Why SaaS companies must become octopuses to survive AI

Sixty-six million years ago, an asteroid wiped out 75% of species on Earth. Octopuses survived, however, because of their ability to radically adapt their biology in hours, not eons. Today, SaaS companies face their own asteroid: AI. And the octopus points the way to survival.

We see the signs in the firms that are thriving today. When Upwork Senior Vice President Dave Bottoms rebuilt the company’s AI stack, he made a counterintuitive choice. Rather than optimizing for today’s best model, he architected for disposability.

“What we think is the best model today may not be the best model tomorrow,” Bottoms explains. His team built an “optionality layer” that lets them swap AI models like changing batteries.

As Bottoms recognized, AI is evolving faster than SaaS architectural cycles, and rigid AI implementations are becoming legacy systems the moment they ship.

Design for your customers’ jobs, not your technology

The fatal mistake many SaaS companies make with AI is starting with what the technology can do rather than what customers need it to do. Aarthi Ramamurthy, chief product officer at CommerceHub (now Rithum), begins differently. “I start with empathy,” she says. “I know what the retail ecosystem goes through and all the complexity in it.”

That empathy led CommerceHub to focus AI on a deceptively simple problem: supplier onboarding. When one company calls a sweater pink and another calls it fuchsia, manual matching creates friction.

But Ramamurthy’s team didn’t just throw AI at the problem. They mapped the specific “Jobs to Be Done”—such as getting suppliers connected faster with fewer errors—and then selected the right AI approach, starting with simple algorithmic matching before layering in machine learning for demand prediction.

Contrast this with an all-too-typical approach: We’ve got LLMs, let’s find somewhere to use them. Sushma Kittali-Weidner, former chief product officer at Rheaply, frequently sees this mistake. She explains: “People are looking for magic but not thinking enough about how AI can create efficiencies in existing processes.”

Enable octopus organizations

The most valuable thing SaaS companies can do with AI is enabled by the technology, but much bigger: Helping customers distribute intelligence and authority throughout their organizations. Allow them to be like the octopus, which has two-thirds of its neural tissue outside its central brain. Octopus organizations move faster and more responsively because decisions get made closer to the frontline.

SaaS company Movable Ink’s Da Vinci platform demonstrates this enablement. CEO Vivek Sharma built a system to mass-send hyper-tailored emails by combining vision models, generative AI, insight engines and prediction algorithms.

The platform pushes sophisticated personalization decisions to frontline marketers who previously needed executive approval for campaign variations. The system determines what stories are delivered to customers, which imagery and creative are used, and when, how often and where to deliver them.

This is authority devolution at scale. Each marketer becomes vastly more capable, teaming with AI to make thousands of micro-decisions that would have been impossible under traditional hierarchies. Movable Ink’s customers can now generate hundreds of thousands of email variations where they once created one.

Within one of the world’s largest commerce networks, CommerceHub’s 2.4 billion daily transactions create similar distributed intelligence, pushing supplier matching and inventory decisions to procurement teams. CommerceHub’s AI mines poorly structured data and surfaces patterns that enable frontline employees to act without escalation.

In short, winning SaaS products help customers become octopus organizations—distributed, adaptive and intelligent at every edge.

Break your own silos

The uncomfortable truth is that most SaaS companies can’t help customers become octopuses because they’re not octopuses themselves.

The octopus has a “neural necklace,” a ring of nerve bundles that connects all its arms, enabling instant information sharing among them without involving the central brain. But SaaS companies frequently have broken connections. Just look at customer success and product teams.

Customer success teams hear about where products fail, where workflows create friction and where latent needs go unmet. Product teams have usage telemetry and performance data. When this information flows freely between teams, you create extraordinary sensing capability. But typically, these teams have separate reporting lines. They exchange sanitized summaries while critical signals vanish.

CommerceHub’s Ramamurthy addressed this by starting AI deployment on internal insights dashboards before adding it to external features. This created shared understanding across functions. When customer success, product and engineering teams access the same AI-generated insights about customer behavior, they develop a common language and aligned priorities.

Build for continuous transformation

The octopus can reconfigure its RNA in hours, adjusting biological processes faster than evolution allows. This is how it’s survived for 300 million years without external defenses. SaaS companies need to adapt at a similar speed because AI capabilities shift weekly.

Kittali-Weidner experienced this. Her team at Rheaply was resource-constrained and couldn’t afford over-engineering, so they designed modular AI implementations that could evolve without massive refactoring. The research and prototyping process that once took weeks now happens in real-time co-creation sessions. That’s a true competitive advantage.

On-demand adaptation demands a new team composition. You need engineers who embrace disposable code, product managers who ship features knowing they’ll be replaced in months, and executives willing to deprecate yesterday’s breakthroughs for tomorrow’s improvements.

Design for the 80/20 rule

SaaS companies stumble by automating too little, leaving AI as a novelty or automating too much, triggering resistance. At Upwork, Bottoms has learned that “80% of the work can be automated, but the last 20% still requires human judgment.” Upwork’s AI, for example, generates job posts and proposals, but humans make the hiring decisions.

Similarly, Movable Ink succeeds by making AI suggestions initially optional and editable. Users see value while maintaining control. Only after establishing trust does the system shift toward AI-as-default.

Adapt for the future

The octopus teaches us that survival belongs to the adaptable.

Externally, your product must help customers become octopus organizations: Distributing intelligence, devolving authority and adapting rapidly. Internally, you must become an octopus yourself: Connecting information across silos, building for continuous transformation and balancing autonomy with coordination.

The AI asteroid is already here. Become an octopus and thrive.

This article is published as part of the Foundry Expert Contributor Network.
Want to join?