Shannon Eldridge-Kuehn is a Principal Solution Architect at AHEAD, where she has been the FinOps voice on cloud conversations totalling over half a billion dollars of spend. Her career spans internal IT, managed services, a boutique Azure systems integrator, and five and a half years at Microsoft.
Shannon Eldridge-Kuehn became technical because her DJ mixer would not speak to her laptop. She wanted to be booked at the clubs in her hometown of Lincoln, Nebraska. The gear she had bought with her own money would not behave once a computer was in the loop. Her developer friends were tired of telling her to reboot the machine. So she taught herself enough computing to keep the set running. That problem was the first one she ever solved with code. Twenty years later it is still the shape of the work she does.
Today Shannon is a Principal Solution Architect at AHEAD, where she has been the FinOps voice on cloud conversations totalling over half a billion dollars of spend. Before that she spent five and a half years at Microsoft, two years at a boutique Azure systems integrator, a year at a managed services provider, and eight and a half years in internal IT. The throughline is the same as the night she rewired the mixer. The job is bridging two systems that do not naturally talk to each other. Today the two systems are FP&A and engineering.
Shannon argues that any FinOps program staffed entirely on one side or the other is structurally broken from the start. Hand it to procurement and the team loses the technical ingenuity that operates the cost levers. Hand it to engineering and the team loses the business context that makes the levers worth pulling. The hybrid team is not a preference. It is the only configuration that survives contact with a cloud bill. Once the team is in place, the job is not a quick win. It is a long set: care, feeding, governance, and automation, sustained over years.
Reading the Room: Trust Before the Spreadsheet
Before Shannon can argue for a hybrid team, she has to be in the room. The work begins with stakeholders who are usually apprehensive. They are often outranked by the people they need to convince. They have been burned by previous attempts. FinOps does not start with a spreadsheet. It starts with whether the person on the other side of the table is willing to trust the person making the case.
Q: A lot of FinOps consulting starts with the data. You start with the room. Why?
You have to gain the trust. You have to speak confidently. A lot of it is active listening. A well-timed self-deprecating joke goes over very well, because these folks are also apprehensive. They need to reach out because something is not working right. They do not want to be perceived as dumb or incompetent. If you make yourself out to be the dumb one for a moment, you disarm them almost immediately. They are like, oh, I can trust Shannon, or I can trust so-and-so.
Q: What does the discipline look like once you are in the room?
Cameras on, show the whites of our eyes. In person whenever we can do it. Regular stand-ups, office hours, and the playbook taught in the work rather than emailed in advance. Customers who want to build FinOps out long term learn how we do it by watching us do it every week.
Two Decks, One Mix: Why the FinOps Team Must Be Hybrid
The hybrid team is Shannon's structural argument. She gets pushback on it. Customers try to plant FinOps entirely inside procurement or entirely inside engineering, then wonder why the discipline does not gain traction. Her answer is that both configurations are broken in symmetrical ways. The only model that holds is the one most organisations are uncomfortable commissioning.
Q: Where does the work actually live if not in procurement or in engineering?
It lives in both. Procurement teams take FinOps seriously. They get the business context and the reporting. But they miss the technical ingenuity and the cost levers. There is no technical team member to support them. If it is heavy, heavy engineering, you 100% lose the business context. The financial context is not something you need to hire in. It is already in the company. Accounting and finance teams are not going anywhere. The people on them already handle the bills and the reconciliations. The hybrid team puts those people in the same room as the engineers who operate the cost levers. Not next to them. In the room.
Q: How do you size the team you put in place?
We have T-shirt-sized approaches. Ten million a year of cloud is a different team size than fifty million, different again at a hundred and fifty. We give them our playbook and right-size the pod against their consumption pattern. The technical resources are the full-time job, cranking the levers and building automation. The financial folks are part-time in the mix. They wrap the data in and make sure the reporting makes sense. The slot fits an existing finance role; we are not asking the customer to hire.
Working the Long Set: The Assessment That Decays
This is where the argument moves from the model to its failure mode. Shannon has watched the same pattern across many engagements. The customer commissions an assessment. They take the quick wins. They walk away before the governance and the automation are built. Eighteen months later they call back. The work has to begin again from a worse position. Shannon's response was to stop selling the assessment as the product.
Customers come in and ask for a FinOps assessment. The team puts the assessment together, runs the initial remediation paths, maybe runs a handful of sprints to reduce the cost structure. The customer sees the savings, thanks the team, and walks away. They have not implemented the governance program, the automation, or the reporting. The split is amicable. Eighteen months later the customer comes back. They are in a worse spot than where the engagement first found them. The reason is mechanical. Without the governance hooks the cost discipline regresses. Sprawl returns. The team that knew the playbook has moved on. The customer has to run another hygiene cycle just to get back to where they were a year ago.
Shannon's answer was to lead with sustained engagements, year-minimum, and call the offering FinOps services. Savings show up regularly, and by around month six a lot of the low-hanging fruit is behind the customer. After that, the work turns to governance, automation, and the cadence that keeps the cloud bill from regressing. Her framing internally: FinOps is more like Site Reliability Engineering (SRE) for finance than like a one-shot audit. There is always something to clean up. It never really goes away. It is kind of like always having a janitor around.
Q: What changed once you saw the pattern?
We stopped leading with the assessment. The bet is that the customer commits for at least a year. There is so much that happens over a year in the cloud, so many projects. Cloud is way more ephemeral than people give it credit for. Implementing governance and change is not an overnight sensation. People are still in the mix.
Q: How do you talk to a customer who insists on doing it themselves?
I do not tell them no. I ask them whether they want to. They could just do it themselves. Do they want to? Or would they rather have a team that gets up and loves to do FinOps? Their own people then focus on what their company actually competes on. Companies that die on the FinOps sword want to own every piece of it. They are getting outpaced.
Working with Procurement: The Six-Month Proof Window
Every consultant has a procurement story, and most of them are complaints. Shannon's is not. Procurement is executing a mandate it was given, and a team that treats that mandate as an obstacle only hardens it. She has built specific mechanisms to clear the gate before the work begins.
Q: Procurement gets framed as the antagonist in a lot of FinOps stories. You frame it differently.
Procurement loves to make my life hell. Maybe I should be nicer to them. They have a job to fulfil. A lot of them take it very seriously. Certain procurement teams are really difficult to work with. We factor in a way to play nice on that side. Our consultants are no stranger to working with difficult procurement teams who want to see the value of this work before letting us in.
Q: What did you actually build to address the friction?
Sometimes customers ask for a level of accountability around projected outcomes. One way to address that is by aligning a portion of the commercial model to the expected results, so stakeholders are confident that both parties have skin in the game. In practice, the value of the engagement typically becomes evident over time. The projected savings often materialise within the first several months.
FinOps Plus, Before It Was Called That: On-Prem, SaaS, and the Single Pane of Glass
Last year the FinOps Foundation announced FinOps Plus, extending the discipline beyond cloud to on-prem and SaaS. The announcement landed at Shannon's firm as both validation and a moment of incredulity. The work had been happening under a different name for years. The field is converging on a wider scope. The wider scope is more demanding of the team that delivers it.
Q: When the FinOps Foundation announced FinOps Plus, what was your reaction internally?
Are they tapping into our phone calls with customers? We have a whole side of the business called financial consulting. They stitch on-prem costs together, look at fully-loaded FTE costs of the folks who support environments, review contracts, and build a model that encapsulates on-premises. When the announcement happened, I sent a picture of the keynote slide to that team. They were like, what? It felt like the industry had caught up to where we had been operating for a while.
Q: What does FinOps Plus actually demand from the team?
The scope has expanded beyond traditional cloud and on-premises infrastructure. It now includes a broader set of technology investments. Organisations no longer rely on a single platform. They bring together data from multiple sources and create a more complete financial picture. That broad scoped view makes for better investment decisions. Savings can be redirected elsewhere, to fund modernisation or strategic initiatives.
From Cloud Bill to Token Bill: AI Cost and the Next Governance Frontier
Shannon's forward look is about AI. The timing of this conversation turned out to matter. Three weeks after she predicted the FinOps Foundation would build out a framework specifically for AI, FinOps X 2026 in San Diego did exactly that. The Foundation introduced Token Economics as a primary focus area. It launched the Tokenomics Foundation initiative, a collaboration with the Linux Foundation to standardise AI billing data across providers. It expanded FinOps Scopes to include AI services. The Foundation reported that ninety-eight percent of FinOps practices now treat AI cost management as a priority. The tokenisation question is the cost-discipline question of the next two years. Most organisations are still nowhere on it.
Q: Where does AI cost sit relative to traditional cloud cost in 2026?
AI adoption is accelerating quickly. AI assistants and models become broadly available to the workforce. Usage costs grow much faster than initially expected (twice or more becomes the norm). AI spend becomes a meaningful component of the cost of delivering work. Most organisations already plan for significant growth in AI expenditures. They evaluate different deployment and infrastructure strategies to balance cost, performance, and governance.
Q: What is the discipline that has not yet been built?
Tokenisation is the new optimisation surface. Prompting is the new code review. Asking AI to rewrite a fifty-page paper eighty times is not a great use of AI. Asking it to handle a section, or to co-write, is. The vendors are starting to put cost telemetry in places it was not before. Kong's AI Gateway, for one, treats costing and security telemetry as part of the same control plane. There will be a moment, probably not far out, when the FinOps Foundation builds out the framework specifically for AI. I feel it will look a lot like ITIL.
Key Takeaways
Shannon's perspective sharpens the FinOps argument with these essential insights:
Hybrid team. A FinOps team without both technical and financial seats is structurally incomplete. Procurement-led teams lose the levers. Engineering-led teams lose the context. The hybrid is the only configuration that survives contact with a real cloud bill.
Long set, not single drop. Assessments without governance, automation, and a sustained cadence regress within eighteen months. Lead with the program, not the audit.
Trust before spreadsheet. Active listening, camera-on culture, in-person whenever possible, and a well-timed self-deprecating joke all earn the credibility a spreadsheet cannot.
Scope expansion. FinOps Plus and Token Economics formalise what mature practitioners were already doing. On-prem, SaaS, cloud, and AI inference sit in the same model. The single pane of glass is built, not bought.
Token bill. A developer's licence and a developer's salary can land at the same number within a year if no one is watching. The discipline is the same. The unit is new.
North star. Bridging two systems that do not naturally talk to each other. That is what got Shannon into technology. It is still the job.
Implementation Roadmap
Right-size the team to consumption
Use a T-shirt-sized approach against annual cloud spend (ten, fifty, one hundred and fifty million tiers). Land both technical and financial seats in every pod.
Trade the assessment for the program
Lead with a year-minimum engagement. Build governance, automation, and reporting into the first year, not after the savings have already arrived.
Open with trust, not the spreadsheet
Camera-on by default, in-person where feasible, regular stand-ups, office hours, and the playbook taught in the work.
Disarm procurement structurally
Put the consultant fee against the estimated savings, credit-back if missed. Expect savings to land around month six.
Build the single pane of glass yourself
No single FinOps tool covers cloud, on-prem, SaaS, and AI inference together. Plan for the SRE-style team that builds and maintains the stitches.
Stand up an AI cost surface now
Tokenisation, prompting hygiene, per-licence cost ceilings, and AI gateway telemetry are the new optimisation work. The Tokenomics Foundation gives the standard. Treat AI cost as a first-class FinOps control plane.
The Bottom Line
FinOps is not a fix called in after the bill arrives. It is a sustained governance discipline that operates a hybrid team across the lifetime of a cloud estate. That discipline is now expanding into on-prem, SaaS, and AI tokenisation. Organisations that treat FinOps as a one-shot assessment will keep regressing on a predictable cycle. Organisations that build the long set will keep the bill within reach.
About Shannon Eldridge-Kuehn
Shannon Eldridge-Kuehn is a Principal Solution Architect at AHEAD, where she leads FinOps conversations across the firm's customer base. Her career spans internal IT, managed services, a boutique Azure systems integrator, five and a half years at Microsoft as a cloud advocate and then in the Microsoft Identity Product Group, and now three years at AHEAD. She has spoken at the FinOps Weekly conference on Reserved Instances and savings plans. Connect with Shannon on LinkedIn.
Cloud Value Lab publishes practitioner-led thought leadership at the intersection of FinOps, GreenOps, and AI Economics. If you are a practitioner or subject matter expert interested in sharing your perspective, reach out to David May.