TL;DR
- IT asset management software keeps a record of what hardware you own, who has it, and what state it is in. The record is a claim about the physical world, and it starts going out of date the moment it is written.
- If you have 30 machines in one office and a person who can walk round and look, you do not need this. You need a spreadsheet somebody maintains.
- Three kinds of tool exist and they know different things. Discovery tools know what connects to your network. Register tools know what somebody typed in. Lifecycle platforms know what they themselves shipped.
- No single one of them sees the whole fleet, and the gap between them is where missing laptops live.
- Pricing runs from genuinely free to five figures a year. Snipe-IT is free self-hosted; the lifecycle platforms publish nothing at all.
- Measure register accuracy, not feature count. If you cannot say what proportion of your records are correct today, the tool is not your problem yet.
The Audit That Found 14 Laptops Nobody Had Lost
A 260-person company had asset management software. It had been bought two years earlier, configured properly, and populated with every device the company owned at the time. The IT lead could produce a report in about four seconds.
Then an insurer asked for a schedule of equipment and somebody checked the report against reality. Fourteen machines in the system had no current holder. Eleven of those belonged to people who had left. Three had been reassigned internally and never updated. Separately, nine laptops that definitely existed, because people were working on them, were not in the system at all, having been bought on a card by a team lead during a hiring push when procurement was slow.
Nobody had done anything wrong, and the software had worked exactly as designed. It had faithfully stored what it was told and had no mechanism for noticing what it was not told. An asset register is not a record of what you own. It is a record of what somebody remembered to tell it, and the distance between those two things is the entire problem. Software reduces that distance. It does not close it, and a tool bought in the belief that it will is a tool that will disappoint quietly for two years.
Best tools for Device Management
This is what IT asset management software is supposed to solve.
When You Don't Actually Need IT Asset Management Software
When the manual way is genuinely fine
Under about 30 machines in one or two locations, with somebody who knows where they all are. A maintained spreadsheet is accurate at that size and a database is not more accurate, it is only more impressive. Spend the effort on keeping the sheet current rather than on migrating it.
When friction starts appearing
The signal is the first question you cannot answer in under a minute. How many machines are out of warranty. Who has the laptop with that serial. What did we issue to the person who left in June. One of those going unanswered is a bad afternoon. Three of them is a system problem, and it typically appears between 60 and 120 devices.
When it becomes a liability
The point at which somebody outside IT needs the number. An insurer asking for a schedule. An auditor asking about device-level access. A customer security review asking how you track company hardware. At that moment an inaccurate register is worse than an obviously absent one, because it produces a confident answer that is wrong.
The edge case that forces it
An acquisition, or a sudden expansion into remote hiring. Both double the fleet without doubling the attention paid to it, and both introduce devices that were never in your process. Companies that adopt asset management after one of these events spend most of the first year cleaning rather than tracking.
Five Questions People Ask Before Buying
"How many devices do we actually have?" Most teams can produce a number and cannot defend it. The gap between the register and reality is the thing you are buying a tool to reduce.
"Will it find devices automatically?" Only the ones that connect to something it can see. A laptop that has not been switched on for four months is invisible to every discovery tool there is.
"Who keeps it up to date?" A person, for the parts no system observes. The tools that work best are the ones that need the least of this.
"Can it tell us what is out of warranty?" If it holds purchase dates, which depends on whether anybody entered them, which depends on where devices were bought.
"What does it cost?" Anywhere from nothing to five figures a year, and the spread is real rather than a function of quality.
The Three Kinds of Tool, and What Each One Is Blind To
This is the distinction that decides the purchase, and it is rarely made in vendor material because every category claims to cover the others.
Discovery tools
What it is. Software that finds devices by scanning the network or by an agent installed on each machine, and builds the register from what it observes.
When it's right. When most hardware is on a network you control, and you want hardware and software inventory without anybody typing. It is very good at specification, operating system version, installed applications and patch state.
When it fails. It cannot see what does not connect. A laptop in a drawer, a machine held by somebody who left, a device switched off for a quarter: all invisible, and all exactly the devices you most need to find. Discovery tools are strongest on the part of the fleet that is least at risk.
Register tools
What it is. A database of assets with people, locations, purchase dates and status, updated through forms, imports or an API.
When it's right. When you need a record of ownership and history rather than live technical state, particularly for insurance, audit and finance. It will hold things no scanner can know, like who signed for a device and what it cost.
When it fails. It degrades in proportion to how much manual updating it needs. Every reassignment, every leaver, every machine bought outside the process is a chance for the record to drift, and nothing in the tool notices when it has.
Lifecycle platforms
What it is. The register is a by-product of the platform having procured, shipped and retrieved the device itself.
When it's right. When devices are distributed and the platform handles the movement. The record stays accurate because it is generated by the actions rather than reported after them, which removes the main source of decay.
When it fails. It knows its own devices. Hardware bought before you adopted it, or bought elsewhere since, is invisible unless somebody enters it by hand, at which point it is a register tool with the same drift as any other. Companies with a large existing fleet should expect a migration rather than a switchover.
| Kind of tool | Knows | Blind to | Decays because |
|---|---|---|---|
| Discovery | Anything that connects | Switched off, stored, held by leavers | Nothing, it just cannot see them |
| Register | Whatever was entered | Anything nobody entered | Humans forget to update it |
| Lifecycle platform | What it shipped itself | Devices it did not ship | Only at the edges, via manual entries |
So the honest answer for most companies with a distributed fleet is that they need two of these, and the pairing that works is a lifecycle platform for the devices it moves plus discovery for technical state. Running three is usually somebody's unfinished migration.
Measuring Register Accuracy, Which Is the Only Number That Matters
Feature comparisons are a poor way to choose here, because every tool in the category can store a serial number. What separates them is how fast your data goes wrong. So measure that first, on whatever you have now.
Take a random sample of 30 records from your current system, however informal it is. For each one, establish two things independently of the record: does the device exist, and does the person named still have it. Phone calls and messages are fine; this is not meant to be elegant.
Then compute the proportion correct. That is your register accuracy, and it is almost always lower than the team's estimate.
| Accuracy | What it means | What to do |
|---|---|---|
| Above 95 per cent | Your process works | Do not buy a tool, protect the process |
| 85 to 95 per cent | Normal drift | A tool helps, pick for low manual upkeep |
| 60 to 85 per cent | The record is a rough guide | Fix intake and offboarding before buying |
| Below 60 per cent | Not a register, a historical document | Rebuild from purchasing, then choose a tool |
The rows at the bottom matter most, because a company at 55 per cent accuracy that buys software will import 55 per cent accuracy into a new system and conclude the software failed. Clean, then migrate. Not the other way round.
And take the sample again 90 days after you implement anything. The difference between the two numbers is the only evidence that will tell you whether the purchase worked.
Where Registers Actually Go Wrong
Four events cause nearly all drift, and knowing them tells you which tool features are worth paying for.
Internal reassignment. Somebody changes team and their laptop goes with them, or does not. This is the single most common cause of a record naming the wrong person, and almost nothing catches it automatically.
Purchases outside the process. A team lead buys three machines on a card because procurement was slow. Those devices are real, in use, and in no system. Discovery will find them if they connect to something you monitor; otherwise they surface during an audit.
Leavers. The record says the device is with a person who no longer works here. This is the drift that matters most, because it is also a security and insurance gap rather than only an accuracy one.
Repairs and swaps. A machine goes away and a loaner arrives. Two records are now wrong, and the loaner is frequently never recorded at all.
What to do:
- Tie the register to the leaver process so a departure forces a device decision.
- Make intake the only route to a company device, and make it faster than a card purchase.
- Record loaners as assets, not as a temporary arrangement somebody remembers.
- Reconcile against purchasing quarterly, because finance has invoices for everything that was bought properly.
- Sample 30 records a quarter and track the accuracy figure over time.
The Fields That Earn Their Place
Every tool in this category will let you store 60 fields per asset, and every company that enables 60 fields ends up with 50 of them empty. Empty fields are worse than absent ones, because a column that is populated for a third of the fleet cannot be reported on and quietly teaches people that the register is unreliable.
Ten fields carry almost all the value. Start with these, insist they are complete, and add others only when somebody asks a question you genuinely cannot answer.
Serial number. The only identifier that survives reassignment, repair and renaming. A register keyed on anything else is a register you cannot match against a physical device.
Current holder. A person, linked to your directory rather than typed as free text, so that a leaver shows up as a leaver without anybody updating the asset.
Status. In use, in stock, with a repairer, retired. Four values are enough and more than four invites inconsistency.
Location, at country level. Not a desk. For a distributed fleet, country determines shipping, duty, warranty cover and whether retrieval is even practical.
Purchase date and price. Finance and insurance both need these, and they are the fields most often missing because the device was bought outside the process.
Supplier. Decides who you talk to about a warranty claim, and reveals how much hardware entered outside procurement.
Warranty expiry. The one field that lets the register answer a forward-looking question rather than a historical one.
Assigned cost centre. Makes the register useful to somebody other than IT, which is what keeps it funded.
Condition at last handover. A single line recorded when a device changes hands. Without it, damage disputes are unresolvable.
Date the record was last verified. The most underrated field in asset management. It lets you sort by staleness and tells any reader how much to trust a given row.
| Field | Answers | Usually missing because |
|---|---|---|
| Serial number | Which physical device is this | Nobody recorded it at handover |
| Current holder | Who do I ask | Typed as text, not linked to the directory |
| Purchase date and price | What is it worth, is it due a refresh | Bought outside procurement |
| Warranty expiry | What breaks next and who pays | Never imported from the supplier |
| Last verified date | How much should I trust this row | The field does not exist in most setups |
So configure 10 fields properly rather than 40 partially. A register where every row is complete is a register people use, and use is what keeps it accurate.
Reading the Disagreements Between Your Systems
By the time a company is large enough to want asset management software, it already has three or four partial pictures of the fleet, and they disagree. The disagreements are not noise. Each pattern means something specific, and learning to read them is faster than any migration.
You usually hold four sources. Purchasing knows what was bought. Your identity provider knows who is active. Your MDM knows what is enrolled. The register knows what somebody typed. Compare them in pairs once a quarter and the gaps tell you where your process leaks.
In purchasing, not in the register. Hardware entered the company outside your intake process. Normally a card purchase during a hiring push. The fix is making intake faster, not scolding the buyer.
In the register, not in the MDM. A device that is recorded but not enrolled. Either it was never set up properly, or it is not in use. Both are worth knowing and the second is a recovery candidate.
In the MDM, not in the register. A machine is being actively managed that your asset record does not contain. This is the most alarming gap and the most common in companies that grew quickly, because it means a device is in use, holds company data, and is financially invisible.
Register says a leaver still holds it. The highest-priority row in any reconciliation, because it is simultaneously an accuracy problem, a security exposure and an insurance misstatement.
In no system, but somebody is using it. Found only by asking people. This is why a periodic confirmation message to staff is worth more than it sounds.
| Appears in | Missing from | What it means | Priority |
|---|---|---|---|
| Purchasing | Register | Bought outside intake | Medium, fix the process |
| MDM | Register | Live device, financially invisible | High |
| Register | MDM | Unused, or never configured | Medium, check for recovery |
| Register, held by a leaver | Reality | Security and insurance gap | Highest |
And run the comparison as a quarterly routine with a named owner rather than as a project. A reconciliation that happens four times a year at 85 per cent effort beats a perfect one that happens once and is never repeated, because the value is in the trend rather than in any single snapshot.
How to Choose: Five Questions Before You Talk to Any Vendor
What is your register accuracy today? Run the 30-record sample. A vendor conversation held without that number becomes a feature discussion, and features are not what is wrong.
Where do your devices physically live? A fleet on a corporate network suits discovery. A fleet in people's homes in nine countries does not, and the tools that serve it well are the ones that move hardware rather than scan for it.
Who will do the manual updating? Name them, and be realistic about how much of their week this is worth. The honest reason most registers decay is that updating them is nobody's actual job, and that does not change because the interface improved.
What do you need to produce, and for whom? An insurance schedule, an audit trail, a warranty report and a security attestation need different fields. Write down the output you are asked for, then check the tool produces it without manual assembly.
Do you already own something that does this? Your MDM holds a device list. Your identity provider holds one. Your purchasing system holds invoices. A surprising number of companies buy asset management software to produce a view they could assemble from systems already paid for, and it is worth an hour to check before a procurement cycle.
A Word About Software Licences
Everything above concerns hardware, and the term IT asset management covers software licences too. The two are routinely sold together and they are different problems, so it is worth being explicit about which one you are solving.
Software asset management asks who holds a licence, whether you are over-deployed, and what you are paying for seats nobody uses. It is answered mostly by discovery tools and by the administrative consoles of the applications themselves, and the money at stake is usually larger than the hardware number because unused seats renew silently every year.
Hardware asset management asks where the physical devices are. It is the problem this article addresses, and it is the one that produces audit findings, insurance gaps and unreturned laptops.
A tool strong at one is frequently weak at the other. Discovery platforms see installed software well and physical custody badly. Lifecycle platforms see custody well and have no view of what is installed. If both matter to you, expect two tools, and decide which is urgent rather than buying something that claims both and does each adequately.
But do not let a software licence problem be solved by a hardware purchase, or the reverse. They share a name and almost nothing else.
Six Options Worth Knowing
Prices below were read from each vendor's own page on 6 and 7 October 2026, with the date noted, and where a vendor publishes nothing that is stated rather than estimated.
Snipe-IT
Best for: companies that want a real asset register without a procurement cycle, and teams happy to run their own software.
Why companies choose it: the self-hosted version is open source and free, permanently, which makes it the honest starting point for anybody whose problem is that they have no register at all. The hosted versions are published and legible: $399.99 a year or $39.99 a month for Basic, $999.99 a year or $99.99 a month for Small Business, and $2,499.99 a year or $249.99 a month for a small dedicated instance, with medium and large dedicated tiers at $5,000 and $7,500 a year. In a category where most vendors publish nothing, that transparency is worth something on its own.
Where it struggles: it is a register, so it knows what you tell it and drifts exactly as fast as your process allows. The free version requires somebody to run and update a server, which is a real cost even though it is not an invoice. There is no discovery and no logistics.
Lansweeper
Best for: fleets largely attached to a network you control, where technical inventory matters as much as ownership.
Why companies choose it: discovery is the strength. It builds the picture by looking rather than by asking, which removes the largest source of register decay for on-network hardware and gives genuinely useful detail on specification and software estate.
Where it struggles: it publishes no price. Checked on 7 October 2026 in a browser, the pricing page carries no figures at all; there is a free tier and everything above it routes to sales. And the structural limit applies: devices that are switched off, in storage or held by former employees are invisible, which is precisely the population an audit asks about.
RemoAsset
Disclosure: RemoAsset is owned by the same people who publish HROpsLab. It is listed here because it competes in this category and because what it cannot see is the most useful thing to understand about it.
Best for: distributed teams where the register's accuracy problem is caused by devices moving between countries and people.
Why companies choose it: the record is a by-product of the platform having bought, shipped and retrieved the device, so it does not depend on anybody updating anything. Offboarding triggers a return and the record changes because the device moved, not because somebody filed a form. For a fleet in people's homes that removes the main cause of drift.
Where it struggles: it knows the devices it handled. A laptop bought three years ago through a reseller is not in it unless somebody enters it manually, and at that point those records drift like any other register's. It publishes no price and requires a demo. It performs no network discovery, so it will not tell you what is installed on a machine or whether it is patched, and it is not an MDM. It is also not a certified disposition vendor, so end-of-life certification is a separate arrangement.
Workwize
Best for: multi-region fleets where storage and reissue are part of the tracking problem.
Why companies choose it: regional warehousing means the register reflects devices in stock as well as devices with people, which matters when reissue is a real workflow rather than an aspiration.
Where it struggles: no published price, and the same blind spot as any lifecycle platform regarding hardware it did not supply.
Deel IT
Best for: companies already running Deel, who want equipment records attached to employment records.
Why companies choose it: the joining and leaving signals already exist in the system holding the employment relationship, which is where most register drift originates.
Where it struggles: quote-based, and most compelling as an extension of an existing Deel relationship rather than as an asset management decision on its own merits.
Firstbase
Best for: companies tracking whole home setups rather than laptops alone.
Why companies choose it: desks, chairs and peripherals are in the register as well as computers, which matches what was actually issued in many companies and is awkward to track anywhere else.
Where it struggles: no published price, and the broader scope is only an advantage if you genuinely need to track furniture.
What Each One Published
| Option | Published price | Unit | Kind of tool |
|---|---|---|---|
| Snipe-IT, self-hosted | Free, open source | n/a | Register |
| Snipe-IT, Basic hosted | $399.99 or $39.99 | per year, or per month | Register |
| Snipe-IT, Small Business | $999.99 or $99.99 | per year, or per month | Register |
| Snipe-IT, Dedicated | $2,499.99 to $7,500 | per year | Register |
| Lansweeper | Not published, free tier exists | n/a | Discovery |
| RemoAsset | Not published, demo required | n/a | Lifecycle |
| Workwize | Not published | n/a | Lifecycle |
| Deel IT | Not published | n/a | Lifecycle |
| Firstbase | Not published | n/a | Lifecycle |
Read from each vendor's own pricing page, Snipe-IT and Lansweeper on 7 October 2026 and the lifecycle platforms on 6 October. The pattern is worth noticing: the register and discovery tools publish, and the platforms that move physical hardware do not, because their cost depends on geography and volume.
The Decision Table
| Situation | Scale | Setup | Primary Pain | Recommended Starting Point |
|---|---|---|---|---|
| Everyone in one office, somebody knows | Under 30 | Single site | Nothing is wrong | A maintained spreadsheet |
| No register at all, no budget | Any | Any | Starting from nothing | Snipe-IT self-hosted, free |
| Register exists, accuracy below 60 per cent | Any | Any | Data, not tooling | Rebuild from purchasing first |
| Devices on a network you control | 100 plus | Office or hybrid | Technical state and software estate | A discovery tool |
| Devices in homes across several countries | 50 to 500 | Remote | Drift from movement | A lifecycle platform |
| Need an insurance or audit schedule | Any | Any | Defensible output | A register tool with purchase data |
| Already running Deel for employment | Any | Remote | Joiner and leaver signals | Deel IT, inside the existing relationship |
| Tracking furniture as well as laptops | Any | Remote | Scope beyond computers | Firstbase |
Most companies need a register row and a movement row at once, which is why the common end state is a lifecycle platform for distributed hardware plus discovery for whatever sits on the network. That is two tools and it is not a failure of planning.
What Getting This Wrong Costs
The obvious cost is the licence, and it is trivial next to the real one. The expensive outcome is a register everybody trusts and nobody has checked, because a confident wrong answer travels further than an admitted gap. It goes into an insurance schedule, a security questionnaire and a board paper, and each of those is harder to correct than the original record was.
The second cost is the migration you did not plan. Importing an inaccurate register into better software produces the same inaccuracy with a nicer interface, and the team concludes within a year that the tool was oversold. What actually happened is that the cleaning step was skipped because it is tedious and nobody wanted to own the number it would produce. Cleaning first is slower to start and is the only version that works.
The third is paying twice for the same view. Your identity provider knows which devices authenticate. Your MDM knows which are enrolled. Purchasing knows what was bought. A company buying asset management software to answer questions those three could already answer between them is buying a report, and an hour spent checking that before procurement has saved companies a five-figure annual commitment more than once.
So ask the diagnostic question plainly. Is this a data problem, a visibility problem, or a movement problem? A data problem wants cleaning and a process, not a purchase. A visibility problem wants discovery. A movement problem, which is what most distributed teams actually have, wants a platform that generates the record by doing the work rather than by asking somebody to report it.
When You're Ready to Move Beyond a Spreadsheet
The signals are specific. You ran the 30-record sample and the number was uncomfortable. Somebody outside IT has asked for a schedule you could not produce quickly. Devices are moving between countries rather than between desks. And there is a named person whose job includes keeping this current, rather than an assumption that everybody will.
When those are true, the choice is mostly determined by where your hardware physically lives, and the feature comparison matters far less than the question of how much manual updating the tool requires to stay honest.
If the drift in your register is caused by devices moving, RemoAsset generates the record from the movement itself, which is why it suits distributed fleets and why it is the wrong answer for an office fleet already covered by discovery. Its limit is equally specific: it knows what it shipped, so an existing fleet bought elsewhere needs importing and then needs maintaining like any other register. It is worth a look alongside the alternatives here, and expect a demo rather than a price.
And if you have no register at all, start with the free one. Snipe-IT self-hosted costs nothing and will tell you within a quarter whether your problem is the tool or the process, which is a cheaper way to find out than a procurement cycle.
Frequently Asked Questions
What is IT asset management software?
IT asset management software maintains a record of the hardware an organisation owns, including who holds each device, where it is, what it cost, when it was bought and what state it is in. Some tools build that record by scanning a network and observing what connects, some by storing what people enter through forms and imports, and some by generating it as a by-product of having physically procured and shipped the device. The distinction matters more than any feature list, because it determines what the system can know and, more importantly, what it is structurally blind to.
What is the difference between asset discovery and an asset register?
Discovery finds devices by looking at your network or through an agent on each machine, so it produces accurate technical detail about anything that connects and knows nothing about anything that does not. A register stores what somebody told it, so it can hold information no scanner could ever observe, such as who signed for a device, what it cost and which department owns the budget, at the cost of degrading whenever a change is not recorded. Neither sees the whole fleet. The devices most likely to cause a problem, those switched off, in storage, or held by former employees, are exactly the ones discovery cannot see and a register will list inaccurately.
How much does IT asset management software cost?
The range is genuinely wide. Snipe-IT's self-hosted version is open source and free permanently, with hosted tiers published at $399.99 a year or $39.99 a month for Basic, $999.99 a year or $99.99 a month for Small Business, and $2,499.99 a year or $249.99 a month for a small dedicated instance, rising to $5,000 and $7,500 a year for larger dedicated hosting. Lansweeper publishes no figures and has a free tier with everything above it quoted by sales. The lifecycle platforms, RemoAsset, Workwize, Deel IT and Firstbase, publish nothing at all. All checked against the vendors' own pages on 6 and 7 October 2026.
Is there a free IT asset management tool?
Yes, and for a company with no register at all it is the sensible starting point. Snipe-IT's self-hosted edition is open source and free with no device limit, which means the only cost is running the server and maintaining it, and that cost is real even though it never appears on an invoice. Lansweeper also offers a free tier, though it publishes no pricing for anything above it. Starting free is particularly useful because it tells you within a quarter whether your difficulty is the absence of a tool or the absence of a process, and those have very different solutions.
Can asset management software find laptops we have lost track of?
Only in the specific sense that discovery tools will report any device that connects to a network they can see, which will surface machines bought outside your process and still in use. What no tool can do is find a laptop that is switched off, sitting in a drawer, or in the possession of somebody who left eight months ago, because there is nothing to observe. Those devices are found by reconciling your register against purchasing records and against your identity provider, which is an exercise rather than a feature, and it is usually how companies discover the true size of the gap.
How often should an asset register be checked?
Sample it quarterly rather than attempting a full audit annually, because a full audit is a project that gets deferred and a sample is an hour that actually happens. Take 30 records at random, verify independently that each device exists and that the named person still holds it, and record the percentage correct. Tracking that single number over time tells you whether your process is holding or degrading, which is far more useful than a complete count taken once a year and out of date by the time it is finished.
Does an MDM replace asset management software?
No, though it overlaps enough to cause confusion and occasionally to prevent a purchase that was not needed. An MDM knows about devices enrolled in it and can report their configuration, compliance and patch state in detail, which is genuinely valuable and covers part of what people want from asset management. What it does not hold is the commercial record: purchase price, supplier, warranty expiry, assigned cost centre, and the history of who held the device before the current person. If the questions you are being asked are financial or contractual rather than technical, an MDM will not answer them.
Should we clean our data before or after buying a tool?
Before, without exception, and this is the most common and most expensive sequencing mistake in the category. Importing an inaccurate register into new software produces the same inaccuracy in a better interface, and the team concludes the tool underperformed when what actually happened is that it faithfully stored bad data. Rebuild the record from purchasing invoices and your identity provider first, accept whatever number falls out as the honest starting position, and only then choose a tool. The cleaning is tedious and it is the part that determines whether the purchase was worth making.
HROpsLab takes no vendor money and publishes no paid placements, which is why the free option is listed first in the table rather than last.