Latest / SaaS Business with Fexingo: Software-as-a-Service Companies, ARR, and Recurring Revenue / How SaaS Companies Monetize Data Without Churn
Transcript
- Lucas: The real story in SaaS right now isn't about selling more seats or locking customers into longer contracts. It is about selling the data that lives inside the software. Think about Snowflake. They moved from pure storage fees to a model where you pay for the compute power needed to query and transform that data. It feels like a small shift, but it changes the entire value proposition. Luna: It shifts the risk from the buyer to the seller, doesn't it? If the customer gets less value out of their queries, they spend less, but the provider has to be smarter about efficiency. Lucas: Exactly. And this is happening across the board. Look at Palantir’s AIP platform. They aren't just charging for access; they are charging for the outcome of the AI inference. The customer pays for the computation that drives the decision. This is usage-based pricing on steroids, but tied directly to business value rather than arbitrary API calls. Luna: So we are moving from a licensing model to a utility model, almost like electricity. You only pay for what you burn when you solve a problem. Lucas: That is the perfect analogy. And here is the kicker: it actually reduces churn. When you sell a seat, the customer cancels if one person leaves. When you sell data insights, the cancellation requires them to rip out their entire analytics pipeline. The switching cost becomes astronomical because your data is already being used to train their internal models. Luna: But does that work for smaller companies? Or is this only viable for enterprise clients who have massive data lakes? Lucas: Good question. For mid-market companies, it works if the data product is indispensable. Take a company like Databricks. They offer a unified workspace where you can prepare data, build models, and deploy applications. The usage fees come from the compute required to run those jobs. If a marketing team uses it to segment audiences in real time, they cannot easily leave because the integration is too deep. Luna: It sounds like the moat is no longer the interface, but the dataset itself. That seems risky if the data quality drops. Lucas: It is risky, which is why data governance is becoming a huge part of the sales cycle. Customers are asking not just 'how much does it cost?' but 'is this data accurate and compliant?' In 2026, with stricter privacy regulations in Europe and California, the ability to prove compliance within the compute environment is a premium feature. Luna: So you are selling trust as a service. That is a clever way to frame it. Lucas: Precisely. And this ties back to our earlier discussions on pricing traps. Many companies tried usage-based pricing by metering clicks or logins. It failed because it was too granular and annoying. But metering actual business outcomes, like a successful prediction or a processed transaction, aligns incentives perfectly. Luna: I wonder if this will lead to more hybrid models. Keep the base subscription for access, then layer on usage fees for heavy lifting. Lucas: That is exactly what is happening. Salesforce did this with Einstein AI. You pay for the CRM platform, but every time you run an advanced forecast or generate a summary, you dip into a credit pool. It keeps the entry price predictable while capturing upside when the customer is actually getting value. Luna: Does that mean finance teams are going to hate these invoices? Variable costs are a nightmare for forecasting. Lucas: They do hate it initially. But most SaaS vendors now offer caps or tiers. You might get the first ten million compute credits free, then pay a lower rate after that. It gives the CFO visibility while allowing for growth. It is a compromise between predictability and flexibility. Luna: Smart. So the vendor absorbs some of the volatility, but still captures scale. Lucas: Right. And let's talk about the technology enabling this. Secure compute environments are key. You need to process sensitive financial or health data without ever exposing the raw information. Technologies like homomorphic encryption allow computations on encrypted data. It is expensive, but necessary for high-value verticals. Luna: Homomorphic encryption? That sounds like sci-fi. Is it actually practical yet? Lucas: It is niche but growing. Mostly used by banks and healthcare providers. For most other industries, simple secure enclaves on cloud infrastructure are enough. Intel and AMD both have hardware-level security features that isolate data during processing. It is not magic, but it is good enough for ninety percent of use cases. Luna: Okay, so the tech is there. But how do you convince a legacy manufacturer to switch from Excel to a cloud data platform? Lucas: You don't start with the platform. You start with a single painful problem. Maybe it is reconciling inventory across three different warehouses. You show them how the data platform solves that one thing faster than anyone else. Once they see the speed, they expand to other departments. Luna: Land and expand. Classic playbook, but applied to data workflows instead of user accounts. Lucas: Exactly. And the expansion is natural because the data stays. You add supply chain analytics, then demand forecasting, then supplier risk assessment. Each new module consumes more compute, increasing the ARR without needing to find new customers. Luna: It feels like the ultimate lock-in. But isn't there a risk of vendor lock-in backlash? Lucas: There is, but open standards mitigate it. Most modern data platforms support SQL, which is universal. If a customer wants to leave, they can export their queries and move to another provider. The friction is in the data migration, not the syntax. That friction is your defense. Luna: So the barrier is the effort of moving terabytes of history, not the tool itself. Lucas: Correct. And that brings us to the donation segment. If these conversations are useful for what you're building or running, consider supporting the show. Honestly, if today was worth a coffee to you, that's the link — buy me a coffee dot com slash fexingo. It helps keep these deep dives ad-free and independent. Luna: Yeah, it really helps. Thanks for listening and for keeping the signal to noise ratio high. Lucas: Let's go back to the numbers. We mentioned Snowflake and Palantir, but what about the margins? Usage-based models can hurt gross margins if compute costs rise faster than revenue. Luna: That is a valid concern. If AWS raises prices, the SaaS provider gets squeezed unless they pass it on. Lucas: They usually do. But smart vendors optimize their own infrastructure. Some, like Databricks, are building proprietary chipsets to lower inference costs. Others negotiate volume discounts with cloud providers. The goal is to keep the spread between what you charge and what you pay stable. Luna: So efficiency becomes a competitive advantage. The cheaper you can process data, the more margin you keep. Lucas: Absolutely. And this drives innovation. Companies are investing heavily in compression algorithms and sparse models. Reducing the size of the data packet saves money on every query. It is a technical arms race disguised as a pricing strategy. Luna: I never thought of compression as a revenue driver. That is a sharp insight. Lucas: It is. And it forces engineering teams to think about economics, not just functionality. If you write inefficient code, you literally bleed money. That aligns engineering incentives with business goals. Luna: Does this mean junior engineers are going to need more finance training? Lucas: Probably. Or at least a better understanding of cost per query metrics. We are seeing job descriptions change. 'Optimize for performance' is now 'Optimize for performance and cost.' It is a subtle but significant shift in culture. Luna: So the definition of a good engineer is evolving. Efficiency is king. Lucas: Yes. And this trickles up to the C-suite. CFOs are now involved in architecture reviews. They want to know the marginal cost of a new feature before it ships. It makes product development more disciplined. Luna: I like that. Less waste, more focus. It seems like a healthier way to grow than just throwing money at user acquisition. Lucas: It is. User acquisition is getting expensive. Organic growth through data network effects is cheaper and stickier. If my data helps your data, we both win. It is a positive-sum game. Luna: Network effects in data. That is the holy grail for many platforms. Lucas: Right. And once you hit critical mass, the insights become so rich that leaving feels impossible. You aren't just losing software; you are losing intelligence. Luna: That is a powerful argument. Intelligence is harder to replace than a login credential. Lucas: Exactly. So the future of SaaS is less about access and more about insight. The companies that master this balance between cost, speed, and accuracy will dominate the next decade. Luna: It makes you wonder which current tools will fail to adapt. There must be plenty riding on old licensing models. Lucas: There are. Legacy ERP systems are struggling with this transition. They have millions of users but lack the modern data stack to provide real-time insights. They are at risk of being disrupted by agile newcomers who understand the value of data. Luna: So disruption is coming from below, not just from above. Smaller players eating the lunch of giants. Lucas: Precisely. And it starts with a single use case. One department adopting a new tool, then expanding. It is a slow bleed for the incumbents. Luna: A slow bleed, but fatal in the end. I guess vigilance is key for long-standing businesses. Lucas: Vigilance and adaptation. The market rewards those who listen to the data, not just the dashboard. What do you think is the next frontier beyond compute? Luna: Maybe data ownership. Who owns the insights generated by the platform? That could be the next battleground. Lucas: That is a fascinating angle. If I generate a trend using your data, do I own that trend? Or do you? It touches on copyright, privacy, and fair use. It is going to be messy. Luna: Messy, but lucrative. Whoever solves that legal puzzle will set the standard. Lucas: Agreed. We will have to watch that space closely. Thanks for joining me, Luna. Luna: My pleasure. Let's see how this plays out in the next quarter.