How LIMS requests get approved

1. Four reasons get a LIMS approved

Of the 20 approved requests, 17 named at least one of four reasons, and 13 named two or more.

The same headings run through corporate capital plans. A large refiner reports spending "for sustaining the business, including costs for turnarounds, catalysts and regulatory compliance." One utility ranks its capital projects on "regulatory requirements, system reliability, safe and secure operation of the system, cost savings."

2. The strongest requests name a deadline and what missing it would mean

Two wastewater labs faced the same accreditation deadline. "By October 2023, the laboratory will have to be fully compliant with TNI in order to remain in compliance and maintain its accreditation," one wrote. The other spelled out what missing it would mean: "the City could incur significant fines and penalties."

Two others faced the end of support. A water quality lab's support was running out, and without it "the LIMS would need to be shut down absent security patches." A third wastewater lab ran its system on servers using "a programming language no longer licensed to the City," and wanted the replacement in place "prior to any potential disruption to regulatory compliance."

3. Approved requests turn everyday complaints into risks

The labs we interviewed for our replacement study describe their systems the way the bench lives with them. None of the 20 approved requests opened with how hard the system is to use or how staff feel about it. They described the same problems in terms an approver can weigh:

When lab data goes wrong, the bill can be large: our study of what lab data errors cost reviewed documented costs from $7,000 to $1.2 billion, and the size depended on how far the product traveled before someone acted on the result.

Reviewers notice when a request leaves the problems in the bench's terms:

4. Requests fail when the plan isn't settled

Five requests in our records never reached approval. Three were canceled before an award with no reason given, or with only a legal citation. One drew no acceptable bids. One, a budget request, was left out of the budget.

Some made the same case as the approved requests: results typed in by hand, a system past support. They lacked a settled plan.

Two of these labs went back out to bid and got approval on the second try, about 14 and 18 months after their first bid deadline. For the nine approved requests that went out for proposals, the time from the request for proposals to the award averaged about a year, and ranged from about four months to more than three years.

5. Finance asks three questions, and payback is the usual yardstick

A reliability leader who sells programs to finance put the questions plainly: "How much does the current state cost us?" "How much will the proposed change cost?" "What's the expected return, and when?" His advice: "Use conservative estimates. If your projections look too good, the CFO will assume you're inflating them."

At companies under $1 billion in revenue, payback is used far more often than net present value or internal rate of return.

Reviewers mark down the same gaps. "Only the 'do nothing' option was mentioned," the technology review board wrote of one lab's request. "Maintenance at 10% could easily be over-optimistic."

The approved requests answered finance's questions up front. They separated the one-time cost of getting started from the yearly cost of running the system, showed the full cost over five years (our cost study breaks that total down), and set aside contingency as its own line: 10 percent in one request, 20 percent in another.

6. Price matters less than labs expect

In the scoring behind four approved purchases, price was worth 10 of 110 points in one and 20 to 30 percent of the score in the other three. A lab that is still choosing its system gives price 10 points of 100. Other things weighed as much or more. One request gave an oral interview and system demonstration 30 points and price 20. The lab still choosing gives efficiency and ease of use 20 points and price 10. A third gave security, and training and support, each as many points as price.

None of the 20 approved requests gave cost as the reason to replace. What a LIMS really costs over five years, beyond the license, is the subject of our cost study.

7. Approvers fear a failed project more than a high price

In a study of 1,471 IT projects, one in six ran over budget, by an average of 200 percent. In our replacement study, labs that kept an old system said leaving looked too hard.

A plan for handling the change lowers that risk. A pilot is one way; as the reliability leader put it, "A pilot de-risks the investment." Others are starting at one site and rolling out from there, keeping the old system running until the new one matches, and setting milestones. Labs write this into their own requirements: one asked for a parallel run of at least five business days, another for "continuity of operations between existing systems and new systems."

8. Labs with no LIMS make a different case

Several of the approved requests were for a lab's first LIMS, replacing paper, spreadsheets or a home-built database. These requests make the case on traceability, paper records and audit risk, as well as efficiency.

Spreadsheets carry risks of their own, which our study of lab data errors covers.

What this means if you're making the case

Lead with the risk, not the annoyance. Approvers fund risk, so translate each complaint into what it puts at risk: a missed deadline, a result that can't be defended, hours the lab doesn't have.

Tie the request to a date with a consequence. An accreditation deadline, an end-of-support date, a security patch that won't come, and what happens when it passes.

Settle the plan, and have the answers, before the request goes in. A budget figure, requirements that won't change midstream, a schedule people believe, a place in the annual plan, IT's sign-off and a sponsor.

Answer finance's three questions, conservatively. What the current state costs, what the change costs, and when it pays back, with the options side by side, including doing nothing.

Show how the risk of the change is handled. A pilot, one site first, the old system running until the new one matches.

How we did the study

We reviewed the written justifications behind 20 LIMS requests that were approved and 5 that were canceled or never funded between 2019 and 2026, from the same records as our replacement study, along with the scoring sheets, award decisions and budget records they went through. We added published approval policies, capital plans and regulatory filings from utilities, refiners and pipeline companies, interviews with labs in refining, pipelines, midstream and electric power, and research on how companies decide to buy software.

About Sabine Industrial

Sabine Industrial builds SI LIMS for fuel, energy, utility and chemical quality labs. Our team has spent fifteen years working with data extracts from every major LIMS. This is the second study in a series on how labs buy, run and replace a LIMS.

Questions about the data: info@sabineriverpartners.com