Lubinpla

Why customer sample turnaround time stretches to weeks

Summary

Customer sample turnaround in technical service usually stretches to weeks because samples wait between steps: at receipt, in transit, for interpretation or for approval. The lab analysis is rarely the slow part when the test itself takes hours or days; laboratory studies report that non-analytical delays may account for up to 96% of total turnaround (Hawkins, 2007).

  • In laboratory turnaround studies, non-analytical delays may be responsible for up to 96% of total turnaround time (Hawkins, Clin Biochem Rev, 2007).

  • Turnaround time distributions are skewed to the right, so the median plus a tail measure such as the 90th percentile is preferred to the mean alone (Hawkins, 2007).

  • In a 1998 survey, 41% of laboratories started the emergency turnaround clock at laboratory receipt, while only 9% of physicians did (Hawkins, 2007, citing CAP Q-Probes).

  • Under US DOT rules, a sample whose hazard class must be determined by testing may ship under a tentative classification, with the word Sample in the shipping description and in combination packaging of no more than 2.5 kg net per package; self-reactive, organic peroxide and explosive samples follow separate rules (US DOT, 49 CFR 172.101, 2026).

This diagnosis fits chemical, lubricant and coating suppliers whose technical service tests customer samples in house or at a contract lab. It does not fit long exposure tests, such as aging or corrosion runs, where the test duration itself sets the turnaround, nor regulated release testing with mandated timelines.

What does a slow customer sample test look like in technical service?

From the outside, a slow test is a short lab result inside a long case. The instrument work takes a day, yet the customer hears back after three or four weeks (illustrative). Part of the gap is how each side counts: the customer starts the clock on the day they asked, while the lab often starts it on the day the sample was logged.

Clinical laboratories have the same split. In a 1998 survey, 41% of laboratories defined emergency turnaround as the time from receipt in the laboratory to report, while only 9% of physicians started the clock at laboratory receipt (Hawkins, 2007, citing CAP Q-Probes). A technical service team that reports lab time alone will look fast to itself and slow to its customers.

The case history of a slow sample usually shows one or more of these:

  • The sample sat in receiving because the request named no test method or no specification.

  • The sample crossed a border and waited for shipping papers or customs.

  • Results were entered, and then the report waited for the one engineer who knows the application.

  • The draft report went through two or three rounds of approval comments.

  • A retest was needed because the first sample was contaminated or mislabeled.

None of this is analysis time, and none of it shows up in an average lab turnaround figure.

What does each stage of a customer sample test add to turnaround time?

Each stage adds its own working time plus a queue, and the queues are the likely source of the weeks. In clinical laboratory work, studies report non-analytical delays of up to 96% of total turnaround, though one CAP study of urgent hospital samples still found 28.2% of delays in the analytical phase (Hawkins, 2007).

Splitting the case into five stages shows where the queue sits, but only if everyone counts from the same events. Laboratory literature lists start points as varied as sample receipt, registration or analytical sampling time (Hawkins, 2007), so write down the start and end event of each stage before you measure.

Receipt and request review

The sample is logged and checked against the request: test, method, specification and safety data sheet (SDS). A missing item stops the clock until the customer answers.

Sample transit

Transit runs from customer dispatch to lab receipt. It grows when a sample needs dangerous goods paperwork or crosses a border, and the SDS transport section and each jurisdiction's shipping rules apply.

Analysis

Analysis covers the queue before the instrument as well as the test itself. It runs long when tests are batched or one instrument serves many requests.

Interpretation

Interpretation turns the numbers into an answer to the customer's problem. It waits for whoever knows the application.

Approval and release

Approval waits for sign-off, and each review round starts that wait again.

How can you tell which stage is the bottleneck from case records?

Timestamp when every sample enters and leaves each stage over a closed period. The bottleneck is the stage with the longest median wait and the largest share of slow cases, which is not always the stage with the most work.

The diagnostic table below maps what each stage looks like in the records to the bottleneck it points to.

Stage

What you see in the records

Points to

Receipt and request review

Days between receipt log and test start; notes such as "awaiting SDS" or "which spec?"

Incomplete request form; no agreed test method or decision rule

Sample transit

Gap between customer dispatch date and lab receipt; held at courier or customs

Shipping papers, dangerous goods classification, wrong address or lab

Analysis

Long queue before test start for some tests only

Instrument or analyst capacity; batching rules

Interpretation

Result entered, report not drafted for days

One person holds the application knowledge

Approval and release

Several report versions; long waits after each

Unclear approval criteria; approver capacity

Retest

Second result entry for the same sample or case

Rework from sample, method or labeling problems

Read the table by stage, not by person. A stage is the bottleneck when its median wait is long and it holds a large share of the slowest cases. A stage with one long outlier usually points to a single event, such as a lost courier parcel.

Watch the interpretation row most closely. When results sit entered but unreported, more lab capacity will not shorten the case. The same stage-and-owner method, applied to all support work and not only samples, is laid out in how to find where technical support cases get stuck.

How do you separate rework from first-pass results when measuring sample turnaround?

Tag every attempt in the record and report two numbers: the turnaround of cases closed on the first attempt, and the share and added time of cases that needed a retest, resample or report revision. Mixing the two hides both problems.

A first-pass case is one where the first test on the first sample produced the reported result. A rework case had at least one extra attempt: a retest of the same sample, a new sample from the customer, or a revised report after the customer disputed the method (illustrative definitions). Count rework per case, not per test, so one bad sample does not count three times.

Report the first-pass median turnaround and a tail figure such as the 90th percentile, because turnaround distributions are skewed to the right and the median and tail size are the preferred measures (Hawkins, 2007). Then list the rework causes.

For the tail, clinical laboratory practice offers a model. Steindel and Novis recommended setting an outlier criterion where about 10% of samples exceed it and establishing the cause for every sample beyond it (Hawkins, 2007). In technical service, the slowest tenth of sample cases is usually small enough to read one by one.

When rework traces back to the request, such as the wrong method or no specification, it belongs to the receipt stage, not to the lab.

How do you shorten sample turnaround once the bottleneck is found?

Cut the queue at that one stage. Speeding up any other stage leaves the total almost unchanged. Little's law shows why: in a stationary system, the long-term average number of items in the system equals the average arrival rate times the average time each item spends there (Little, 1961). If 30 samples wait for interpretation and 10 arrive each week, the average sample spends about three weeks in that stage (illustrative).

Match the fix to the bottleneck:

  1. Receipt: use a request form that requires the test method, specification, decision rule and SDS before the sample ships. ISO/IEC 17025:2017 clause 7.1.1 requires a review procedure that ensures "the requirements are adequately defined, documented and understood" (ISO/IEC 17025:2017, via PJLA, 2022).

  2. Transit: ship on the day of sampling where possible, as one commercial lab advises (POLARIS Laboratories, 2024), and prepare dangerous goods papers from the SDS before pickup.

  3. Analysis: set batching rules by test, and publish them to sales so that promises match them.

  4. Interpretation: train a second person for each application, and cap the open cases per engineer.

  5. Approval: write down approval criteria, and allow one review round for routine reports.

After one period, recheck the same stage with the same definitions.

When should a customer sample test go to an outside lab?

Send a sample out when the in-house team lacks the method, the capacity in time, or the accreditation the customer needs, and only when the added transit and contract review still finish sooner than the internal queue. An outside lab moves the bottleneck; it does not remove the receipt and interpretation stages.

ISO/IEC 17025:2017, Edition 3, confirmed in 2023, is the standard accreditation bodies use to assess testing laboratories (ISO, 2023). Where a lab uses external providers, clause 7.1.1 c) requires that the laboratory "advises the customer of the specific laboratory activities to be performed by the external provider and gains the customer's approval" (ISO/IEC 17025:2017, via PJLA, 2022). When the customer asks for pass or fail against a specification, clause 7.1.3 requires the decision rule to be clearly defined (ISO/IEC 17025:2017, via PJLA, 2022).

Before sending, check that the outside lab's stated turnaround covers your case. One oil analysis lab states that it strives to return results in approximately two days (POLARIS Laboratories, 2024), but that figure comes in a sample lifecycle written for oil, diesel and coolant samples, so confirm it for your test. Hazardous samples ship under the SDS transport section and local rules; in the US, 49 CFR 172.101(c)(11) allows a tentative classification for samples whose hazard class must be determined by testing (eCFR, 2026).

Lubinpla's technical support performance analysis reads confirmed results at each stage as performance and attempt history as rework, then names the bottleneck and the improvement priorities.

Frequently asked questions

Should sample turnaround be counted in calendar days or working days?

Use calendar days for the customer view, because the customer waits through weekends and holidays. Working days help when you compare staff workload. State which one each report uses, and never mix them in the same table.

Is the mean or the median better for sample turnaround?

Use the median with a tail figure such as the 90th percentile, since a few very slow cases pull the mean up. For tests with long turnaround, a study cited by Hawkins found the mean the most reproducible of four measures (Hawkins, 2007), so the mean still helps when tracking change over time.

What if a customer sample arrives without an SDS?

Log the sample as received but on hold, and record the wait as owned by the customer. Do not open or test it until the SDS arrives, because safe handling depends on it.

Related posts