- Why Cost Optimisation Tops the 2026 CIO Agenda
- The Three Aims of Modern Cost Optimisation
- Layer One: Visibility
- Layer Two: Waste Elimination
- Layer Three: Contract Structure
- Build vs Buy and the Real Cost Question
- Cutting Without Cutting Capability
- Process, Automation, and Sustainability
- Benchmarking and the AI Spend Problem
- Defending the Budget You Keep
Why Cost Optimisation Tops the 2026 CIO Agenda
Enterprise IT cost optimization has moved from a periodic finance exercise to a permanent strategic discipline. In the 2026 Gartner CIO and Technology Executive Survey, 52% of leaders said reducing costs will become an even more important objective over the next two years. The reason is scale: global cloud and SaaS waste is now estimated to exceed $200 billion a year, and the waste inside any single enterprise is rarely below 20% of software spend.
The figures are uncomfortable when made specific. Around 30% of SaaS licences in the average organisation go unused, and roughly 53% are underutilised or unused in a given month. Gartner describes about 30% of SaaS spend as "toxic". Companies waste roughly 32% of their cloud budgets on idle or over-provisioned resources, and 94% of enterprises overspend on cloud. SaaS spend now averages around $4,830 per employee — up nearly 22% year on year — and the average organisation wastes about $21M annually on unused SaaS alone. Across the US and UK, $34 billion is wasted on unused licences every year.
Made concrete, the opportunity is striking. For an enterprise spending $50M a year on software, a 25% waste rate represents $12.5M of annual spend producing little or no value — more than most CIOs could free up through any other single initiative. The reason it persists is not incompetence but invisibility: the waste is distributed across hundreds of contracts, thousands of unused seats, and cloud resources nobody owns, so no single line item is large enough to trigger scrutiny. A framework exists precisely to aggregate those small leaks into a number large enough to act on.
This guide is the pillar for our cost-optimisation cluster. It sets out the framework; each layer links to a focused guide that goes deeper. If you want the operating model rather than the strategy, start with our IT cost optimization framework for 2026, which turns the principles below into a repeatable programme built on Run/Grow/Transform, TBM, and FinOps.
The Three Aims of Modern Cost Optimisation
The defining shift in modern practice is that cost optimisation is not cost cutting. Gartner frames it as three aims pursued in parallel: reduce low-value spend, improve enterprise performance, and reinvest the savings in future sources of value. The enterprises that treat optimisation purely as subtraction destroy capability and demoralise the organisation; the ones that treat it as reallocation fund their own growth.
That reframing matters because it changes what "success" looks like. A 25% reduction in a tool nobody uses is a pure win. A 25% reduction in a platform the revenue team depends on is a liability dressed as a saving. The whole point of a framework is to tell those two apart before the cut is made — which is why visibility comes first and service reduction comes last.
Cost cutting asks "what can we remove?". Cost optimisation asks "what are we paying for that creates no value, and what could that money fund instead?". The second question is the one that survives contact with the business.
Layer One: Visibility
You cannot optimise what you cannot see, and most enterprises see far less than they think. Organisations are typically aware of only around 40% of the SaaS applications actually in use — meaning roughly 60% of the portfolio sits in the dark, bought on departmental cards and never recorded centrally. Every cost programme that skips this step ends up optimising the visible 40% while the waste compounds in the rest.
Two disciplines build visibility. The first is spend analytics: instrumenting your software and cloud spend so consumption, ownership, and renewal dates are all queryable in one place. Our guide to IT spend analytics tools and practices sets out the tooling and the metrics that matter. The second is bringing shadow purchasing into the light, because unmanaged tools are both a cost leak and a compliance and security exposure — the full picture is in our analysis of shadow IT licensing risks and cost. Underpinning both is a credible software asset register; if you are building the business case for that capability, our software asset management ROI calculator shows the return.
| Visibility Gap | Typical Scale | What It Hides |
|---|---|---|
| Unknown SaaS apps | ~60% of portfolio | Shadow IT spend and risk |
| Unused licences | ~30% of seats | Reclaimable cost |
| Underutilised apps | ~53% of apps | Right-sizing opportunity |
| Idle cloud resources | ~32% of cloud budget | FinOps waste |
Layer Two: Waste Elimination
Once you can see the estate, the waste layer is where the fastest, lowest-risk savings live — because waste has, by definition, no capability cost. There are three moves, and they run in order.
Rationalisation
Licence rationalisation removes duplicate tools and consolidates overlapping capabilities onto a smaller, intentional set. It is the structural counterpart to reclamation: where reclamation recovers unused seats of a tool you keep, rationalisation decides which tools you keep at all. Our step-by-step software licence rationalisation guide walks through the inventory, overlap-mapping, and decommissioning sequence.
Reclamation
Reclamation programmes systematically recover the ~30% of licences sitting unused and return them to a pool or remove them at renewal. Because the average enterprise wastes around $21M a year on unused SaaS, a standing reclamation process is one of the highest-return activities in the whole framework — detailed in our guide to licence reclamation programmes.
Right-Sizing
Right-sizing moves users off premium editions and over-provisioned tiers they do not actually use — the E5-paying-for-E3-usage problem, repeated across the estate and the cloud. With 53% of applications underutilised, right-sizing is rarely a one-off; our guide to right-sizing enterprise software deployments treats it as continuous discipline. Run together, these three moves typically recover 15–25% of software spend before a single contract is renegotiated.
Layer Three: Contract Structure
Waste elimination cleans the estate; contract structure makes the savings permanent and compounds them at renewal. Three structural levers do most of the work. The first is how the enterprise licence agreement itself is built — bundling, metrics, true-up mechanics, and price protection — covered in our guide to enterprise licence agreement structure optimisation. The second is co-terming: aligning renewal dates so contracts can be negotiated together rather than piecemeal, with the trade-offs set out in our piece on co-terming software contracts. The third is contract consolidation, the disciplined reduction of contract count and supplier sprawl described in our IT contract consolidation playbook.
Structure is where cost optimisation meets negotiation, and the two should never be separated. The cleanest estate in the world still overpays if the contract terms are weak. This is the bridge between this cluster and our negotiation work — the same benchmark data and leverage principles that drive a good renewal also tell you which structural changes are worth pushing for. To anchor the pricing side, our price benchmarking report shows what comparable enterprises actually pay.
Structure also compounds. A waste-elimination saving is captured once; a structural saving — a price-escalation cap, a downsizing right, a co-termed renewal that creates an annual competitive event — keeps paying out every year of the agreement and every cycle after. This is why the structure layer is where independent advisory earns its keep: the gains are not the headline discount but the terms that stop the discount eroding, the protections that let licence counts fall with headcount, and the renewal architecture that keeps the vendor competing for the business rather than coasting on lock-in. An enterprise that rationalises its estate but signs the same weak contract terms simply re-accumulates waste over the following three years.
Build vs Buy and the Real Cost Question
Some of the largest cost decisions are not about how to buy software but whether to. The open-source-versus-commercial question is frequently decided on licence price alone, which is the wrong basis — the real comparison is total cost of ownership, including support, integration, security, and the internal engineering time that "free" software quietly consumes. Our analysis of open source versus commercial total cost works through where each genuinely wins, so the build-versus-buy call is made on lifetime cost rather than the sticker.
The discipline matters because the headline number and the lifetime number frequently point in opposite directions. Commercial software carries a visible licence cost but bundles support, security patching, and a roadmap; open source removes the licence line but transfers those obligations to internal teams whose time is rarely costed into the comparison. Conversely, a commercial tool used by a handful of power users may cost more per genuine user than an open alternative the organisation could run at scale. The framework does not have a default answer here — it has a default method: compare on three-to-five-year total cost, including the people, and revisit the decision as scale and usage change rather than treating it as settled forever.
Cutting Without Cutting Capability
The discipline that separates a cost programme that lasts from one that triggers a backlash is sequencing by capability risk. Waste first, structure second, service last. Done in that order, the great majority of savings are captured before anything the business relies on is touched at all. Our guide to how CIOs cut IT costs without cutting capability turns this sequence into an operating playbook, with the governance to keep service-affecting cuts as a genuine last resort rather than a first instinct.
The same logic applies under acute pressure. When a downturn forces the pace, the order does not change — it just compresses. The disciplined approach to in-year cuts is the same waste-first sequence, executed faster and paired with the contractual protections that stop the saving reversing.
Sequencing also protects the programme politically, which is what determines whether it survives past the first year. A cost initiative that opens by cutting tools people rely on generates resistance that outlasts the saving; one that opens by removing waste nobody misses builds the credibility to make harder decisions later. The order is therefore not only a risk control but a change-management strategy: visible early wins on shelfware and idle cloud earn the mandate to tackle the structural and service-level questions that always arrive eventually. Enterprises that invert the order — leading with the painful cuts to "show seriousness" — typically stall when the backlash arrives, leaving the larger structural savings unclaimed.
Process, Automation, and Sustainability
Optimisation that depends on heroic annual projects does not last; optimisation built into the process does. Two enablers make the framework continuous rather than episodic. The first is procurement automation — reducing the cycle time and manual effort in sourcing and renewals so the organisation can act on optimisation opportunities at the speed they appear; our guide to IT procurement automation and cycle time covers the tooling and the workflow. The second, increasingly material in 2026, is sustainability: green-licensing and efficient-deployment choices that lower both cost and carbon at once, examined in our guide to sustainable IT procurement. The growing adoption of FinOps — now reporting into the CTO or CIO in 78% of practices — is the organisational expression of this shift from project to process.
Benchmarking and the AI Spend Problem
Two forces make 2026 different from previous optimisation cycles, and both demand specific attention. The first is benchmarking. Eliminating waste cleans the estate, but it does nothing about paying an above-market rate for the software you keep — and most enterprises have no idea where their unit prices sit relative to comparable buyers, because vendors price by segment and never disclose the comparison. Without external transaction data, you are negotiating against list price, which is not the benchmark the vendor uses internally. A structured benchmark turns "the renewal looks reasonable" into "we are paying 18% above the median for our size band", which is the difference between accepting a number and challenging it.
The second force is the AI surcharge. Vendors are increasingly hardcoding 15–25% renewal increases justified as the cost of new AI capabilities, regardless of whether those capabilities are used. Because the increase is tied to the vendor's investment rather than the buyer's consumption or return, it sidesteps every conventional value argument — and at enterprise scale it is reintroducing waste faster than rationalisation removes it. The optimisation response is to insist AI features are unbundled and separately priced, to refuse to fund capabilities not yet adopted, and to benchmark the underlying product against its pre-AI baseline. An estate that has been rationalised and instrumented is far better placed to make that argument, because you can show exactly what is used and what is not.
These two forces are why optimisation and negotiation cannot be run as separate programmes in 2026. The cleanest, best-instrumented estate still overpays if the renewal is negotiated from list price and the AI surcharge is accepted without challenge — and the sharpest negotiation still leaves money on the table if it is run against a bloated, unmeasured estate. The framework only delivers its full 20–30% when the two halves operate together.
Defending the Budget You Keep
The final layer is political, not technical. Having optimised, you still have to defend the spend that remains — and in an environment where 52% of CIOs face rising cost pressure, every retained line is questioned. The answer is evidence: tying each investment to a quantified business outcome so the budget conversation is about value, not vision. Our guide to IT budget defence and justifying software investments sets out how to build that case, so the savings you have worked for are reinvested rather than simply clawed back.
That is the full arc of the framework: see the estate, eliminate the waste, restructure the contracts, decide build-versus-buy on real cost, protect capability through the sequence, build optimisation into the process, and defend what remains with evidence. None of it requires gutting the organisation — it requires knowing, precisely, what creates value and what does not. To put the framework to work on your estate, or to pressure-test where your 20–30% of waste is hiding, request a confidential briefing, or download our SaaS optimisation guide and CIO contract governance white paper to start in-house. You can also see how the structural levers connect to live negotiation through our SaaS optimisation practice.