Cloud conference season is here. Woohoo! Microsoft Ignite and AWS re:Invent sandwich Thanksgiving. I’ll be attending re:Invent (send me a note if you want to connect) and I’m working on Ignite.

A week at one of these costs bank. Passes, flights, hotel, and a week of not doing the other work. I'm not saying you (or I) shouldn’t go. Meeting people in person is hard to replace, and I have been fortunate to meet great people and companies when attending. But the cash burn has to give something back to the business. New customers. Better support for the ones you already have. A technique or skill that you bring home and use.

Here's the problem. Most people can tell you what the trip cost. Very few can tell you or will spend the time to determine what it returned, because nobody wrote down what it was supposed to return before and calculate after.

That's a value problem in a small package. The cost shows up on its own. The value only shows up if someone decided in advance what to look for.

For the last few issues I've been writing about the savings side of the client work. Tags, owners, orphaned disks, the checks behind the agents. All of that matters, and right now savings is the only question that is being asked by the executives at the company I’m advising. I know it won't last.

A savings number tells them what can be cut. Sooner or later they'll ask what the technology is helping provide (Value), and the same spreadsheet can't answer it.

Here's a real life example. Azure Site Recovery disks were in my first list of waste for orphaned disks. They look orphaned: nothing attached, no obvious owner. They're also the replicas that bring the business back after an outage. Delete them and you get savings and no recovery plan (not ideal). That would have been bad. The checks caught it, but only because I knew to look. Nothing in the environment said what those disks were for.

A cost with no record of what it supports (or good tagging) looks like waste every time. And the more aggressive the savings target, the more of that you'll cut by accident.

FinOps is one of the few functions that brings spend by owner, sits between engineering, finance, and IT, and gets asked the awkward questions. The FinOps Foundation lists business value driving decisions as a core principle. Most programs I see practice the cost half of that sentence and leave the value half for someone else.

Ramp published its October AI Index on October 14. Using September data from Ramp customers, 55.9% of businesses now pay for AI. Anthropic sits at 43.8% and OpenAI at 40.6%. Token spend has barely moved for two months, while total spend hit a new peak in early October, 1.5% above August.

Ramp's economist, Ara Kharazian, reads this as the top of the S-curve, with falling model prices doing much of the work in holding spend flat. That's a fair read. But this fails to ask a harder question?

Here's my opinion. The flattening is what it looks like when organizations stop buying on faith and start asking whether the value is there. The first wave was buy it and see. The next one is finance asking what it bought. Ramp's piece doesn't claim to answer that, and it can't. Spend data from the outside only sees the purchase. It can't see what the purchase did inside your company.

If prices keep falling and spend stays flat, usage is going up. Is more usage more value? Only you can answer it for your company, and only if you focus on value.

Write the return (aka Value) before you spend. Conference, AI seats, a cloud commitment. One sentence on what it should return, who checks, and when. For a conference, that might be "two customer conversations, one partner intro, and one technique we try in the next 30 days." Then check it after you get back to the office. Same for AI, cloud resources, architecture … what is the Value return on investing in these areas. Common sense, so try it.

Pick one unit the business already uses and put the cost next to it. Cost per order, per claim, per release, per ticket. You don't need a perfect unit. You need one that someone in finance would recognize. If the cost moves and the unit doesn't, that's a conversation. If both move together, that's the start of a value story.

Record / Tag / Identify what each major resource supports, along with the owner because the owner will tell you who to call. its purpose will tell you whether it's waste. This info that would have helped flag the Site Recovery disks before I had to catch and remove.

Remember - an agent can find the waste. It still can't tell you who owns it (unless there is metadata there)

How have you shown the value of technology spend to a CFO/CIO/CTO? Not savings, value. I'd like to hear what worked and what got ignored. Send it to [email protected] and I'll respond to every inbound that I receive.

And if you're headed to re:Invent or Ignite grab time with me here.

Written with the help of AI. All the ideas expressed are mine and mine alone.

If you would like your company included in the Spotlight, contact the CloudXray AI Team

Company: CloudXray AI

Category: Managed Services & Consulting

What They Do: FinOps Consulting & Advisory Services; Owners & maintainers of the single largest FinOps company directory (finops.cloudxray.ai)

Why It Matters: Companies still need guidance on implementing FinOps and understanding the landscape of companies that exist