●cybersecurity

What It Costs to Outgrow Your IT Asset Inventory Tool

AR by Anneliese Roth published 01.09.2026 read time 10 min

The renewal quote is the smallest number on the bill. A line-by-line account of what actually happens to a UK IT budget when the estate grows past the tool that was sized for it, and why per-asset pricing quietly rewards you for looking at less of your own network.

Most teams don’t discover they have outgrown their asset inventory tool when the renewal quote lands. They discover it on a Tuesday, when somebody notices the inventory hasn’t updated since Friday.

Lansweeper’s own documentation is blunt about what happens at the ceiling. Once you reach your licensed asset limit, “scanning of new and existing assets may stop until you delete some assets or upgrade your subscription.” Not just new assets. Existing ones. The tool you bought to tell you what you own stops telling you what you own, and it does it at exactly the moment your estate is changing fastest.

That is the first cost of outgrowing a tool, and it appears on no invoice. What follows is an attempt to put numbers on the rest of them. There is a great deal of published content comparing asset inventory products; there is almost none modelling what the transition between them costs. This is that model.

How the figures were arrived at

Prices are vendor list prices taken from public pricing pages and checked on 1 September 2026. Labour costs are modelled from published UK salary data, with the arithmetic shown so you can substitute your own. Hour estimates in the worked example are estimates, and are labelled as such. No figure here comes from an unsourced industry statistic.

Cost 01: The licence step-up, which is the part everyone budgets for

Lansweeper is a fair reference point because it publishes its prices, which much of the category does not. The structure, from its own pricing page: a free tier up to 100 assets; Starter from £219 a month billed annually, including 2,000 assets; Pro from £399 a month, starting at 2,000 assets and expandable to 9,000 in bands of 1,000; Enterprise starting at 10,000 assets on custom pricing, in bands of 2,000.

Lansweeper list pricing, checked 1 September 2026

Plan	Assets included	Per month, billed annually	Per year
Free	Up to 100	£0	£0
Starter	2,000	£219	£2,628
Pro	2,000, expandable to 9,000	£399	£4,788
Enterprise	From 10,000	Custom	Custom

One caveat if you check this yourself: the pricing page shows three figures per tier as a currency comparison, so alongside £219 you will also see 199 and 239 in euros and dollars. The GBP prices above are the ones a UK billing address is quoted.

Look at the gap between the first two rows. An organisation that grows from 95 assets to 130 does not buy 130 assets’ worth of tool. It buys 2,000, an entitlement roughly fifteen times what it needs. Nothing is wrong with that; it is how tiered pricing works in every category. It is worth naming only because it is the first of several places where your bill moves in steps while your estate moves in ones.

This is also where almost every published comparison of asset inventory tools stops. The licence is the visible cost and the easy one to put in a table. On the modelling below, it is somewhere between a fifth and a third of what outgrowing a tool actually costs.

Cost 02: The asset count you are billed on is not the asset count you have

Lansweeper’s documentation lists what counts toward the licensed limit: computers, servers, network devices, OT devices, OT cards, public cloud workloads and manually created assets. Monitors without extended display data do not count.

Read that list again with a cloud estate in mind. Public cloud workloads count. An autoscaling group that spins up eleven instances on a busy Thursday has, for licensing purposes, added eleven assets. The test environment somebody stood up in March and never tore down counts. So does the virtual machine you decommissioned in January, if nobody deleted its record, because the record is what the tool counts, not the machine.

This is the difference between the assets you operate and the assets your tool has ever seen. The second number only ever goes up, unless somebody is tasked with making it go down. In most teams, nobody is.

Cost 03: Per-asset pricing quietly penalises an accurate inventory

Here is the structural problem underneath all of this, and it is the part nobody writes about.

The purpose of an asset inventory is to be complete. The pricing model of most asset inventory tools is a function of how many assets it holds. Those two things pull in opposite directions.

Every improvement in coverage is an increase in cost. Widening the scan scope to the OT segment, adding the second cloud account, picking up the network hardware, discovering the estate the acquisition brought with it: each of these is the tool doing its job better, and each of them moves you toward a band. Meanwhile the only lever that reduces the count is deletion, and deletion is unglamorous hygiene work that competes with everything else in the queue and produces no visible benefit when it succeeds.

None of this is a vendor conspiracy. Charging by asset is a defensible way to price a tool whose cost of delivery scales with assets, and the alternatives have their own distortions. But the incentive is real, and it is worth recognising before you find yourself acting on it. If your team has ever narrowed a scan scope in the weeks before a renewal, you have already priced your own visibility.

The test

Ask whoever owns the tool a single question: if we discovered 400 more assets tomorrow, what would happen to our bill? If the answer is “we’d move up a band,” then your discovery scope is a budget decision, and it should be made deliberately and in the open rather than by default at renewal time.

Cost 04: Credentials, an operating cost dressed as a configuration task

Agentless discovery works by authenticating to things. WMI credentials for Windows, SSH keys for Linux, SNMP community strings or v3 credentials for network hardware, API keys for each cloud account, a read-only service account in the hypervisor, another for the directory.

Every new segment or platform brought into scope adds credentials, and credentials are not fire-and-forget. They rotate. They expire. They need documented scope and least-privilege justification. They appear in every audit you will ever sit. The count grows with the estate, and the maintenance grows with the count.

I have not found a published figure for what this costs, and I would distrust one if I had, because it varies far too much by environment. But it is real, recurring labour, and it belongs in the model rather than in the “that’s just admin” bucket where it usually sits unpriced.

Cost 05: Scan windows have a growth rate too

Credentialed discovery across 800 devices fits comfortably into the small hours. The same discovery across 4,000 devices, over more segments, with more authenticated checks per device, does not necessarily fit anywhere.

When the window stops fitting you have three options, and each costs something. Run scans during business hours and accept the load on the network and the endpoints. Narrow the scope and accept worse data. Or add infrastructure, such as a second scanning server or distributed collectors, and accept both the spend and another thing to maintain.

Teams almost always take the middle option, because it is the only one that is free at the point of decision. It is not free. It moves the cost off the budget and into the data, where it is invisible until the day someone needs the data to be right.

Cost 06: The agent estate, if you have one

Installing an agent on 500 endpoints is a project. Keeping an agent healthy on 4,000 is a job.

Agents drift out of date. They fail silently on machines that were rebuilt. They conflict with endpoint detection tooling and get excluded by an exception nobody documented. They stop reporting on the laptop that has been in a drawer since April. The install cost is one-off and gets budgeted; the maintenance cost is continuous and does not.

The failure mode is the expensive part. An agent that has quietly stopped checking in does not appear in the inventory as a gap. It appears as a device whose details have not changed, which looks exactly like a device that is fine.

Cost 07: The switching bill

If the conclusion is that you should move, here is what moving costs. Almost none of it is the new subscription.

Parallel running. You do not cut over an inventory in an afternoon. You run both tools for a period, typically a quarter, while you satisfy yourself that the new one sees everything the old one saw. Two licences, one estate.

Reconciliation. The two tools will disagree, and every disagreement is a judgement call made by somebody who knows the estate. This is the line item that consistently gets underestimated, because it scales with how messy the incumbent data already was.

Integration rebuild. Everything downstream that consumes the inventory is pointed at the old tool’s schema and API: the service desk’s CMDB, patch management, the SIEM’s asset context, procurement’s licence reconciliation. Each is a small piece of work. There are always more of them than anyone remembers.

Retraining. Not only the administrator. The first-line analyst who looks a device up mid-call, the security analyst checking an asset’s exposure, the procurement lead pulling a licence count at year end.

Putting a UK rate on the hours

ITJobsWatch puts the UK median salary for a systems administrator at £45,000, drawn from 330 salaries quoted in the six months to 1 September 2026. That splits into £56,750 in London and £43,500 outside it. Employer national insurance for 2026/27 runs at 15% above a £5,000 secondary threshold, adding £6,000. The minimum auto-enrolment employer pension contribution is 3% of qualifying earnings, which on that salary is about £1,160. So the post costs the business roughly £52,200.

Spread that over about 1,740 working hours (232 days after holiday and bank holidays, at seven and a half hours each) and you get £30 an hour, fully loaded. That is the number to multiply the estimates below by, or to replace with your own.

Modelled switching cost, 3,000-asset UK organisation

Line item	Hours	Cost
Parallel running, one quarter of duplicate licence	n/a	£1,197
Discovery scope design and credential setup	40	£1,200
Data reconciliation and cleanup, 3,000 records	80	£2,400
Integration rebuild, four consuming systems	60	£1,800
Retraining and documentation	24	£720
Total, excluding the new subscription	204	£7,317

This is a model, not a case study. No real organisation was measured, and the hour figures are estimates offered so you can argue with them. The point is the shape rather than the precision: the cost of switching is larger than the first year of the thing you are switching to.

That is the number to take into a budget conversation. If you present the subscription alone and the project then runs eighty hours over, the overrun reads as a failure of delivery. It was arithmetic that nobody did.

So how do you actually decide?

Two questions matter, and neither one is “which tool has more features”.

First: is the growth structural or a one-off? If your asset count jumped because of an acquisition or a cloud migration that has now finished, the cheapest answer is very often to buy the next band, delete your ghost records and stay where you are. If the growth is structural, you will pay the switching cost eventually, and paying it at 3,000 assets is materially cheaper than paying it at 8,000, because every line item above scales with the size of the estate you are reconciling.

Second: what basis of charge do you want to be on in three years? This is the question a feature comparison cannot answer. Per-asset, per-user, per-site and flat-rate models produce very different bills at the same headcount, and the basis of charge is what determines whether you are having this same conversation again in two renewals’ time. A rundown of what the main alternatives to Lansweeper for larger estates charge, and on what basis, is a more useful starting point than a feature matrix. Read the pricing model first and the feature list second.

And whichever way you go: delete something. The highest-return hour in this entire exercise is the one spent removing assets that no longer exist. It lowers your band, it improves every downstream decision that reads the inventory, and it costs nothing but the hour.

The actual point

You do not outgrow an asset inventory tool because the tool got worse. You outgrow it because you got bigger, which is the outcome everyone in the building was working toward. The uncomfortable part is that growth arrives as a series of small increments and the bill arrives as a step, so the two never line up and the step always feels like a surprise.

The licence is the number on the invoice. The credentials, the scan windows, the agent estate, the parallel run and the four integrations nobody remembered are the number in your team’s calendar. The second one is bigger. Cost it before you argue about the first.

Method and limitations

Vendor pricing was taken from public pricing pages on 1 September 2026 and is subject to change; third-party pricing trackers quoted figures in USD and EUR that correspond to the same list prices shown here in GBP. Salary data is from ITJobsWatch’s rolling six-month sample and reflects advertised permanent roles, which tend to run above the market’s actual paid median. The switching-cost table is a model built from estimated hours, not from measured project data. Treat it as a structure to populate with your own numbers rather than as a benchmark. No vendor commissioned or reviewed this article.