Contents
The Voice of Volge is Volge’s thought-leadership series, unsigned and institutional, every figure sourced, quantified and year-stamped. Volge is a Flexera delivery partner; interest disclosed.
volge.com/voice
About Flexera
Flexera is a technology-spend management software company, and this book reads it from the side it started on: the owned landscape. Before it governed rented cloud and SaaS, Flexera governed installed software, licences, entitlements, and the audit exposure they carry, and it has since added vulnerability intelligence and lifecycle data to make the landscape counted, secured and retired.
The landscape now includes AI. Flexera's 2026 State of ITAM Report (24 June 2026) finds only 31 per cent of organisations with accurate visibility into AI software, the newest asset class to enter the owned landscape, and the least governed.
| See | IT Visibility · Technopedia · hardware & software inventory |
| Comply | Software Asset Management · Effective Licence Position · audit response |
| Secure & retire | Software Vulnerability Manager (Secunia) · lifecycle / end-of-life data |
Source: Flexera 2026 State of ITAM Report, 24 June 2026 (n=512). Figures describe the landscape, not Flexera outcomes.
Flexera's oldest strength is the owned landscape: counting software and defending audits. Its newest challenge is the same landscape's newest resident, AI, which almost no one can yet see. The pitch is a single view across both; the test is whether it reaches the assets you cannot currently find.
You are liable for everything you have ever installed.
Not just what you use, what you deployed, forgot, and left running. To an auditor, an attacker, and the end of its support life, it is all still yours.
The rented landscape bills you monthly;
the owned landscape bills you all at once
The companion to this book, The technology tax, was about the technology you rent: the meters that run every second and arrive as a monthly bill. This book is about the technology you own, and it bills differently. The owned landscape does not send an invoice. It sends an auditor's letter, a breach notification, or a support-expiry notice, and each arrives without warning and all at once. Renting is a cash-flow problem. Owning is a liability problem.
Everything an organisation has ever installed and not deliberately removed is still, in the eyes of the parties that matter, its responsibility. The software vendor counts it and can audit the count. The attacker scans it and can exploit the unpatched. The support calendar expires under it and turns a working system into an unsupported one. None of these depends on whether anyone is still using the thing. Liability attaches to installation, not to use.
This is why the owned landscape is dangerous in a way the rented one is not. A meter you ignore merely overspends. An asset you ignore accrues exposure silently until the day it is called, the audit true-up you did not budget, the vulnerability you did not patch, the end-of-life system that was quietly running your payroll. The cost is invisible right up to the moment it is total, and by then the choices that would have prevented it are years in the past.
The 2026 figures show the liability growing, not shrinking. Flexera's State of ITAM Report finds 48 per cent of organisations audited in the past year and 44 per cent spending more than a million dollars on software audits over three years, with Microsoft audits reported by 64 per cent and Oracle audits climbing to 38 per cent (Flexera, 24 June 2026). Complete IT asset visibility, meanwhile, has fallen to 36 per cent. The exposure is rising as the ability to see it falls.
And a new asset class has entered the landscape. The AI an organisation deploys, the models and agents running inside its walls, are assets to be inventoried, secured, and eventually retired, not merely a bill to be paid. Yet only 31 per cent of organisations have accurate visibility into their AI software, and 59 per cent report rising wasted AI spend (Flexera, 24 June 2026). The newest resident of the landscape is the one almost no one can find.
Flexera's answer is to make the owned landscape visible and then discharge its liabilities in turn. Software Asset Management defends the audit. Software Vulnerability Manager, backed by Secunia Research, secures the unpatched. Technopedia's lifecycle data flags what must be retired. Volge licenses, configures, and operates that instrument on our clients' behalf, and we wrote this book to make the case: with our commercial interest marked wherever it appears.
Read on for the three liabilities in turn, comply, secure, retire, and then for the operator's read: how to carry the owned landscape as a managed risk rather than an unbudgeted surprise.
What the audits cost
SAM & the licence position
2 · The audit exposure
3 · Entitlement vs deployment
The audit exists because you do not know what you have deployed
A software audit is a bet the vendor makes that you have lost count, and it is usually a good bet. Between what an organisation is entitled to run and what it has actually installed lies a gap that no one is paid to close, servers re-imaged, licences double-assigned, editions upgraded without a paper trail, virtual machines cloned with their software aboard. The vendor knows this gap exists. The audit is simply the mechanism for pricing it, at the moment least convenient to you and most convenient to them.
Entitlement and deployment drift apart for structural reasons, not negligent ones. Enterprise licensing is bafflingly complex, per core, per user, per server, with virtualisation rules that can multiply a single licence requirement across an entire cluster. A well-meaning engineer who moves a workload to a bigger host can, under some metrics, quadruple the licence liability without touching the software. The rules are designed to be hard, and the difficulty is not incidental to the vendor's revenue; it is the source of it.
The audit is therefore best understood as a sales channel rather than a compliance exercise. Vendors under pressure to grow revenue from a mature product turn to their installed base, and the audit converts a customer's poor record-keeping into an unplanned purchase, the "true-up", often bundled with a conveniently-timed offer to migrate to the cloud edition. The finding is real; the licence gap usually does exist. But the timing, the framing, and the price are the vendor's to choose unless the customer arrives with its own numbers.
That is the whole game: whoever has the better count sets the terms. An organisation that cannot state its own effective licence position negotiates blind, accepts the vendor's figure, and pays the true-up the vendor computed. An organisation that can state its position, with normalised, defensible deployment data, negotiates from evidence, contests the inflated finding, and frequently discovers it was over-licensed rather than under. The count is the advantage, and most organisations do not have it.
| Liability | Called by | Arrives as | Flexera instrument |
|---|---|---|---|
| Comply | The software vendor | An audit and a true-up bill | Software Asset Management |
| Secure | An attacker | An exploited unpatched flaw | Software Vulnerability Manager |
| Retire | The support calendar | An unsupported, un-patchable system | Technopedia lifecycle data |
Half of large organisations were audited last year, and the bill is climbing
The software audit is not a tail risk; it is a near-certainty on a long enough horizon, and in 2026 the horizon shortened. Flexera's State of ITAM Report finds 48 per cent of organisations audited in the past year alone, and 44 per cent having spent more than a million US dollars responding to software audits over three years (Flexera, 24 June 2026). Audit response now consumes 22 per cent of the average ITAM team's time: a fifth of a scarce function spent defending against a bill the vendor chose to send.
Oracle audits rose to 38% (from 24%) and Adobe to 32% (from 24%); 44% of organisations spent over $1m responding across three years.
The concentration is instructive. Microsoft, Oracle, and Adobe, the vendors with the most complex metrics and the largest installed bases, audit most, because that is where the entitlement gap is widest and the true-up largest. These are not obscure suppliers; they are the software every enterprise already runs, which is precisely why the exposure is universal rather than niche. If you run their products, you are on the list; the only question is timing.
The rising Oracle and Adobe figures point to a broader pattern: as licensing models shift, to the cloud, to subscription, to consumption, vendors use the audit to accelerate the migration. A finding of non-compliance on the old model becomes the lever for a move to the new one. Reading the audit as a technical compliance check, rather than as the opening move in a commercial negotiation, is the mistake that makes it expensive.
Software Asset Management gives you the count before the vendor sets the price
Software Asset Management, the discipline Flexera has practised longest, deepened by the 2024 Snow acquisition, exists to produce one artefact the audit turns on: the effective licence position, a defensible reconciliation of what you are entitled to against what you have actually deployed. Produce it before the vendor does, keep it current, and the audit changes character from an ambush into a negotiation between two parties who both have numbers.
The work starts with discovery, because you cannot reconcile what you cannot see. Flexera's inventory and IT Visibility capabilities, normalised against the Technopedia catalogue of more than five million hardware and software items, turn the raw and inconsistent output of an enterprise landscape, the same product named five ways across four tools, into a single, clean count. This normalisation is the unglamorous foundation of everything that follows; without it, the licence position is built on sand.
The reconciliation then maps that deployment against entitlement, the contracts, the purchase records, the licensing metrics, to compute where the landscape is over-licensed, under-licensed, and correctly licensed, product by product. The under-licensed positions are the exposure to close before an auditor finds them. The over-licensed positions, which are common and which no vendor volunteers, are money to reclaim at the next renewal. Both are invisible without the reconciliation, and both are worth real money.
An audit-ready posture is the standing result: a licence position current enough that a vendor's letter prompts a review rather than a panic. When the finding arrives, the organisation checks it against its own numbers, contests what is inflated, and settles what is real: often for a fraction of the opening figure. The saving is not a discount; it is the difference between negotiating from evidence and negotiating from fear.
We will state the limit, as we do throughout. SAM produces the position; it does not negotiate the contract. The reconciliation gives you the evidence, but a licensing negotiation with Oracle or Microsoft is a specialist commercial exercise involving judgement, timing, and nerve that no tool supplies. That combination, the defensible count and the negotiation around it, is precisely where a delivery partner earns its place, and where an unaided software purchase tends to stall.
| Position | What it means | The action |
|---|---|---|
| Under-licensed | Deployed more than entitled | Close before the auditor finds it |
| Correctly licensed | Deployment matches entitlement | Maintain the evidence |
| Over-licensed | Entitled to more than deployed | Reclaim at renewal: money the vendor will not volunteer |
The publishers who set the audit weather
Audit exposure is not evenly distributed across the landscape. Four publishers, Microsoft, Oracle, IBM, and SAP, lead the enterprise audit and true-up risk, and each has built a licensing mechanism that prices the gap between what you measured and what you run. Know the mechanisms and you know where a SAM programme earns its keep first.
Oracle moved Java SE to the Universal Subscription in January 2023, and with it to an employee-count metric: the subscription is priced against the whole workforce, not the developers who use Java. A single JDK install on one server can put the whole employee base in scope, so Java discovery, every install, edition, version, is now the first question in an Oracle conversation. Database options, enabled with a command and licensed separately, remain the older trap.
SAP's Digital Access, introduced in April 2018, charges for indirect use by the document: orders, invoices, and other core documents created in SAP by non-SAP systems carry the licence. Every integration that writes into SAP, a web shop, a CRM, now an AI agent, is a potentially metered actor. The exposure hides in the architecture diagram, not the user list.
IBM permits the cheaper sub-capacity counting, the virtual machine rather than the whole cluster, on one condition: the IBM License Metric Tool deployed, collecting, and reporting. Absent ILMT, the default is full capacity, and on a large virtualised estate the difference is a multiple, not a margin.
Microsoft's mechanism is quieter: not the audit hammer but the drift, E3 renewed where usage justifies less, E5 features paid for twice over point tools, Hybrid Benefit claims resting on unverified licence positions. The exposure arrives as a renewal quote, which makes it easier to miss and no cheaper.
The list is open. Broadcom's VMware acquisition closed in November 2023; within weeks the portfolio was subscription-only and per-core: a benign vendor turned audit-weather publisher by a change of ownership. Adobe, Salesforce, and ServiceNow run their own variants. The defence has to be the landscape, not the publisher.
| Publisher | The mechanism | What it prices | Since |
|---|---|---|---|
| Oracle | Java SE Universal Subscription: per-employee metric | Every employee, not every user of Java | January 2023 |
| SAP | Digital Access: document-based indirect use | Documents created in SAP by non-SAP systems | April 2018 |
| IBM | Sub-capacity PVU, conditional on ILMT reporting | The full cluster, wherever ILMT is absent | Passport Advantage terms |
| Microsoft | Tier mix, CSP billing, Azure Hybrid Benefit | The gap between tiers bought and features used | Continuous |
Prioritise by criticality
SVM & Secunia Research
You cannot patch what you cannot see, and most of what you cannot see is not Microsoft
The unpatched software in an landscape is its standing invitation to an attacker, and the dangerous part of it is rarely the operating system. Windows patches itself on a schedule most organisations have automated. The exposure lives in the third-party software layered on top, the browsers, runtimes, PDF readers, media libraries, and hundreds of niche applications that each carry their own vulnerabilities and none of which the operating-system patch cycle touches. It is a large surface, and most organisations cannot see the whole of it.
The scale of the problem is set by two numbers moving in opposite directions. The count of disclosed vulnerabilities rises every year into the tens of thousands, spread across thousands of products. The capacity of a security team to test, schedule, and deploy patches does not rise to match. The result is a permanent triage problem: far more known vulnerabilities than any team can remediate, and no obvious way to tell the handful that will be exploited from the many that will not.
Responding to volume alone makes the problem worse. A team that treats every disclosed vulnerability as equally urgent spreads its finite patching capacity evenly across a threat landscape that is anything but even, and exhausts itself remediating low-risk flaws while a genuinely dangerous one waits in the queue. The metric that matters is not how many vulnerabilities exist but which of them are present in your landscape, reachable, and actually being exploited: a far smaller and more actionable set.
That reframing requires two things most organisations lack: an accurate inventory of what third-party software is actually installed, and reliable intelligence about which vulnerabilities in that software are real and serious rather than noise. The first is the discovery layer from Part I, reused. The second is a research capability: the difference between a raw feed of every disclosure and a validated, prioritised advisory that tells you where to spend the day's patching capacity.
If tens of thousands of vulnerabilities are disclosed a year and a team can meaningfully patch a fraction of them, the entire value is in choosing the right fraction. Patching more is not the goal. Patching the ones that would actually have been exploited, first, is.
Secunia Research turns a feed of every flaw into a list of the ones that matter
Software Vulnerability Manager exists to solve the triage problem, and its distinguishing asset is Secunia Research: a team that has validated and prioritised vulnerability disclosures since 2002. The difference between SVM and a raw vulnerability feed is the difference between data and intelligence: Secunia does the work of testing disclosures, rejecting the illegitimate ones, scoring the real ones for criticality and attack vector, and pairing them with remediation, so a security team receives a ranked list rather than an undifferentiated stream.
The research is deliberate rather than automated. Secunia's specialists conduct code audits and binary analysis to find issues that automated fuzzing misses, and, just as important, they reject illegitimate or duplicate reports so that teams are not made to chase phantoms. In a field drowning in disclosures, the act of authoritatively saying "this one is not real, and this one is" is itself much of the value. Noise is not a neutral cost; it is the thing that hides the signal a team needed to act on.
SVM then joins that intelligence to the landscape. By matching validated, scored advisories against what is actually installed, the normalised inventory from Part I, it produces the only list that matters operationally: the vulnerabilities that are both serious and present in your environment, ranked so the scarce patching window is spent on the flaws most likely to be exploited. Flexera pairs this with what it describes as the largest repository of third-party patch data in the industry, closing the loop from "what is dangerous" to "here is the patch."
The honest limit is the same one every security team knows. Identifying and prioritising a vulnerability is not the same as remediating it; the patch still has to be tested against a live environment and deployed inside a change window that operations, not the tool, controls. SVM makes the triage decision defensible and the patch available. It does not absorb the operational risk of deployment, and it does not cover the vulnerabilities in software you wrote yourself. It sharpens the judgement and shortens the list. Someone still has to do the patching.
The AI you now own
Lifecycle & asset governance
6 · The AI-in-landscape gap
End-of-life is the quietest liability: nothing breaks, until everything is exposed
Obsolescence is the liability that announces itself last. When a piece of software reaches end-of-life, nothing visible happens, it keeps running, keeps serving users, keeps looking like an asset. What has actually happened is that the vendor has stopped shipping security patches for it, so every vulnerability disclosed from that day forward is one you can never close. The system has quietly converted from a working asset into a permanent, un-patchable exposure, and it will go on doing its job right up to the moment it becomes the way in.
The reason obsolescence accumulates is that retiring software is nobody's urgent priority. Replacing a working system costs money and effort now to avoid a risk that is diffuse and later, which is exactly the trade organisations are worst at making. So the end-of-life database server keeps running the report it has always run, the unsupported operating system stays on the machine in the corner, and the risk compounds silently on a balance sheet no one reads until an auditor or an attacker reads it for them.
Seeing obsolescence before it bites requires knowing two things the landscape does not natively record: precisely what versions of what products are installed, and what the support lifecycle of each of those versions is. The first is the discovery layer again. The second is lifecycle intelligence: a maintained record of when every product version reaches end of support and end of life. Neither an inventory tool nor a vendor's website gives you both in a form you can act on across an landscape of thousands of products.
This is where a normalised catalogue with lifecycle data earns its place. Technopedia's five-million-plus items carry identity and lifecycle status, so the landscape can be read forward in time: what is unsupported today, what expires this year, what to budget to replace before it becomes an exposure. Obsolescence stops being a surprise discovered during an incident and becomes a schedule managed in advance: which is the only affordable way to manage it.
| State | What it means | The liability |
|---|---|---|
| Supported | Patches still shipping | Normal: keep current |
| End of support approaching | Expiry dated and known | Budget the replacement now |
| End of life | No further security patches | Every new flaw is permanent: retire or isolate |
The AI you deploy is an asset in the landscape, and almost no one can find it
The companion book governed the AI you rent: the token bill that arrives each month. This book's concern is the AI you own: the models, the agents, the AI-enabled features now installed and running inside the landscape, which are assets to be inventoried, secured, and eventually retired exactly like any other software. The difference is that this asset class arrived faster than any before it, and the landscape has no record of most of it. It is the obsolescence and vulnerability problem of this book, applied to software that did not exist in the inventory two years ago.
Only 31% have accurate visibility into AI software; complete IT asset visibility overall has fallen to 36%.
The figures describe an asset class outrunning its governance. Only 31 per cent of organisations have accurate visibility into their AI software, even as roughly half have begun tracking it as a line of software spend and 59 per cent report that AI waste is rising (Flexera, 24 June 2026). A deployed model an organisation cannot see is a model it cannot secure, cannot license correctly, and cannot retire, the three liabilities of this entire book, reappearing in their newest and least-understood form.
Treating deployed AI as an asset rather than only a cost changes the questions. What models are running in our landscape, and under what licence terms? Which of them carry known vulnerabilities or were trained on data we must account for? Which are shadow deployments a team stood up without record? These are inventory-and-lifecycle questions, and they are answered with the same discipline used for any other software, provided the discovery layer can find AI at all, which for most organisations it cannot yet.
Retiring the landscape, and bringing AI into it, is one discipline, not two
The instrument for retirement is the same landscape view used to comply and to secure, read forward in time. Technopedia's lifecycle data lets an organisation schedule obsolescence rather than discover it, and the same discovery-and-normalise layer that finds installed software is what must be extended to find installed AI. The claim Flexera makes for 2026 is that AI is not a separate governance problem but the existing landscape discipline applied to a new asset class: provided the discovery reaches it.
For obsolescence, the operating pattern is a rolling one: read the landscape against lifecycle data, identify what expires in the planning horizon, and budget the replacements before end-of-life turns them into permanent exposures. This converts a reactive scramble, discovering an unsupported system during an incident, into a line in the technology plan. It is unglamorous asset hygiene, and it is far cheaper than the breach or the audit it forestalls.
For AI, the honest position is that the discipline is clear and the tooling is early. Bringing deployed models and agents into the inventory, tracking their licence and lifecycle status, and flagging shadow deployments is the right frame, and Flexera is extending its landscape view toward it in 2026. Whether that view yet reaches the full sprawl of AI an organisation has deployed is a question a buyer should test against their own environment, not take on faith. We recommend the discipline confidently and the maturity cautiously.
End-of-life is the quietest liability: nothing breaks until everything is exposed, and only lifecycle data read forward in time prevents it. Deployed AI is the newest resident of the landscape and the least visible, 31 per cent see it accurately, and it must be inventoried, secured, and retired like any other asset. Part IV carries the three liabilities into one operator's read.
The evidence, honestly
Six buying questions
Objections · example · routes
Evidence · the survey read
8 · Three liabilities, one landscape
9 · Three routes
The three liabilities share one root cause: an landscape no one can fully see
Comply, secure, and retire look like three problems owned by three functions, asset management, security, and infrastructure, but they are three symptoms of one condition: an organisation that cannot see, accurately and in one place, what it has installed. The audit exploits the gap between deployment and entitlement. The breach exploits the unpatched software no one inventoried. The obsolescence accrues on versions no one tracked. Fix the visibility and you do not fix three problems; you remove the ground all three grow in.
This is the argument for one landscape view rather than three point tools. When the same normalised, Technopedia-referenced inventory feeds the licence position, the vulnerability triage, and the lifecycle schedule, the three functions are working from one truth instead of three partial ones. The security team's patch list and the asset team's licence position are drawn from the same count of what is installed: which is the only way the AI a team deployed last month shows up in all three views at once rather than in none of them.
Governed separately, the liabilities hide in each other's blind spots. A security tool built for Microsoft misses the third-party sprawl; an asset tool built for licences misses the vulnerability; neither is looking for the end-of-life date or the shadow model. The single landscape is what makes an asset governable on all three axes at the moment it is discovered, rather than on one axis by one team while the other exposures wait to be called.
And we keep the disclosed scepticism from the companion volume. Flexera assembled this breadth over years and by acquisition, and the single-landscape promise is conditional on those pieces sharing one clean inventory in practice, not just on one slide. A buyer should ask to see the licence position, the vulnerability list, and the lifecycle schedule drawn from one discovery of their own environment before believing the one-landscape story. We ask on our clients' behalf.
What the survey shows, and what we will not pretend it is
A disclosed-interest analysis should be candid about the quality of its own evidence, and here candour requires a distinction the companion book did not. Much of this book's data comes from Flexera's own 2026 State of ITAM Report: a survey of 512 practitioners that is genuinely independent in its respondents but produced by the vendor selling the remedy. We treat those figures as directional rather than decisive, and we lean the argument on the structural facts, which do not depend on any survey.
| Finding | Figure | Source |
|---|---|---|
| Organisations audited in the past year | 48% | Flexera ITAM, 24 Jun 2026 |
| Spent >$1M on software audits over three years | 44% | Flexera ITAM, 24 Jun 2026 |
| Reported a Microsoft audit in three years | 64% | Flexera ITAM, 24 Jun 2026 |
| Complete IT asset visibility | 36% | Flexera ITAM, 24 Jun 2026 |
| Accurate visibility into AI software | 31% | Flexera ITAM, 24 Jun 2026 |
| Reporting rising wasted AI spend | 59% | Flexera ITAM, 24 Jun 2026 |
The structural argument stands without the survey. That enterprise software vendors use audits as a revenue channel, that licensing metrics are complex enough to guarantee an entitlement gap, that end-of-life software stops receiving patches, and that deployed AI is software an landscape must inventory: none of these is a survey finding. They are how the licensing, security, and support systems are built, and they would be true if no one had run a survey at all. The figures put a size on a problem whose existence does not depend on them.
So we use the numbers to indicate scale and direction, not to close the case, and we would rather you discount them as vendor research than have you feel misled when you notice their source. The case for governing the owned landscape rests on the mechanics of audits, vulnerabilities, and support calendars, and on your own inventory, which is the only evidence about your landscape that matters.
Six questions to ask before you spend on landscape governance
The following six questions are the ones Volge puts to a client before recommending any platform, and they interrogate us as readily as Flexera. A programme that can answer them with numbers and named owners is ready to buy. One that answers with intentions should start with a discovery, which is small, fast, and the honest first step toward carrying the landscape as a managed risk rather than an unbudgeted one.
For Microsoft, Oracle, and the other audit-prone vendors, before a letter arrives. If not, the next audit is negotiated on the vendor's numbers, not yours.
The attack surface is the third-party layer. A security programme blind to it is patching the wrong landscape diligently.
If the answer takes weeks to assemble, obsolescence is being discovered during incidents rather than scheduled in advance.
Only 31 per cent can. If you cannot inventory your models and agents, you cannot secure, license, or retire them.
If asset, security, and infrastructure work from three counts, the single-landscape benefit is theoretical. Ask to see one discovery feed all three.
An landscape view un-operated ages into another unsupported system. Decide who runs the loop, or do not buy the loop.
A programme that answers these six with numbers and names is ready to spend. One that answers with hopes should start with a discovery, small, fast, and the honest first step.
The four objections a sceptical board raises, and the honest reply to each
A well-run board does not buy on an audit statistic, and the four objections below are the ones we meet most often, each reasonable, none dismissed with a slogan. We set them out because an analysis that ducks the hard questions is not worth reading, and because the answers are, in most cases, stronger than the objection.
Very likely, and the scanner produces a list far longer than the team can action. The gap is not detection but prioritisation, which flaws are real, present, and exploitable, and it is where validated research, not another scanner, earns its place.
It is, and that is the point. The audit is old and unglamorous too, and it costs 44 per cent of large organisations over a million dollars. Boring disciplines that prevent expensive surprises are exactly the ones worth funding.
It is, and we have said so. The discipline, inventory, secure, retire deployed AI, is right; the tooling to find all of it is early, and we recommend the frame confidently and the maturity cautiously. Start with software and audit, where the capability is proven, and treat AI as the frontier it is.
You can, and the licence is the easy part. The value leaks in the work around it: reconciling the licence position, negotiating the audit, triaging the patch list, scheduling the retirements. That is judgement and persistence, not a login, and it is what Volge delivers.
Each objection assumes the purchase is the software. It is not. The purchase is a seen, reconciled, continuously-governed landscape, and the discipline to discharge its liabilities before they are called. The software is a means; the governance is the thing.
One landscape, three letters, an audit, a CVE, and an expiry notice
A single composite shows the method more clearly than description, so here is one landscape receiving, in a single quarter, the three letters this book is about, illustrative, drawn to match the shape governance commonly reveals rather than any one client. The point is not the amounts but the pattern: how one landscape view answers all three, and how their absence turns each into a scramble.
An Oracle audit notice. A critical vulnerability disclosed in a widely-installed third-party library. And a vendor's notice that a database version running a core report reaches end-of-life in ninety days. Three functions, three deadlines, one landscape, and, in the un-governed case, three separate panics conducted from three partial inventories.
| Letter | Un-governed response | One-landscape response |
|---|---|---|
| Oracle audit | Weeks assembling a count; accept the finding | Licence position already current; contest the inflated figure |
| Critical CVE | Is it even installed? Where? Unknown | Matched to inventory in hours; patch the exposed hosts first |
| End-of-life notice | Discovered late; emergency migration | Already on the lifecycle schedule; replacement budgeted |
The governed column is not faster because the team is better; it is faster because all three answers come from one inventory that already existed before the letters arrived. The licence position was current, so the audit is a negotiation. The vulnerability matches a known landscape, so triage is hours not days. The end-of-life date was on a schedule, so the migration is planned not panicked. The un-governed column is three scrambles because there was no shared count to answer from: which is the whole cost of not having one.
A vendor deck would total a saving here. We will not, because it depends entirely on how un-governed the starting landscape was. A large headline number usually means the baseline was badly managed rather than the tool being brilliant. Your own discovery produces the only figure that matters.
Build, buy, or have it delivered: an honest comparison
There are three ways to acquire landscape-governance capability, and each suits a different organisation, so we set them side by side rather than pretending only one is defensible. Build the practice yourself. Buy the platform and staff it. Or have it licensed and delivered through Volge. We sell the third, and we will still tell you when one of the others fits you better.
| Dimension | Build in-house | Buy software, self-run | Delivered through Volge |
|---|---|---|---|
| Time to audit-ready | Slow: months to a practice | Medium: licence then learn | Fast: discovery in weeks |
| Where value leaks | Reinventing SAM & SVM | Discovery, reconciliation, triage | Managed as part of delivery |
| Ongoing burden | You own it all | You staff and operate it | Single accountable partner |
| Best for | Firms with a mature ITAM & security function | Firms with spare capacity | Firms wanting the outcome, not the project |
Building the practice yourself makes sense for organisations with a mature ITAM and security function who would rather own the whole discipline. For most, standing up licence reconciliation, vulnerability triage, and lifecycle governance from scratch is an expensive way to arrive where a platform already sits. We would talk you out of it unless you have a specific reason.
Buying the platform and running it yourself is the middle path, and it works when you have spare capacity and patience for the learning curve on discovery, reconciliation, and triage. The familiar risk is that the licence gets bought, the practice never quite forms, and the landscape view ages into shelfware. Most of the stalled programmes we are asked to rescue started here.
Delivery through Volge suits the organisation that wants the outcome rather than the project, a seen landscape, a defended audit, a triaged patch list, and a lifecycle schedule, with one partner accountable end to end. It is what we do, and we price the first step, a discovery, low enough that choosing it is a small decision, not a large one.
Glossary
Plain definitions of the terms used in this book, in the sense we have used them.
- The installed liability
- This book's term for the standing exposure of the owned landscape, audit, breach, and obsolescence, plus the newest asset, deployed AI.
- The owned landscape
- Software and AI an organisation has installed and runs itself, as opposed to the rented cloud and SaaS of the companion book.
- ITAM
- IT Asset Management: the discipline of inventorying and governing owned hardware and software.
- SAM
- Software Asset Management: the licence-and-compliance branch of ITAM. Flexera's oldest strength, deepened by Snow (2024).
- Entitlement
- What a contract permits you to run, by user, core, server, or other metric.
- Deployment
- What you have actually installed and are running, regardless of entitlement.
- Effective licence position (ELP)
- The reconciliation of deployment against entitlement, product by product: the artefact an audit turns on.
- True-up
- The bill a vendor issues when an audit finds deployment exceeding entitlement.
- IT Visibility
- Flexera's discovery and normalisation capability: turning inconsistent landscape data into one clean count.
- Technopedia
- Flexera's normalised catalogue of 5m+ hardware and software items, carrying identity and lifecycle data.
- Software Vulnerability Manager (SVM)
- Flexera's platform for identifying, prioritising, and patching vulnerable software in the landscape.
- Secunia Research
- The team, active since 2002, that validates and scores vulnerability disclosures behind SVM.
- Third-party software
- Non-OS applications, browsers, runtimes, readers, that carry most of the unpatched attack surface.
- End of life (EOL)
- The point at which a vendor stops shipping security patches; every subsequent flaw is permanent.
- Lifecycle data
- The record of support and end-of-life dates per product version, how obsolescence is scheduled, not discovered.
- The AI you own
- Models and agents deployed inside the landscape, to be inventoried, secured, and retired as assets. The subject of this book.
- The AI you rent
- Consumption cost of models and agents, metered by the token. The subject of the companion book.
Methodology & editorial standards
This book follows the same rules as its companion, designed to keep a disclosed-interest analysis honest. Every external figure is attributed and dated. Vendor positioning is labelled as positioning. Newly-shipped capability is labelled as unproven where it is unproven. Illustrative figures are labelled as illustrative. Our commercial interest is disclosed wherever it bears on the argument, and, in this book especially, the provenance of our headline figures is stated plainly.
Most of this book's numbers come from Flexera's own 2026 State of ITAM Report (24 June 2026, n=512). We flag this openly on page 16: it is a survey of independent practitioners produced by the vendor selling the remedy, and we treat it as directional rather than decisive. Unlike the companion volume, we could not anchor the owned-landscape figures to a fully independent data house, and we would rather say so than imply an independence we do not have.
The case does not rest on the survey. That vendors use audits as a revenue channel, that licensing metrics guarantee an entitlement gap, that end-of-life software stops receiving patches, and that deployed AI is software to inventory, these are properties of how the systems are built, not survey findings, and they carry the argument on their own.
Exhibit 9, the three letters, is an illustrative composite, not a measurement of a named client. We quote no single headline saving, because an honest one depends entirely on how un-governed the starting landscape was.
Volge is a Flexera delivery partner and benefits if you buy. We have handled that conflict by disclosing the provenance of our figures, resting the argument on structural facts rather than vendor claims, labelling new capability as unproven, and keeping the first step, a discovery, small.
If any figure here has aged or any attribution is wrong, we would rather know. This is a first edition, and the AI-asset material in particular will have moved by the time you read it.
Sources
Every external claim in this book, with source and date. Listed in order of first appearance.
- Flexera: 2026 State of ITAM Report (24 June 2026, n=512): 48% audited in the past year; 44% spent >$1M on software audits over three years; Microsoft audits reported by 64%, Oracle up to 38% (from 24%), Adobe up to 32% (from 24%); audit response 22% of ITAM team time; complete IT asset visibility 36%.
- Flexera: 2026 State of ITAM Report (AI): Only 31% have accurate visibility into AI software; ~50% tracking AI as part of software spend; 59% report rising wasted AI spend.
- Flexera Secunia Research: Vulnerability intelligence since 2002; validates and scores disclosures via code audit and binary analysis; rejects illegitimate reports; criticality and attack-vector scoring.
- Flexera Software Vulnerability Manager: Identifies and prioritises vulnerable software in the landscape; described by Flexera as backed by the largest repository of third-party patch data in the industry.
- Flexera Technopedia: Normalised catalogue of 5m+ hardware and software items, carrying lifecycle / end-of-life data.
- Flexera Software Asset Management: Effective licence position and audit response; SAM depth extended by the Snow Software acquisition (completed February 2024).
- Flexera corporate history: Spun out of Macrovision (2008); Secunia vulnerability research active since 2002; Snow Software completed February 2024. Majority-held by Thoma Bravo.
The headline audit and AI figures are Flexera's own survey research, labelled as such throughout. We have not dressed them as third-party validation, and we rest the argument on the structural mechanics of audits, vulnerabilities, and support lifecycles, which hold regardless of any survey.
The reader's checklist
Tear-out version. Six questions from the operator's read (page 17), on a single page you can take into a meeting.
| Ask | If the answer is vague |
|---|---|
| Can we state our own licence position today? | The next audit is on the vendor's numbers |
| Do we see third-party software, or only the OS? | You are patching the wrong landscape |
| What is end-of-life right now? | Obsolescence is found in incidents, not scheduled |
| Can we find the AI we have deployed? | You cannot secure or retire what you cannot see |
| Do our three functions share one inventory? | The single-landscape benefit is only theoretical |
| Who operates it after signing? | The landscape view ages into shelfware |
One accountable team, from strategy to shipped software
Volge delivers software asset management on the Flexera platform, or as a manual SAM review, through one team and one arc. Four roles hold the work from the first workshop to operational handover.
One SAM approach, delivered tool-based or manual
The defence does not change with the publisher: the same lifecycle runs in five phases, and only the delivery differs, continuous on the Flexera platform, or periodic as a SAM review inside the audit practice.