2. AI: chaos before clarity
AI presents a double-edged sword for cost optimisation. On the upside, it creates optimisation capabilities that were impossible with historical, deterministic approaches. But it also drives unpredictable cost increases and, potentially, new governance problems.
63% of organisations now track AI spend, up from just 31% in 2024. That doubling signals the velocity of this shift. However, from a value perspective, this AI spending is often supplementary rather than substitutional: it adds new cost layers rather than replacing existing ones. GenAI features are now embedded across software enterprises already own, and vendors are raising prices accordingly.
Conversely, AI-powered discovery and analytics have the potential to outperform traditional rule-based methods. We can now use machine learning to classify vast numbers of usage signals, normalise noisy data, and surface optimisation opportunities that would be missed at human scale.
We have advised ITAM teams for years not to "boil the ocean." Let your MSP experts focus on your Tier 1 vendors, and your internal team focus on Tier 2. Historically, Tier 3 was largely ignored because the manual effort required to manage the long tail wasn't worth the return. Now, AI provides the scale needed to finally tackle Tier 3, though it remains a developing solution rather than a mature one. The strategic move for 2026 is deploying agentic and AI cost optimised solutions faster than AI-driven costs can compound.
3. Audits: gauge your risk
Audit risk and financial exposure are now at levels that warrant executive attention, though your level of exposure will differ significantly according to your vendor relationships.
Some vendors quickly resort to legal notices when customers push back on subscription terms. Others take a different approach. Oracle continues its Java per-employee pricing push, with Gartner predicting one in five Java-using organisations will be audited in 2026. SAP maintains strict audit practises. Microsoft, by contrast, has largely moved away from traditional audits as it secures revenue through subscription models. AWS or Google don’t audit customers for third-party software licence compliance nor participate in vendor audits.
This mixed picture is another reason for gaining holistic visibility across your whole IT estate so you can clarify vendor relationships, hone your audit strategy, uncover non-compliance and identify optimisation opportunities. We're having conversations with customers on this topic every day so they can get a clearer picture of where they stand on audit exposure. They don’t want to rely on vendor goodwill: they want facts and data.
4. SaaS sprawl: governance goals
Organisations consistently underestimate their level of SaaS sprawl. In my experience, many enterprise organisations “think” they run somewhere between 500 and 1000 SaaS vendors and applications only to find that the real number is closer to 1000 – 2000. Less than half of provisioned licences are actively used or even assigned to people that have changed roles or even company, leaving millions of dollars per year tied up in unused or underused SaaS subscriptions. In addition to the cost implications, SaaS sprawl also represents a widely recognised risk for the proliferation of shadow IT.
Monitoring this spend is becoming easier as cloud marketplaces fundamentally reshape how SaaS gets purchased and governed. These marketplaces create centralised visibility where governance can actually function. The complicating factor is that they also make software purchase friction-free, so organisations can easily end up with more software even if they are able to monitor it more accurately.
Despite this, controlling SaaS spend represents low hanging fruit for cost optimisation. Licence cancellations and tier downgrades can produce savings in the next billing cycle, with some guidance suggesting 8–15% of monthly SaaS spend can be recovered in the first 1–4 weeks.