"Spending less is not the goal. Spending better is."
I have seen many companies treat FinOps as a simple synonym for slashing technology costs. Shutting down staging machines on Fridays, squeezing CPU limits until the service chokes, and demanding a justification for every single cent of storage. This is the shallow and defensive version of the discipline. The mature version, which is redefining technology management through 2030, is much more strategic.
The logical absurdity of zero cost
If you take the logic of "our goal is to spend the absolute minimum possible in the cloud" to its ultimate consequences, the technical solution is extremely easy: just log into the provider console this afternoon and shut down absolutely every single server. There you go: next month's bill will come in at zero. Problem solved, right?
Obviously not, because along with zero cost will come zero revenue. The company needs to operate, the product needs to respond in milliseconds, and value needs to be delivered to the customer.
This is precisely why the FinOps Foundation itself defines the discipline not as "saving money," but as maximizing the business value of the cloud and technology. Cutting raw waste (idle machines, orphaned disks, poorly sized reservations) is merely the basic hygiene of the first week. The real game begins after you clean up the trash.
Unit Economics and the Jevons Paradox in the cloud
When we move away from the shortsighted accounting perspective, we realize that the metric that truly matters to a CTO or CFO is not the total amount of the cloud bill at the end of the month, but how much the company spends per unit of value delivered (Cloud Unit Economics):
- Infrastructure cost per completed transaction;
- Monthly cost to serve each active customer (Cost to Serve);
- Compute and inference cost per delivered feature;
- Proportion of cloud spending over gross revenue (Cloud-to-Revenue Ratio).
This distinction gains even more strength when we remember a classic law of economics formulated in 1865 by the British economist William Stanley Jevons, known as the Jevons Paradox. While studying steam engines in the nineteenth century, Jevons proved that when engineering makes the use of a resource much more efficient and cheaper per unit, the total consumption of that resource in the economy does not fall: it increases, because unit efficiency makes new products viable and attracts much more demand.
In cloud and artificial intelligence, the Jevons Paradox happens every single day. If your total cloud bill went up 20% this quarter, but this happened because your team optimized the architecture, cut the unit cost per order in half, and allowed the paying customer base to grow 60%, the bill increase is extraordinary news for the company balance sheet. On the other hand, if you cut the cloud bill by 15%, but system sluggishness dropped sales conversion by 30%, that "savings" destroyed value.
Where I would bet in the coming years
Looking at the data gathered annually in the State of FinOps report and at what I see in the trenches of operations, three movements will separate mature companies by 2030:
1. Direct integration with business OKRs
FinOps will cease to be an isolated report from the infrastructure team and will integrate the strategic objectives of the company. Relevant architecture investments must be born with a clear value hypothesis and margin tracking. Of course, not every shared expense (such as structural security or base network) can be mathematically attributed to a single revenue line, but the general direction must respond to the commercial strategy.
2. Predictive automation with safety guardrails
Recommendation and automatic resizing tools greatly help contain anomalies in real time, especially with elastic workloads and artificial intelligence API calls. However, financial automation can never operate blind: it must respect strict performance limits (SLAs) and allow engineering review before applying risky changes in production.
3. Feature-level Showback
Most companies today can barely split the cloud bill by department or by squad. The next level of maturity is knowing the live cost per product feature. When the product team and engineering see on the same screen how much it costs to keep that specific feature running versus how many customers actually use it, difficult decisions to deactivate zombie features or refactor heavy queries become obvious.
The strategic role: explaining technology in the CFO language
In recent years, the FinOps Foundation officially expanded its scopes (FinOps Scopes) beyond public cloud: today the same discipline of unit economics applies to SaaS licenses, databases, observability platforms, and Artificial Intelligence costs. Expenses involving human teams require their own analysis to avoid falling into simplistic calculations, but the total cost of the digital ecosystem must be transparent.
The engineer, architect, or technical leader who masters this translation will have a guaranteed seat at the table where the company budget is decided. It stops being just "IT asking for more machines for IT" and starts being engineering demonstrating financial leverage in the language of the board.
My final provocation
Do the quick test in your operation today: does your company know how to answer, with reliable data, how much it costs in infrastructure to acquire and serve a new customer? Do you know how much it costs to keep standing that legacy system no one wants to touch? Do you know how much money is burned every month by that product module that less than 1% of users access?
If the answer is still "I don't know," you have an efficiency mine waiting to be worked. And mastering this calculation is the true future of the FinOps practice.
Continue deepening your knowledge: also read the Complete FinOps Guide for Companies and A Good Metric is Not Just for Slides. Also test the Tools and Executive ROI Calculator and get to know the practice of FinOps and Cloud Efficiency.
References and verifiable readings
- FinOps Foundation. FinOps Framework & Cloud Unit Economics Capabilities. Linux Foundation.
- FinOps Foundation. State of FinOps Annual Report & Data. Linux Foundation.
- Jevons, W. S. (1865). The Coal Question; An Inquiry Concerning the Progress of the Nation, and the Probable Exhaustion of Our Coal-Mines / Storrier, J. R., & Mentis, M. (2023). FinOps in the Age of AI and the Jevons Paradox.
Good ideas keep the conversation going.
Share this article or publish the visual deck to your Stories and Feed.

Cesar Zanis
15+ years leading software architecture, database performance, FinOps, and applied AI across enterprise operations. Built and scaled an engineering organization from zero to 20+ developers.
Enjoyed this article?
I publish practical analyses on critical systems architecture, Cloud FinOps, data, AI, and engineering leadership directly on LinkedIn.
Continue Reading
Take this analysis further with data and sources
All visual cards of this article are ready above to share on your Stories, Feed, WhatsApp, or LinkedIn with full sources.



