We bought more asset management tooling than at any point in the discipline's history, and complete visibility fell to 36 percent. We bought more and saw less. The Readiness Test is the answer.
Platform pricing is quote-based and no vendor publishes a rate card, so the six-figure annual commitment, the implementation fraction, and the year-to-trusted-data estimate are drawn from purchase orders the host has signed and are offered as experience, not research. The CMDB failure-rate figure is flagged in the episode as an unsourced impression.
The complete transcript — 3,464 words. Search it, quote it, or send the relevant section to a colleague who needs to see it.
I'm Bill Van Nort and today I am finally answering the question I have been dodging since the very first episode. Here is a number that should bother you. Flexera runs an annual survey of asset management professionals. Two years ago, 47% of organizations reported complete visibility into their IT estate. Last year it was 43. This year, the 2026 report, 512 respondents worldwide, it is 36%.
Down 7 points in a single year. Down 11 across 3. Now over that same stretch, organizations bought more asset management tooling than at any point in the history of this discipline. More platforms. Bigger platforms. Better discovery. Better normalization. More automation. And lately, more AI stapled to all of it. We bought more, and we saw less. That is the entire subject of today's episode, because if tools created visibility, that line would be going the other way.
Good morning, good afternoon, or good evening, wherever you're listening from. This is the show where we take the unglamorous machinery of enterprise technology and make it make sense. Grab your coffee. Today we answer the tool question properly. Not just whether to buy one, but which kind, and when. Back in episode one, a listener named Mike from Columbus wrote in, his leadership had told him they could not afford an asset management tool, and he asked me what to do.
I said you cannot afford not to have one, but you also do not need one to start. Your endpoint platform already knows what is connected. Your procurement system already knows what you bought. Your directory already knows who works there. The gap between those three datasets is your first project, and it costs you nothing but attention. That answer was true.
It was also incomplete, and it has bothered me for nine episodes. Because the real question underneath Mike's question was never, can we afford a tool? It was, should we buy one? And how would we know? That is a completely different question. It has an actual answer. And today, Mike gets the full version. Let me lay out the evidence, and I will name the sources, because on this show, we do that.
Flexera's 2026 State of ITAM report, 512 Technology Professionals. Complete visibility, 36%, down 7 points year over year. Maintaining accurate inventory is the number one priority for 78% of asset management teams. 48% were audited in the past 12 months. 44% have spent over a million dollars on audits across three years, and that figure has now been essentially flat across three consecutive reports.
64% of audited organizations reported a Microsoft audit. And watch the movement underneath. Oracle audit activity jumped from 24% to 38% in a single year. Adobe went from 24% to 32%. Sit with that combination. Inventory accuracy is the top stated priority in the profession. Audits are constant, expensive, and spreading to more publishers. And visibility is going backwards anyway.
Now here is the statistic you have absolutely seen everywhere. 80% of configuration management database projects fail or add no business value attributed to Gartner. It is in vendor decks. It is on consultancy landing pages. I want to do to that number exactly what I told you to do in episode 7. Chase it back to its source. When you do, you find it quoted constantly and sourced almost nowhere.
The versions in circulation put the figure at 70%, at 80%, and at 85%. Several of the citations trace back to research that is more than 10 years old, and a striking number of the sites repeating it sell configuration management software. I am still going to use it today because the direction matches everything I have seen in three decades, but I am telling you exactly what it is.
That is an impression with a large number attached to it. It is not a measurement, and you should not put it in a business case. Alright, today's thesis, and it will not make me popular with a certain kind of sales team. A tool does not create a program. A tool amplifies a program. It multiplies whatever discipline you already have.
And if what you already have is close to zero, it multiplies that too, faster, at greater expense, with a very attractive dashboard on top.
So how do you know if you are ready? Five questions. I call it the readiness test. Answer these honestly, and you will learn more than any vendor demonstration will ever tell you. Question 1. Who owns the data? Not who administers the system. Who is accountable when a record is wrong? If a laptop assignment is incorrect right now, whose name is attached to that failure?
If the answer is a team, that is not an answer. If the answer is nobody, then a platform gives you a substantially larger volume of unowned data, generated faster and disputed by more people than before. Question two. When two systems disagree, which one wins? Every organization has several partial inventories that contradict each other. Procurement knows what was purchased.
Your endpoint platform knows what checked in. Finance knows what was invoiced. Security knows what its agent can reach. Each of those is answering a different question honestly, which is why they never agree. A platform does not resolve that. A platform ingests all four and hands you the disagreement at higher resolution and greater speed. Somebody has to decide the precedence rules.
That is a human decision, made before the purchase order, or it does not get made at all. Question 3. Can you produce entitlements as documents? Back to the ledger from Episode 3. Not what you believe you bought. The contracts, the order forms, the amendments, the proofs of entitlement, in one place, indexed by publisher. Every discovery tool on the market is genuinely good at the right-hand column, which is deployments.
Not one of them can invent the left-hand column for you. Entitlement evidence is paper you either have or you do not. If you do not, the platform will render you a beautiful compliance position built on half a ledger, and you will present it to somebody who has the other half. Question four. Does a lifecycle event change a record today without a tool?
Somebody leaves the company. Does a record change? Somebody swaps a laptop at the service desk. Does a record change? Remember the seven gates from episode two. Those gates are process, and process is either running or it is not. If custody changes today and the record does not follow, then automating that pathway automates something broken. You will get to wrong data sooner, and you will have paid for the privilege.
Question five. Can somebody write down what the tool is for in one sentence with a number in it, not improve visibility, not mature the program, something closer to reduce the time to produce a defensible Microsoft license position from 11 weeks to two before the March renewal. If nobody in your organization will put a sentence like that in writing with their name on it, the project has no owner and no definition of done.
18 months from now it will be evaluated against whatever people privately hoped for, and it will lose. That is the readiness test. Owner, precedence, entitlements, process, purpose. Fail any one and a platform makes your problems more expensive instead of more visible.
Now let me give you two pieces of evidence from the market itself. Because the market tells on itself if you listen carefully. First, there is an entire consulting sub-industry that exists to rescue asset management tool implementations. Not to implement them, to rescue them. Firms advertise it under exactly that name, "implementation rescue." Health checks for platforms that are live but not trusted.
That market exists because the failure mode is common enough to sustain it, and the language those firms use to describe the problem is remarkably consistent. The module is deployed. The data is not trusted. Nobody uses it to make a decision. Notice what is completely absent from that description. Anything at all about the software being bad. Second, look at what the vendors actually compete on.
Flexera acquired Snow Software in February of 2024, creating the largest dedicated software asset management vendor by revenue. And the capability the merged company leads with is its normalization catalog. Millions of software titles, versions, and license metrics maintained continuously. Their competitors argue with them in precisely the same terms. The industry's own marketing says the hard part is translating messy discovery output into clean statements about products, editions, and versions.
That is a data quality problem, and it is exactly the problem you can start working on this week, with a spreadsheet and one publisher, before you have spent a dollar. Let's take a quick break. If you're getting value from this show, subscribe wherever you're listening. And if you know somebody who is three months into a tool selection right now, send them this episode before they sign.
Every episode with full transcripts is at operationalitam.com.
Okay, part two, what to do instead if you just failed a question, or three. You are not going to love this, because the answer is much smaller than the thing you were about to buy. Pick one publisher. Your largest by spend is the obvious candidate. Give yourself 90 days and no platform. Before you start, do the single highest leverage thing in this profession, which almost nobody does.
Go to financial planning and analysis and get written agreement on three points. What counts as a hard saving? What baseline it will be measured against? And who validates that it was actually realized? Not you, somebody in finance. That conversation feels like bureaucracy, and it is the whole game. A function that is the sole judge of its own results will not be believed indefinitely.
Then produce one clean, fully evidenced result. One publisher. Entitlements you can prove. Deployments you have validated. One page. And report it in three separate lines, never as one aggregate. Hard savings, meaning reductions traceable to a lower invoice. Cost avoidance, documented against something external, like a vendor's written price increase notice. And risk, stated as exposure, never added to the other two.
A verified $400,000 reduction that finance has confirmed is worth more to your next funding request than a $4 million theoretical position that nobody outside your team believes. And you will find out within 90 days whether your organization can actually sustain the discipline. That is precisely the thing a tool cannot tell you, and precisely the thing you need to know before buying one.
Now, let's say you pass. All five questions, honest answers, names attached, the 90 days behind you. This is where I stop being the skeptic, because past that gate, the tools have a genuinely strong case, and I will make it fairly. Here is what the platforms actually do well, and I will use the market's own definition.
Gartner published its most recent Market Guide for software asset management tools in January of 2026, and it describes the core capabilities as discovery, normalization, reconciliation, optimization, and reporting. The first two are where the machines earn their money. A commercial normalization catalog tracks millions of titles, editions, and versions, updated continuously. No internal spreadsheet survives contact with that problem at enterprise scale.
And reconciliation on a platform is continuous, where your spreadsheet is a photograph. A license position you built in March is history by June. A platform rebuilds it every night. Then there is the case where the tool is not even a choice. And every listener with IBM in the estate needs to hear this. Under IBM's own Passport Advantage terms, sub-capacity licensing, meaning paying for the cores your software can actually reach instead of the full physical box, requires the IBM License Metric tool, or an approved alternative, to be installed and reporting.
Since 2022, manual counting is no longer permitted for those metrics. Run without the tool, and IBM's published remedy is charging you at the full capacity of the hardware. For that estate, the tool is not a maturity question, it is a license term. Now the other side, stated just as plainly. The platform cannot produce a single entitlement document, cannot decide precedents, and cannot own a record.
Those were questions one through three, and they are still yours. The data going in is the ceiling on everything coming out, and Gartner's same January guide notes that 61% of procurement leaders describe their own data as disorganized, inaccurate, or needing major quality improvements. Feed that to a platform and you get question two at machine speed. And the money.
This one I am flagging as experience because pricing is quote-based and nobody publishes a rate card. I have signed these purchase orders. At enterprise scale, expect a six-figure annual commitment. Expect implementation to cost a meaningful fraction of that again, and expect a year before the data is trusted enough to take into a negotiation. Budget all three, or budget none of it.
Forrester wrote back in 2013 that the worst outcome of under-investing in people is an expensive SAM tool full of data that is rarely used. That sentence is 13 years old, and the implementation rescue industry exists because it is still true.
So when is it actually time to buy? Four signals, and I would want at least two before signing. 1. You pass the readiness test. In writing. 2. A contract requires it, the IBM situation being the clearest case. 3. Scale has outrun hands. If you did Episode 9's homework and counted more than a handful of major agreements renewing inside 18 months, and your audit exposure spans Microsoft, Oracle, and Adobe at the rates Flexera is reporting, manual reconciliation is now costing more than the platform would.
4. There is a dated event where speed pays, which is exactly question 5's sentence. And when you do buy, buy the smallest thing that answers your question. This market sells full estate suites, modules native to your service management platform, SaaS management point tools, and publisher-specific depth. Those are four different purchases for four different questions. Match the tool to the sentence you wrote, not to the demo that impressed you.
Time for a listener question, and this one is the follow-up I was expecting. Dana in Dayton writes in, she says, we bought the platform two years ago. It is live. Nobody trusts the data, and I have now been asked whether we should rip it out. What do I tell them? Dana, do not rip it out. Opinion, flagged as an opinion, but it is a strongly held one.
Ripping it out is the same mistake running in reverse. Another large project driven by frustration instead of diagnosis, and in 18 months, you will be exactly here again with a different logo on the login screen. Do this instead. Run the five questions retroactively and be honest about which ones you would have failed two years ago. In my experience, it is usually two and one, in that order.
Nobody set precedence rules, so the platform is faithfully reporting a disagreement between four systems that were never reconciled. And nobody owns the records, so when the data is wrong, there is no one whose job it is to make it right. Neither of those is fixed by changing vendors. Both of them are fixable exactly where you are, with no migration and no capital request.
Then pick the smallest question your platform could answer correctly, and make it answer that one thing right. One publisher, one asset class, whatever is nearest to clean. Trust does not come back through a relaunch. It comes back through one number somebody checked and found correct, and then another one, and then another. And when you go back to leadership, do not tell them the tool failed.
Tell them the tool was asked to do a job that had no owner. Here is the owner now, and here is the first thing it will get right by the end of the quarter. That is a fundable sentence. We should replace it is not.
Now let's go back to the library, and today we are in the room this entire series has been walking toward, the catalog. For most of the 20th century, the catalog was cards. Wooden drawers, one card per work, filed by hand, maintained by people whose profession was descriptive standards. Then computers arrived, and every library in the world faced the same project.
Get the catalog into the machine. That project had a name, Retrospective Conversion, and it sorted libraries into two groups with almost brutal clarity. The libraries whose card records were consistent, complete, and maintained to a standard converted and came out the other side with something better than what they had, faster, searchable, shareable with every other institution in the network.
The libraries whose card records were inconsistent, incomplete, and years out of date converted too. And what they received was every one of those problems, now searchable, now shareable, now visible to absolutely everybody, and considerably harder to quietly fix than a drawer of cards had ever been. Same software, same project, same budget, completely opposite outcomes, decided entirely by the state of the records before anybody plugged anything in.
The machine did not make the catalog good. The cataloging did. The machine only made the answer arrive faster, whatever the answer already was.
Which brings us to today's principle. And let's take the whole run from the top. Hardware Asset Management is a custody discipline. Software Asset Management is an evidence discipline. Audit Defense is a process discipline. Settlement is a commercial discipline. Shadow IT is a service discipline. Refresh is an economics discipline. Disposal is a liability discipline. Renewal is a leverage discipline And tooling?
Tooling is a judgment discipline, not a technology decision. A judgment about whether the thing you are about to amplify is worth amplifying. Every other discipline on that list is something you do, continuously, forever. This one is something you decide, once, and then live with for three to five years. And I want to be fair here, because this is the episode where it would be easy not to be.
The vendors are not lying to you. The platforms mostly work. I have bought these, implemented them, and inherited failed implementations from people who bought them for the wrong reasons, and in none of those cases was the product the problem. The failure is buying an amplifier before you have a signal.
Class dismissed. Here's your homework, and this one takes about an hour. Take the five questions and answer them in writing about your own organization. Owner. Precedence. Entitlements. Process. Purpose. One honest paragraph each, and be harder on yourself than any vendor would ever be. Then, whichever ones you failed, write question 5 anyway. One sentence with a number in it, describing what a tool would actually be for.
Take that sentence to whoever would have to sign the purchase order, and ask them if they agree with it. If they change it, you have just learned what the project is really for. If they will not commit to it at all, you have just saved your organization a great deal of money and saved yourself about two years.
Next episode, we go after the newest asset class in the estate, and the first one to arrive with a regulator already attached. On the 2nd of August, days before this episode, the European Union's AI Act transparency obligations under Article 50 began to apply. The European Commission adopted its guidelines on the 20th of July, and fines run up to 15 million euros or 3% of worldwide turnover.
Meanwhile, in that same Flexera report, only 31% of organizations reported accurate visibility into their AI software at all. If you have ever wondered whether AI belongs in your asset program, somebody else has now answered that for you. One more thing before I go, the case files. I've asked for these, three episodes running, and I'm going to stop asking in a vague way, because vague asks get vague results.
Here's what I actually want. One situation, one page, what the constraint was, what you did, and what happened. Anonymized, sanitized, no company names. Details at operationalitam.com. Send me one worth working, and I will build an episode around it. I'm Bill Van Nort, this is the Operational ITAM Podcast. Answer the five questions, buy the amplifier last, and never automate a process that isn't running.
Talk to you next week. Take care.
Send it over — anonymized, sanitized, no company names. Real constraints, real politics, real budgets. Situations get worked on air.