Investing in NGS automation is the most common response to a throughput problem and the most common way to spend money without solving one. The instinct, when a sequencing lab cannot keep up with demand, is to look at the sequencer, because the sequencer is the expensive, visible centerpiece of the operation. But the sequencer is almost never the constraint. In most labs the queue forms upstream, at library preparation, and downstream, in analysis, while the instrument itself sits idle between runs waiting for enough samples to justify starting. Buying a faster sequencer into that situation raises capacity that was never the limit and leaves the actual bottleneck exactly where it was.
This overview is about moving a sequencing lab from pilot to production volume by fixing the constraints that actually limit it: the library-prep bottleneck, partially filled flow cells, untracked samples, and undisciplined scheduling. It starts with the discipline that makes every later investment pay off, which is diagnosing where the real bottleneck sits before spending anything to fix it.
Key Takeaways
|
Where the Real Bottleneck Sits
The most useful thing a lab manager can do before spending anything on throughput is to find out where the throughput is actually lost, and that requires measurement, not intuition. Time each stage of the workflow for a representative period, from sample receipt through library preparation, sequencing, and analysis to delivered result, and crucially include the queue time between stages, not just the hands-on time within them. The stage with the longest total elapsed time, counting the waiting, is the constraint. Everything else is secondary until that one is relieved.
Samples pile up at two points, before the sequencer and after it. The instrument between them is the stage least likely to be the constraint.
Flow (2026)
The reason this matters is that the constraint is frequently not the stage that feels worst. Manual library preparation feels like the hardest job in the lab because it is repetitive, exacting, and hands-on, so it draws the automation budget. But if finished libraries then wait three days for a scheduled run slot, the real constraint is scheduling, and automating library prep will simply produce libraries faster that then wait the same three days. The symptom and the cause are different stages, and only measurement tells them apart.
There is a second, subtler reason measurement beats intuition here: constraints move. Relieve the library-prep bottleneck and the constraint does not disappear, it relocates, often to analysis, which was the second-longest stage all along and becomes the limit the moment prep stops being it. A lab that automates without re-measuring afterward frequently finds delivered throughput barely changed, because it fixed one stage and the queue simply re-formed at the next. The discipline is not a one-time diagnosis but a habit: measure, relieve the constraint, then measure again to see where it went.
Observed Symptom | The Likely Real Constraint | The Action That Helps |
Libraries wait days for a run slot | Scheduling, or under-filled flow cells | Batch to flow-cell capacity; publish a run schedule |
Library prep cannot keep up with demand | Library preparation is the true constraint | Automate prep, if volume justifies it |
Runs complete but results are slow | Analysis and interpretation downstream | Add analysis capacity, not sequencing capacity |
Instrument sits idle between runs | Insufficient batched volume to fill a flow cell | Improve batching and sample scheduling |
Frequent reruns consume capacity | Quality or process failure upstream | Root-cause the failures before adding capacity |
Table 1. A bottleneck-diagnosis framework. Match the symptom you observe to the constraint actually causing it, then take the action that relieves that constraint rather than the one that feels most urgent.
Batching and Flow Cell Utilization
Flow cell utilization is the single least visible driver of cost per sample, because the cost of a partially filled flow cell appears in the reagent budget rather than in any line labeled waste. A run that fills a large flow cell efficiently costs far less per sample than the same work spread thinly across a flow cell that is mostly empty, since the flow cell and reagent kit cost roughly the same whether the run carries six samples or eighty. The economics of this, with real figures across flow cell tiers, are worked through in How Much Does NGS Cost? Budgeting Instruments, Reagents, and Sequencing.
Interested in lab design?
The discipline that follows is simple to state and requires steady nerve to hold: batch to the flow cell, not to the calendar. Accumulate samples until a flow cell is economically full before running, rather than running partial flow cells to clear a backlog quickly. This trades some turnaround for substantially lower cost per sample, and the tradeoff is real, because at low volume the wait to fill a flow cell can itself become the turnaround problem. That interaction between batching and turnaround, and the point at which it argues for outsourcing overflow instead, is examined in In-House vs. Outsourced Sequencing: Core Facility or Service Provider?.
Partial Runs Are the Quietest Waste in the LabA half-filled flow cell pays the full-flow-cell price for half the output, and nothing in the run’s result flags it, because the run succeeded. The waste is invisible precisely because the work looks like it worked. The only place it shows up is the cost per delivered sample, and only if someone is tracking that number. Making flow cell utilization a metric you watch, run by run, is one of the highest-return habits a production sequencing lab can adopt, and it costs nothing but attention. |
Automating Library Preparation
When measurement shows that library preparation genuinely is the constraint, automation becomes a strong answer, and this is the point at which the automation investment actually pays off rather than merely relocating the queue. Automated liquid handling for library prep delivers consistent, repeatable pipetting, reduced hands-on time, lower contamination risk, and, when integrated with laboratory informatics, a digital record of every sample’s handling. For a lab whose throughput is truly prep-limited and whose volume is high enough to keep the automation busy, these benefits compound quickly.
The qualifier matters as much as the claim. Automation pays off when prep is the constraint and volume justifies the capital and the setup effort; it does not pay off automatically, and a liquid handler bought for a lab whose real constraint is scheduling or analysis becomes an expensive way to make the wrong stage faster. The growing demand that makes automation worthwhile for many sequencing labs, and how automation supports it, is discussed in this overview of how automation can support increasing demand for next-generation sequencing. The detailed decision, when the volume and constraint genuinely justify automating prep, is developed in Automating NGS Library Prep: When Liquid Handling Pays Off.
Tracking Samples Through the Workflow
As volume rises, the informal sample tracking that worked for a pilot project quietly stops working, usually without anyone deciding it has. A sequencing sample passes through receipt, extraction, quantification, library preparation, pooling, sequencing, and analysis, accumulating identifiers and metadata at every step, and every sample needs an unbroken record linking its result back to its origin. At low volume a spreadsheet holds this together. At production volume it becomes a source of error and a bottleneck in its own right, as staff spend time reconciling records instead of processing samples.
The transition from a spreadsheet to a laboratory information management system is one most growing labs make later than they should, because the spreadsheet keeps almost working right up until a mix-up or an audit forces the issue. Planning that transition before it becomes urgent, and choosing a system that fits the genomics workflow rather than a generic one retrofitted to it, is the subject of LIMS and Informatics for Genomics Labs: What to Look For, and the specific discipline of maintaining an unbroken chain of custody through the workflow is covered in Sample Tracking and Chain of Custody in Sequencing Labs.
Scheduling and Turnaround Commitments
Scheduling is the least expensive throughput lever and one of the most powerful, because it costs nothing but discipline and directly determines both flow cell utilization and turnaround. Three practices do most of the work. Publish a run schedule and hold to it, so that submitting groups can plan backward from known slots, which improves batching without anyone having to negotiate each run. Set a rerun policy in advance, deciding who pays and under what circumstances before the first failed run rather than during the argument that follows it. And define turnaround commitments honestly, against the batched reality of how the lab actually runs, rather than promising a turnaround that only holds if a flow cell happens to be full when a sample arrives.
When runs do fail, and they will, how quickly and how well the lab diagnoses the cause determines how much capacity failures consume. A disciplined approach to root-cause analysis and a clear rerun policy keep failures from silently eating throughput, and both are developed in Managing Run Failures: Root Cause Analysis and Rerun Policy. The detailed mechanics of batching, multiplexing, and scheduling to keep flow cells full are worked through in Batching, Multiplexing, and Scheduling to Maximize Flow Cell Use.
None of these levers requires buying a faster sequencer, which is the point. Sizing the data each run actually needs, so that flow cells are filled deliberately rather than by habit, is supported by the Sequencing Coverage and Cost Calculator: How Much Data Do You Actually Need?. And the way throughput fits into the whole operational picture of running a sequencing program, from cost through staffing, is in Next-Generation Sequencing in the Lab: A Manager’s Guide to Building, Budgeting, and Scaling NGS Capacity. The consistent lesson across all of it is the same: find the real constraint, relieve that one, and measure whether it moved before spending on the next.
This article was produced under Lab Manager's AI Editorial Guidelines.
