How a LIMS Is Priced, and What Per-Seat Does to a Lab

What are the ways a LIMS is priced?#

Four things get counted, and most quotes combine two of them.

Seats. A price per named user, or per user logged in at once. It is the oldest model and the easiest to explain, and it is the one this article is mostly about, because it decides who is allowed in.

Modules. A base system plus a price for each capability switched on: instrument integration, stability, inventory, a customer portal. Cheap to start and expensive to grow, because the lab discovers which modules it needs after it has bought the base.

Samples or volume. A price per sample logged, per test run or per band of annual throughput. Common in cloud offers, and attractive to a small lab, but it turns growth in the lab’s own business into a rising software bill and makes a busy month more expensive than a quiet one.

Working groups. A price per team or department the lab is organised into, with no limit on the people inside each. Rarer, and the one that changes behaviour most, because adding a person costs nothing and the lab stops thinking twice about giving somebody access.

On top of whichever is counted sits the form of payment: a perpetual licence with annual maintenance, or a subscription that includes hosting and support. And beside all of it sit the costs that are not the software at all: implementation, data migration, instrument interfaces, validation and training. On most projects those exceed the first year’s licence, and the implementation is where the real money and the real risk sit.

What does a per-seat licence do to a lab?#

It rations access. When a licence is counted in seats, every login has a price, and someone in the lab has to decide who is worth one. The bench gets seats because the bench enters results. The stockroom does not, because the stockroom only receives deliveries. The QA reviewer gets a read-only login if there is budget this year. The customer never sees anything, because a customer seat is a seat that does not do billable work.

None of those decisions is unreasonable on its own. Together they produce a system that holds part of the lab’s work and a set of people around it who touch samples every day and are not in it. The stockroom keeps a receiving log on paper. The reviewer asks for a printout. The customer telephones. Each of those is a record kept somewhere the system cannot see, and each is a reconciliation somebody does at month end.

The rationing also gets worse over time, not better. A lab hires a technician and the software bill goes up, so the technician shares a login for the first six months. A second site opens and its people are given a single generic account “for now”. The seat count on the invoice stays flat while the number of people using the system doubles, and the audit trail records that four people called Lab User did everything.

Where do the records go when a person has no login?#

They go where they always went before the LIMS. To paper, because a receiving book on the bench is quicker than asking someone with a login to enter a delivery. To email, because a reviewer without access asks the analyst to send the result, and the analyst’s reply is now the record of the review. To a spreadsheet, because the field team has samples to schedule and no screen to do it on. And to a shared account, which is the subject of the next section.

The cost of this is not in the licence. It is in every question the system cannot answer because the answer was never in it. Where is the sample now? The system says received; the stockroom book says it was moved to cold storage on Tuesday. Who reviewed this result? The system says the analyst released it; the email thread says the reviewer approved it two days later. An auditor who finds the system and the paper disagreeing does not conclude that the paper is wrong.

The people kept out are, almost always, the people whose records matter most to traceability: those who receive, move, store and dispose of samples, and those who review and release results. The bench analyst’s work is the best-recorded part of any laboratory. It is the handoffs around it that go missing, and per-seat pricing is designed to keep the people at the handoffs outside the system.

Why is a shared login a compliance problem, not a saving?#

Because every regulation that governs electronic records assumes one account per person, and an audit trail that says “the lab account did it” answers nothing. The US rule on electronic records requires that access be limited to authorised individuals and that each electronic signature be unique to one person and not reused or reassigned. The European equivalent and the accreditation standards for testing laboratories ask for the same thing in their own words. A shared login satisfies none of them, whatever the licence says.

The saving is also smaller than it looks. A shared account saves one seat. It costs the lab the ability to answer who did anything, which is the question every deviation investigation, every customer complaint and every inspection begins with. When the answer is “one of the four people who use that login”, the investigation widens to all four, and the finding is not about the result. It is about the system of control.

One account per person is the lab’s obligation, not the vendor’s. But a pricing model that puts a price on each person is one that makes the obligation expensive to meet, and the lab that meets it anyway pays for compliance in seats. The model that counts something other than people removes the incentive to cut corners, which is why the model matters more than the number.

What should you ask a vendor about price?#

Five questions, asked before the quote, reveal the model and what it will do to the lab. The answers are more useful than the figure.

The fifth question matters because the first-year figure is the one vendors compete on and the one that least predicts the total. Implementation and validation effort usually cost more than the first year’s licence, and the module the lab did not know it needed usually arrives in year two. A quote that fixes the growth path is worth more than one that undercuts on the opening figure.

Count the people in your lab who touch a sample and are not in the current system. That number is the size of the records problem the pricing model has to solve, and the right model is the one under which that number goes to zero without a conversation about budget.

What LabLynx charges for, and does not#

LabLynx has sold LIMS under most of the models above over the years, and the customer portal was the product that made the argument for us: a lab whose customers can see their own work stops fielding the telephone calls, and that only happens when the customer’s access costs nothing.