaS, it is changing what buyers will pay for

Bain’s core warning is simple: software companies can no longer assume the old seat-based playbook will carry them through the AI era. For monday.com, that means the real question is not whether agents arrive, but which parts of the work flow they should own, which parts should stay human-led, and how much value the product can prove once the busywork is stripped away.

The shift Bain is describing

Bain says SaaS first emerged 25 years ago when software moved to the cloud and feature delivery sped up. That model made it easier to ship, easier to adopt, and easier to scale. Now Bain says a new discontinuity is forming, one that forces leaders to decide where AI enhances their offerings and where it may replace portions of the workflow.

Its prescription is blunt: own the data, lead on standards, and price for outcomes rather than log-ons. That is a direct challenge to the assumption that more users automatically means more value. In an agentic world, buyers will care less about how many licenses a company consumes and more about how much work gets completed without human hand-holding.

Bain’s newer research goes even further. It says agentic AI could create a $100 billion U.S. SaaS market opportunity by automating cross-system coordination work. That matters because coordination is where a lot of modern SaaS still leaks value. The product may handle one task well, but the customer still has to move information across apps, chase approvals, and reconcile the work manually.

Why that matters inside monday.com

For monday.com, this is not abstract strategy language. The company’s business sits directly in the path between tasks, tools, and teams, which is exactly where agentic AI can create or destroy value. Product teams now have to decide which capabilities should feel native to the core platform, which should be delegated to agents, and which should remain under human control because the risk is too high to automate fully.

Engineers should hear Bain’s thesis as a warning about context. If an agent is going to take action safely, it needs the right permissions, the right data, and the right understanding of what happened before. That means the product’s real moat may not be the interface alone, but the quality of the workflow data, governance, and standards that sit underneath it.

Sales teams face a different shift. Buyers are already asking whether software should be priced by usage, by outcome, or by some combination of the two. That is not a side question. If Monday.com cannot show a clear line from AI features to measurable business results, the pricing conversation moves away from value and toward discount pressure.

Bain’s bigger point is that AI is not only a threat to software revenue. It is also an opportunity to redefine the category around completed work. If monday.com can make its AI features reduce complexity instead of adding another layer of tools to manage, it can turn that pressure into a selling point.

Monday.com’s AI strategy is already built around that bet

On February 10, 2025, monday.com said its AI Vision would rest on three pillars: AI Blocks, Product Power-ups, and the Digital Workforce. The company said the goal was to help SMBs and mid-market firms scale without increasing resources, while also helping enterprise and Fortune 500 customers speed up processes that get bogged down as organizations grow.

Daniel Lereya, the company’s chief product and technology officer, framed the strategy in language that fits the Bain thesis neatly. He said monday.com’s goal was to "democratize the power of software" and that "people adopt products, not technology." That is the right lens for agentic AI: customers do not want a parade of impressive models, they want work to move faster with less friction.

By September 17, 2025, at Elevate 2025, monday.com said it had gone further with monday agents, monday magic, monday vibe, monday sidekick, and monday campaigns. The key product signal was monday agents, which the company said can orchestrate multi-step processes and execute tasks end to end. That matters because it moves AI from a feature that suggests or summarizes into a system that can actually coordinate work.

That is the line investors, operators, and buyers should watch. A point solution that drafts text is useful. A system that can move work across steps, tools, and approvals is more disruptive because it begins to absorb the labor of coordination itself.

The financial signals already point to a company moving upmarket

The business momentum around monday.com suggests this is not just a product story, but a revenue-model story too. In its 2024 results, the company reported revenue of $972.0 million, up 33% year over year, and said it had surpassed $1 billion in annual recurring revenue. In its 2025 results, it reported revenue of $1.232 billion, up 27% year over year.

Those numbers matter because they show monday.com expanding even before the full monetization of agentic AI is settled. The company also said customers with more than $50,000 in ARR represented 41% of total ARR as of December 31, 2025. That is a sign of deeper penetration into larger accounts, which usually comes with higher expectations around governance, reliability, and integration.

Retention metrics point in the same direction. Monday.com reported net dollar retention of 112% as of December 31, 2024, and 110% as of March 31, 2026. In plain terms, existing customers are still expanding enough to offset churn pressure, which gives the company more room to experiment with AI packaging without relying entirely on new-logo growth.

What employees should take from this

For product managers, the important lesson is that AI features cannot just sit on top of the workflow. They have to change the workflow itself. If monday.com wants agents to matter, they have to remove steps, not merely decorate them.

For engineers, the challenge is to build the rails that make automation trustworthy. That means context, permissions, observability, and a data layer strong enough to support agent decisions across systems. The more monday.com becomes the place where work is orchestrated, the more critical that foundation becomes.

For sales teams, the message is that customers will increasingly ask for proof, not promises. They will want to know whether an agent saves hours, reduces errors, accelerates approvals, or eliminates the need for another tool entirely. That is where outcome-based pricing will start to feel less theoretical and more unavoidable.

Bain’s thesis does not say SaaS disappears. It says the companies that win will be the ones that make AI visibly useful, keep control of the data and standards that power the product, and prove that the software removes work instead of adding to it. For monday.com, that is less a threat than a test of whether its work-OS can become the system that actually gets work done.