I'm going to say something that will upset the entire FinOps tooling industry: visibility is not a strategy. It's a starting point that most companies mistake for a finish line.
You know the playbook. Buy a cost management tool. Connect it to your billing API. Build some dashboards. Schedule a weekly report. Invite the VPs to a monthly review meeting. And then... watch the bill go up anyway.
This happens at almost every company I've worked with. They have perfect visibility. They can tell you, down to the resource level, what every team spent last month. They have trend lines, anomaly alerts, forecasts that would make a data scientist nod approvingly. They can slice by service, region, team, environment, and tag.
And their cloud bill is still growing 30% year-over-year with no corresponding increase in revenue.
Here's the dirty secret nobody in FinOps wants to admit: the correlation between dashboard quality and actual cost savings is approximately zero.
I ran an informal survey of 40 companies in my network last year. The ones with the most sophisticated dashboards — real-time views, custom cost allocation models, predictive anomaly detection, the works — had virtually identical savings rates to companies running off the native AWS Cost Explorer with a spreadsheet on the side.
The companies actually saving money? They had one thing in common: someone with the authority and incentive to act.
Not the authority to see. The authority to do. There's a canyon between those two things, and most FinOps teams are standing on the wrong side of it.
Let me paint a picture that'll feel painfully familiar.
Your FinOps team spots a cluster of oversized RDS instances. They're running db.r6g.4xlarge across 12 production databases, averaging 8% CPU utilization. That's roughly $23,000/month in wasted capacity. Clear as day on the dashboard.
They open a ticket. Tag the database team. Include the data, the recommendation, the projected savings. Textbook FinOps.
Here's what happens next: nothing.
The database team is three weeks into a migration project. The ticket sits in the backlog. When someone finally looks at it, they push back — "we need the headroom for quarterly peaks." No data to support that claim, but nobody's going to argue. The FinOps team doesn't own the databases. They can see the waste, but they can't stop it.
Multiply this by every team in your org. Multiply it by every resource type. That's your savings pipeline: a graveyard of well-documented recommendations that nobody acts on.
The FinOps Foundation's own data shows that the average recommendation acceptance rate across the industry is somewhere around 15-20%. That means 80% of identified savings opportunities die in someone's backlog.
Think about that. Companies spend $50K-$200K/year on cost management tooling that generates recommendations, and four out of five of those recommendations evaporate.
The problem isn't the recommendations. They're usually right. The problem is the operating model. You've built a system where one team sees the problem and a different team has to fix it. That's not cost management — that's an expensive suggestion box.
And the suggestions keep coming. Every month, new anomalies, new rightsizing opportunities, new RI coverage gaps. The dashboard dutifully displays them all. The acceptance rate stays at 15%. And the FinOps team writes a monthly report explaining why the projected savings didn't materialize.
I've seen this cycle continue for years at some organizations.
The companies I've seen genuinely bend their cost curve share a pattern, and it's not a tool or a framework. It's an operating model change.
1. Engineers own their costs, not a central team.The FinOps team's job isn't to find savings. It's to give engineering teams the data, the context, and the guardrails to manage their own spend. When the team that writes the code also sees the bill, the feedback loop closes. When a separate team sends recommendations to a Jira board, the feedback loop dies.
2. Cost has consequences.At one company I worked with, team budgets were real. Not "aspirational targets" — actual budgets that, if exceeded, came out of the team's headcount allocation the following quarter. You know what happened? Engineers started checking their cost dashboards voluntarily. They rightsized without being asked. They terminated idle resources the same day they stopped using them. Savings rate: 34% in six months.
At another company, cost was a "shared concern" with no consequences for overruns. Savings rate after a year of FinOps investment: 4%.
Same tools. Same dashboards. Radically different outcomes.
3. Automation beats recommendations.Stop generating tickets. Start writing policies. If an instance runs below 5% CPU for 14 days, it gets flagged for automatic downsize with a 48-hour grace period. If a dev environment is running at 3 AM, it gets shut down unless there's an active exemption.
The best savings aren't found — they're enforced. Every recommendation that requires a human to read a ticket, context-switch, evaluate risk, and manually click buttons is a recommendation that probably won't happen.
Here's what I want you to take away: if your FinOps strategy starts and ends with visibility, you don't have a strategy. You have a monitoring hobby.
Visibility is the floor, not the ceiling. It's necessary and wildly insufficient. The hard work — the work that actually changes the number on the bill — is the organizational plumbing: who owns what, who's accountable, what happens when someone ignores a recommendation, and what gets automated so humans don't have to make the same obvious decision 500 times.
The FinOps tooling market is worth $5 billion and growing. Most of that money is being spent on increasingly sophisticated ways to look at cloud costs. Very little of it is being spent on increasingly effective ways to change them.
If you're evaluating your FinOps program right now, don't ask "can we see our costs?" You almost certainly can. Ask instead: "When we see waste, what structurally guarantees it gets fixed?"
If you can't answer that question, your dashboards are just expensive screensavers.
---
Want to figure out where your FinOps program is actually stuck? Book a free 30-minute assessment — no pitch, just a clear-eyed look at what's working and what's theater.
