Most organisations I work with can tell you what they spent on cloud last month. Very few can tell you why it changed, which team drove it, or whether the money bought anything worthwhile.

The invoice arrives, finance pays it, and someone asks why it’s 15% higher than last month. Nobody can answer quickly, so the question gets parked until the year-end review. By then the increase is 40% and it’s too late to do much about it.

That’s the gap FinOps fills. It isn’t a cost-cutting exercise. It’s the practice of getting finance, engineering and product to make decisions about technology spend together, using data they all trust.

Why cloud spend gets away from you

Cloud changed who spends money. In a data centre, spend went through procurement and a capital approval process. In the cloud, an engineer with the right permissions can commit the organisation to thousands of pounds a month in an afternoon. That speed is the point of cloud, but the old controls can’t keep up with it.

What usually follows is that everyone owns the bill, so nobody does:

  • Engineering see it as a finance problem.
  • Finance see it as an engineering problem.
  • Leadership ask for it to be cheaper without saying what they’re willing to give up.

I’ve seen this in government, fintech and e-commerce. A better invoice won’t fix it. You need clear ownership, data people can act on, and a regular routine for acting on it.

What FinOps actually is

The FinOps Foundation’s framework is the standard reference, and it’s worth knowing the basics before you buy any tools. Six principles sit underneath it:

  1. Teams need to collaborate.
  2. Business value drives technology decisions.
  3. Everyone takes ownership for their technology usage.
  4. FinOps data should be accessible, timely and accurate.
  5. FinOps should be enabled centrally.
  6. Take advantage of the variable cost model of the cloud.

The work runs as a continuous loop of three phases:

  • Inform: visibility and allocation. Tagging, cost allocation, dashboards, and knowing who’s spending what and why it changed.
  • Optimise: acting on what you’ve found. Rightsizing, commitments such as Savings Plans and Reserved Instances, removing waste, and architecture changes.
  • Operate: making it business as usual. Policies, automation, governance, and teams making cost-aware decisions without being chased.

Then you go round again, because your workloads and the business keep changing.

Measure value, not just spend

A falling bill isn’t automatically good news, and a rising one isn’t automatically bad. If spend went up 20% because usage doubled, you’re doing well.

That’s why mature FinOps practices track unit economics: cost per transaction, per customer, per case processed. It turns “cloud is expensive” into “each case costs us £X to process, and here’s how we bring that down”. Finance and engineering can both work with that number.

Where are you now?

The framework rates each capability against three maturity levels:

  • Crawl: reactive. You see the bill after the fact and investigate when something looks wrong.
  • Walk: processes are in place, teams understand their costs, and optimisation happens regularly.
  • Run: cost is part of how decisions get made, much of it is automated, and teams improve without prompting.

Most organisations are at Crawl and don’t know it. A billing dashboard on its own doesn’t get you to Walk. Run isn’t the target for everything either. The right level depends on how much each capability matters to you.

Building a FinOps practice at the Home Office

When I took on FinOps for our platform, there wasn’t a practice to speak of. Several teams were spending cloud budget with limited visibility, no standard approach, and optimisation that only happened when something went wrong. Our first maturity assessment put us below Crawl on most capabilities.

I set up the FinOps practice from scratch and used that assessment as the baseline for a two-year roadmap, adding capabilities when the organisation was ready for them rather than all at once.

Getting the basics right

  • A tagging and cost allocation strategy, so spend could be traced to teams and services.
  • Dashboards that non-technical stakeholders could read without help.
  • KPIs that separated strategic measures from operational ones.
  • Reporting and alerting, so leadership saw spend and anomalies early.
  • AWS Budgets and Cost Anomaly Detection as quick wins.
  • Basic hygiene: scheduling non-production instances off out of hours, clearing out old EBS and AMI snapshots, and moving data to colder storage where it made sense.

Building the culture

  • A cross-functional FinOps working group, meeting monthly.
  • FinOps by design, so cost is discussed at every stage of delivery rather than after go-live.
  • Optimisation as routine work, not firefighting.
  • Internal benchmarking, so teams could see how their efficiency compared with others.

The results

  • Annual cloud spend down 40%.
  • Maturity up from pre-Crawl to a comfortable Walk across most capabilities.
  • Teams owning their own spend and making cost-aware decisions without being chased.

Who should run it?

FinOps should be enabled centrally, which means someone has to own it. How big a role that is depends on your spend:

  • Under about £50k a month: a senior person spending part of their time on visibility and quick wins is usually enough. Get tagging right now, because retrofitting it later is painful.
  • £50k to £500k a month: this is where a dedicated FinOps lead pays for itself. Waste at this level is real money, and tackling it is rarely anyone’s specific job.
  • Over £500k a month: you’ll probably need a small team, organised by platform or business unit, and FinOps becomes part of how you plan technology investment.

The role works best with a product mindset. Treat your cloud platform as a product, with outcomes, a roadmap and stakeholders, and cost becomes something you manage deliberately rather than react to. Engineers can do much of the optimisation work, but they’re rightly focused on delivering features. Someone needs to translate between finance and engineering, prioritise the work, and keep going once the easy savings are done.

What to expect

Be wary of anyone promising a zero bill, savings with no trade-offs, or instant results. A mature practice usually takes six to twelve months to build. What you should get is:

  • Clear visibility of where the money goes and why.
  • A maturity roadmap with agreed priorities.
  • Engineering, finance and leadership working from the same numbers.
  • Savings that last, because they’re built into how teams work.
  • Teams able to make cost-aware decisions on their own.

How I can help

I’m a product and delivery lead with over nine years’ experience in government and regulated industries, and a FinOps Certified Practitioner and AWS Solutions Architect Associate. At the Home Office I built a cloud platform’s FinOps practice from scratch, taking it from pre-Crawl to Walk and cutting annual cloud spend by 40%.

If you’re spending serious money on cloud and nobody is specifically accountable for it, I can assess where you are, agree where you need to be, and build and deliver the roadmap to get there. Get in touch if you’d like to talk it through.