CloudZero

A cost number on its own doesn't tell you what to do about it. Is it high? Compared to what? Safe to act on? This week's Shipped posts address the context that turns a number into a decision: → See what actually changed, side by side. Compare Periods is now in Explorer. Pick any prior period and get the difference on the chart and in the cost table, % change included, in a view you can share with a link. → Catch a broken rule before it breaks your data. Publishing a dimension now checks every Matches pattern against the same engine that runs it, so a bad regex fails right away with a clear reason instead of quietly stalling your data for days. → Set a budget without leaving the app. Create or edit a budget and forecast in one flow on the redesigned Budgets page, with alert thresholds and a Slack or email heads-up as spend nears the line. → Get the why and the how with every recommendation. Each AWS, Azure, and GCP recommendation now arrives with the evidence behind it, the cost impact, and the next steps, so you can tell what's worth acting on without extra digging. (In private preview with our design partners.) Seeing the spend is the easy part. This week was about what comes next: what changed, whether your rules hold, where the line is, and what to do about it. See everything that shipped → https://lnkd.in/g_6xmDUR

Finance leaders don't need another AI tool pitched to them. They need to see how one of their own actually built something that works. That's the idea behind Campfire's new webinar series, "How I Built This", real finance leaders, real workflows they built themselves, no vendor pitch. First session: our own Emily Allen, Director of Finance at CloudZero, walks through how she vibe coded a CFO dashboard. Link in the comments.

Most of what companies believe about the cost of AI is wrong and the space between what they believe and what's actually true is exactly where the budget disappears. We're doing something about it. Introducing AI Myths, a weekly series where we take one assumption about AI economics and hold it up to the light. No hype or hand-waving. Just the number companies think they're spending vs. the number they actually are. Myth #1 drops right now. Follow along, and tell us which myths you want us to break. AI Myth #1: The subscription or API price is the real cost of AI. Reality: The sticker price is often a fraction of total spend. Usage scales with adoption, and without granular allocation across models and providers, the real number stays hidden until the bill lands. Has your actual AI spend ever surprised you? Drop it in the comments and follow along for myth #2 next Friday. #AIROI #EnterpriseAI

When your cloud spend moves, the first question is always: compared to what? Until now, answering that meant flipping between date ranges and trying to hold last month's numbers in your head. Compare Periods is now live in CloudZero Explorer. Put any two time periods side by side and see exactly what changed, on the chart and down to the dollar. Then share the exact view with a link. One of the most requested things our customers have asked for. Now it's here. Read more in the blog: https://lnkd.in/gBWK_rFh

We just open-sourced something we built to fix our own bad habit: paying for the smartest AI model even when the job didn't need it. Meet Model Rightsizer. Before any task runs, this scores the work on two things: -how much it actually needs a stronger model  -how much cost, speed, and volume matter for that job Then it routes the work to the right tier automatically, so nothing defaults to the biggest, most expensive model just because it's sitting there. No more using expensive Fable for menial Claude jobs, or underserving yourself by using Haiku when you’re trying to design something far more complex. We turned it on inside our own platform and the results were immediate: we cut spend on our top-tier model by 75%, almost overnight. Now it's yours. Free. Open source. No catch. Learn more about the tool: https://lnkd.in/gmUu8uEX

Check out what our very own Emily Allen built at the recent NYC Finance Hack Lab with Campfire and Replit! 🔥

Attending Ai4? Make sure to catch Erik Peterson on stage. He'll be joining four other industry leaders to dig into what's really driving AI spend and what it takes to turn it into a number you can manage, rather than one that manages you. AI Cloud Economics: Controlling Infrastructure Costs and Maximizing ROI 📍 Venetian, Las Vegas 🗓️ Wednesday, August 5, 2:00–2:45 PM PT https://lnkd.in/dBPJdtNu

A savings number tells you money is on the table. It doesn't tell you whether the finding holds up, or what to do next. That can mean recommendations pile up unactioned. When you're staring at thousands, a title and a dollar figure isn't enough to decide which ones are safe to touch. Now every recommendation in CloudZero Optimize, across AWS, Azure, and GCP, arrives with AI-generated context: the evidence behind the finding, the cost impact, and the specific steps to act on it. Faster calls, less digging, no guessing which ones are worth your time. Currently in private preview with our design partners. Want in? Talk to us about the design partner program: https://lnkd.in/e-fG4q4A #AWS #Azure #GCP #AIROI

You can't act on spend you can't see. And the numbers you can see don't do much until you can put them in the terms your team actually works in. This week CloudZero wrote about recently shipped features that close that gap: → See what Claude Code actually costs - point its built-in telemetry straight at CloudZero and get spend allocated by model and token, Anthropic cache detail intact, with no collector to install and no gateway in the path. → Allocate AWS cost by account name, not a twelve-digit ID - the name your team already uses is now a built-in dimension for CUR 2.0 accounts, with no custom lookup to build or maintain. → Take your AI cost table into your own reports - export the AI Explorer view in one click as a clean CSV, your grouping and filters included. → Create and edit an API more quickly and cleanly, with permissions grouped by what they actually do. Spend you couldn't see, named the way you already think about it, ready to take wherever you need it. See everything that shipped → https://lnkd.in/g_6xmDUR #CloudZero #ClaudeCode #AIROI

Uber ran out of its entire 2026 AI budget by April. What did they get for all that spend? Unclear:  "It's very hard to draw a line" between AI usage and outcomes, Uber’s COO said on an earnings call. Connecting AI spend to outcomes is the CFO problem of 2026, and it isn't really a technology problem. It's an intelligence problem. We surveyed 260 finance leaders, more than half of them CFOs. Only 22% can tie AI spend to business outcomes today. 87% say they need to within the year. Boards aren't waiting: two-thirds now gate further AI funding on proof of return. Every CFO running a real AI deployment should be able to answer three questions right now: 1. What did we spend? Not the month-end bill, but the breakdown by team, product, customer, and workflow. 2. Was it worth it? Spend tied to outcomes, not activity. Tokens and usage rates are inputs, not returns. 3. Where do we invest next? You can't answer this one if you can't answer the first two. Experimentation is over and the accountability phase is here. The companies that pull ahead won't be the ones spending the most on AI. They'll be the ones who can see the unit economics clearly enough to invest with confidence. New on the CloudZero blog from David Aponovich: https://lnkd.in/gaGVrZ_y #AI #AISpend #AIROI #CFO