AI Is the New Spreadsheet: The Malleable Expert Wins

Your company runs on spreadsheets. Let's be honest, it always has. You can point to the ERP, the CRM, the BI dashboards and all the approved systems of record. But the actual work happens in spreadsheets: the thinking, the scenario planning, the version someone created for their own understanding.

AI has just created a new spreadsheet problem. Employees and customers are spinning up AI-generated tools and agents to get their work done. People don't adopt the superior tool, or the official one. They adopt the malleable one.

This realization is keeping both enterprise leaders and SaaS providers up at night. The spreadsheet was supposed to be a relic. Instead, it quietly remained ingrained, helping to get real work done before pushing the results to the declared "source of truth" solution after the fact.

Why is that? The spreadsheet never won on capability. ERPs, CRMs, SaaS tools are definitely superior. The spreadsheet won because it easily adapts to the user, to the approach to the task at hand, to a way of thinking that no product roadmap can anticipate. It's malleable. And malleability wins way more often than we like to admit.

A New Playing Field

The pressure is building up from multiple directions at once.

Large AI labs and No-code providers are entering established markets with the ability to create solutions that used to take months and years to build.

Organizations are looking at their large subscription bills and encouraging their teams to build internal tools to lower the operating costs. The buy / build balance, stable for years, is tilting towards build.

Employees are getting more excited about the ability to quickly spin up small tools that help them in a very specific way. The super-charged spreadsheet way.

SaaS companies are seeing the writing on the wall and reinventing themselves around the change. More agentic AI capabilities within the walled garden. Shifting pricing to output or outcome-based models. Looking for new moats that are not based only on the effort to build and maintain a similar solution.

The arena now is about malleability. Yet many are trying to dismiss this trend as a passing fad. "Vibe-coding". Custom spreadsheets were the original "Vibe-ERP", "Vibe-CRM", "Vibe-BI" and many companies still run on them.

Inverting the Value Proposition

How can SaaS companies compete with AI-generated solutions? How can organizations, after encouraging their teams to create a "company-wide" AI-generated solution, compete with the hundreds and thousands of individual AI-generated solutions?

Think this way. The difference between a solution built by a SaaS company, by an internal team, or "vibe-coded" by the user is the moment it was shaped. The one made by the user is closer to their immediate need, but they all freeze after that. They don't learn from how it's being used tomorrow.

This gap between malleable-at-birth and continuous malleability is the reframing of the value proposition. The drivers behind the survival of the spreadsheet are the same behind the AI-generated solution's appeal: malleability.

Deciding what data to use, what steps to take, what decisions and outcomes will be actionable and what will not. Reflecting the ways of thinking and doing of the user.

The answer to this is not the same as adding AI features to a static product. A chatbot box in the corner, a summarize button, a few specks of AI magic dust sprinkled around.

The real solution is harder, because it inverts the value proposition. A tool that offers a solution is not the same as one that morphs into the solution the user wants. A tool that has a single right answer is not the same as one that mirrors the user's judgment and taste. A tool that follows someone else's Standard Operating Procedure is not the same as one that was coached by the user and is trusted by them.

Most SaaS roadmaps are still optimizing the tool. Most organizations are looking to replace one tool with another. People continue to look for ways to solve their problems in their own particular way. That is not a smarter tool. It is a malleable product that becomes the expert.

The malleable expert

"Expert" alone is not enough. In most domains the answer is not deterministic. There is no single right answer.

Ask five reviewers on a legal team to assess the same contract. They share the same foundation, the law, the jurisprudence. Yet the judgment sitting on top of that is a combination of preference and instinct. Where one reviewer flags risk, another sees acceptable ambiguity. Push them to agree on every detail and they will not. The disagreement is not a failure of expertise. It is the expertise.

A product that becomes the expert, then, cannot give a single fixed opinion. It has to be a malleable expert: one that starts from a credible baseline but absorbs the judgment and expertise of the specific person using it.

Picture two marketers using the same AI product to build the same campaign. If the product is genuinely a malleable expert, they will not get the same result. Over time, it will have learned each marketer's way of deciding, preferences and aesthetics, and produce a campaign closer to what that particular expert would have produced themselves. The product does not replace the marketer's judgment. It mirrors and extends it.

The baseline still matters. A malleable expert is not a clueless apprentice. It should arrive already competent, with that competence shaped during onboarding. But from that shared starting point, it diverges, becoming a different team member for every person it works with.

The opportunity

Until now, it was not technically possible to have a product continuously learn from its users.

Empowered users will increasingly spin up AI-generated tools to solve their "spreadsheet problem." Those tools fit the need at the start, but they won't keep improving as they get used.

Organizations and SaaS providers have a window of opportunity to win this brand new space.

For organizations, it's about capturing that tacit knowledge that has evaded every knowledge management initiative ever tried. And doing it before it leaks away again across every ephemeral "vibe-coded" tool their own employees are building. A company-wide team of malleable experts, each embodying the person it works with.

For SaaS companies, enabling the birthing of malleable experts is the ground left to defend. Expertise that has a solid foundation and that learns is the differentiator a weekend project cannot replicate.

The question for organizations and SaaS providers is simple, but it matters. It is not if the SaaS product has AI. Or if the organization should use AI to build its own tool. It is whether your users are already building their own spreadsheets and whether you could give them something better instead.