The Landlord Fallacy
Why calling your ecosystem "parasites" is a confession, not a strategy
On Workday’s Q4 earnings call, Aneel Bhusri said the quiet part out loud.
An analyst asked how Workday plans to monetize third-party tools built on its data. Bhusri’s answer: “Think of us as an evolving layer on top of hyperscalers. In the same way that they charge for consumption of compute cycles, we’re going to continue to flex that muscle. There are some vendors out there, including some of our peers, that I would consider them, at some level, parasites on Workday. They get a free ride on our underlying system of record, and we’re going to put an end to that.”
Parasites. That is the word a $70B enterprise software company used to describe the tools that help its customers get work done.
This is not a Workday problem. This is a pattern. And it reveals a fundamental misunderstanding of where value is moving in enterprise software.
What the Business Model Reveals
To understand why Bhusri said this, you need to understand what is happening to Workday’s business model. The company has gone through four chapters:
Chapter 1 (2005): Cloud HR and finance. Revolutionary idea. Correct bet.
Chapter 2: Hyper-growth. Per-seat licensing. You pay based on how many employees are in the system. Simple. Predictable. Beautiful gross margins.
Chapter 3 (2023): Operational excellence. Margin expansion. Stock went sideways.
Chapter 4 (now): The AI pivot. And this is where it gets interesting.
The problem with per-seat pricing in a world of digital workers is obvious. If AI replaces tasks that humans used to do, and you charge per human in the system, your revenue shrinks. Workday’s stock is down 30% in the past year partly because the market figured this out.
So Workday introduced Flex Credits. A consumption-based model where customers pay for the AI capabilities they use, not the headcount they manage. This is the right instinct. It decouples revenue from the workforce it was designed to track.
But here is the twist. Bhusri does not just want to sell Workday’s own AI. He wants to tax everyone else’s.
Gerrit Kazmaier, Workday’s President of Product, laid out the new pricing tiers on the same call. Customers can subscribe to Workday’s applications (traditional pricing), consume raw APIs (pay-as-you-go), access “data context” (mid-tier), or buy Workday’s premium “agent APIs” that aggregate large chunks of work (top-tier pricing).
Kazmaier said their premium APIs “have a premium price tag because they complete meaningful work. They are not just a simple SOAP or REST API.”
Read that again. Workday wants premium pricing because its tools complete meaningful work. The implication is that everything else accessing the platform is doing something less meaningful. Something parasitic.
The Hyperscaler Analogy Breaks Down
Bhusri’s core analogy is that Workday should function like AWS or Azure. If you run compute on their infrastructure, you pay for consumption. If you run digital workers on Workday’s data, you should pay for consumption too.
It sounds logical. But there is a fundamental difference.
AWS provides compute. Compute is fungible. You can get it from AWS, Azure, GCP, or a thousand other providers. What makes AWS valuable is not that it has unique compute. It is the ecosystem, the tooling, the developer adoption. AWS earns its toll by making it easy to build.
Workday provides a data schema and a set of business rules. That is not fungible. It is proprietary. And it is not valuable because it is easy to build on. It is valuable because it holds the customer’s data hostage. The switching costs are the moat, not the developer experience.
When AWS raises prices, customers have alternatives. When Workday raises prices on API access, customers are stuck. That is not the hyperscaler model. That is the landlord model.
What the Goldman Analyst Got Right
The sharpest question on the call came from Gabriela Borges at Goldman Sachs. She asked, essentially: what if the intelligence layer gets built outside of Workday? What if vendors or customers build solutions next to Workday that leverage all the domain experience you have built, but the incremental value accrues in the intelligence layer, not in Workday itself?
This is the right question. And Bhusri’s answer was telling. He pointed to the API layer. To consumption metering. To tiered pricing. His answer to “what if the value moves outside your platform” was “we will charge a toll on the way out.”
That is a defensive strategy, not a growth strategy. It tells you the CEO views the company’s competitive advantage as data custody, not data utility.
From System of Record to Tollbooth
Here is the pattern playing out across enterprise software. Not just Workday. Every incumbent system of record is facing the same question.
For twenty years, owning the system of record was the game. If you held the data, you held the customer. The strategy was simple. Make switching costs unbearable. Tax every integration. Extract rent from your installed base.
This worked when the system of record was the place where value was created. A human logged in, did their work, logged out. The platform and the work were inseparable.
A digital workforce breaks that coupling. AI workers log in, extract the data they need, execute work across multiple systems, and write results back. The system of record becomes a data source. Important, but not where the work happens.
The value has shifted from “who holds the data” to “who uses the data to complete a task.” A database that stores an employee’s benefits elections is a cost center. A digital worker that processes a benefits change, confirms it across three systems, and notifies the employee is a value multiplier.
Calling the companies that build that capability parasites is like a library calling its readers freeloaders.
The Real Test
Every enterprise platform CEO should ask a simple question: if you removed the “parasite” from the equation, would the customer’s life get better or worse?
Remove the digital workforce that automates month-end close, hiring workflows, and account transfers. What happens? The customer loses capacity. Work slows down. They need more humans to do what the digital workers were doing.
Now flip it. Remove the system of record and replace it with a different one that exposes the same data. The digital workforce keeps working. The customer barely notices.
That tells you where value is being created and where it is being rented.
The Right Response
Bhusri is not wrong that Workday has incredible assets. 1.7 billion AI actions last year. 75 million users. Over a trillion annual transactions. 97% gross retention. Those are real.
But the right response to an agentic world is not to meter access to your data. It is to become the place where digital workers want to work.
Make the APIs so good that every builder chooses your platform as the execution layer. Make the data so accessible that the ecosystem builds on you, not around you. Make the switching costs about quality, not lock-in.
Compete on being the best system of action, not the most expensive system of record.
Workday’s $1.1B Sana acquisition suggests they understand this partially. Sana extends Workday’s reach into Gmail, Outlook, Google Drive, Salesforce. That is the right instinct. Go where the work is.
But you cannot simultaneously build your own tools that span multiple systems and call everyone else parasites for doing the same thing. One of those positions has to lose.
The Broader Pattern
This is not really about Workday. It is about every system of record in every industry facing the same fork in the road.
In wealth management, where I spend my time, the same dynamics are at play. Custodians hold the data. CRMs hold the client relationships. Financial planning tools hold the models. For decades, the power sat with whoever owned the record.
Now digital workers move across all of those systems. They open accounts on custodial platforms, update CRMs, generate plans, and file paperwork. The value is in the orchestration of work across systems, not in any single system’s database.
The platforms that recognize this and make themselves easy to work inside are gaining share. The ones building tollbooths are training their customers to look for alternatives.
When a platform CEO calls his ecosystem parasites, he is telling you three things. He sees his competitive advantage as data custody, not data utility. He plans to monetize through access restrictions, not value creation. He views his customers’ productivity gains as a threat to his pricing model.
Every one of those is a losing position in a world where digital workers do the work.
The future belongs to the platforms that make it easy to get work done. Not the ones that make it expensive to try.