Comparing spatial biology platforms is the step managers reach for first and should probably take last. Every system on the market works, and every vendor can demonstrate it working. What differs is fit, and fit cannot be assessed until you have written down what you actually need. This guide runs the process in the order that produces a defensible decision.
Key Takeaways
|
Start With Requirements, Not Platforms
The most common procurement failure in this category is starting from a comparison of available systems and working backward toward a justification. It feels efficient, and it produces a decision shaped by whichever vendor was most persuasive. Writing requirements first inverts that: it gives you a document that eliminates options on your terms rather than theirs, and it becomes the backbone of the eventual business case.
Six requirements do most of the filtering. State each one as a specific commitment rather than a preference, because a requirement you would abandon under pressure is not a requirement.
Requirement | How to State It | Why It Matters |
Sample type | Name the fixation and preservation types you will actually run, with proportions | Usually the hardest constraint and the fastest eliminator. Determined by your archive, not by preference |
Target scope | Number of targets you need, and whether the panel must be customizable | Separates whole-transcriptome approaches from targeted panels |
Spatial scale | The biological unit you need to resolve, stated in terms of the question | Buying more resolution than the question needs raises data volume and analysis cost together |
Tissue area per run | Physical dimensions of the sections you need to cover | Capture area and imageable area vary and directly affect cost per sample |
Throughput | Sections per week at steady state, from your demand evidence | Realized throughput, not nameplate. Determines whether one instrument suffices |
Modality | Whether you need RNA, protein, or both on the same section | Co-detection requirements narrow the field substantially |
Table 1. The six requirements that eliminate the most options. Complete this before any vendor conversation.
Two of these deserve emphasis. Sample type is where labs most often discover late that their retrospective cohort is incompatible with the platform they wanted, so confirm it against your actual blocks rather than against what you intend to collect in future. And throughput should come from measured demand rather than aspiration, which is why the spatial biology readiness assessment is worth completing first. If you have not yet settled whether to buy at all, the volume at which ownership overtakes outsourcing is worked through in in-house vs. outsourced spatial biology.
How Do You Build a Shortlist?
Spatial platforms divide into three functional families, and your requirements will usually point at one of them before you compare individual systems. Sequencing-based transcriptomics captures tissue-resolved material for downstream sequencing. Imaging-based transcriptomics detects transcripts in situ on the instrument. Multiplexed proteomics detects protein targets by cyclic imaging or by mass-based methods.
The systems most commonly evaluated across those families include Visium and Xenium from 10x Genomics, CosMx and GeoMx from Bruker, PhenoCycler from Akoya Biosciences, MERSCOPE from Vizgen, and Hyperion from Standard BioTools. That is a list of what exists, not a ranking, and deliberately carries no specifications: plex counts, resolution figures, and throughput numbers change with each release, so take them from current vendor documentation rather than from any article, including this one.
For the underlying science of how the families differ, the methods treatment on Technology Networks goes deeper than is useful here: Exploring the Latest Advances in Spatial Transcriptomics. Detailed operational comparisons within each family are covered in Imaging-Based vs. Sequencing-Based Spatial Transcriptomics: Which Fits Your Lab?, Spatial Transcriptomics Platforms Compared, and Spatial Proteomics Platforms Compared.
Shortlist to Three, Not Two Two options invite a binary argument and tend to produce a decision by preference. Three forces explicit trade-off reasoning and, in most institutions, satisfies procurement requirements for competitive quoting. It also gives you leverage: a vendor who knows they are one of three negotiates differently from one who knows they are one of two. |
What Should You Ask Vendors?
Send the same written question set to every shortlisted vendor and require written answers. Verbal answers in a meeting are not comparable and are not quotable later. The questions below are grouped by what they protect you from.
Category | Questions to Ask in Writing | What It Protects You From |
Sample compatibility | Which fixation and preservation types are validated, not merely compatible? What section thickness and quality thresholds apply? What is the documented failure rate on tissue like ours? | Discovering your archive is unusable after purchase |
Consumables | What is the per-section consumable cost at our volume? Are reagents proprietary? What minimum order quantities and shelf lives apply? | Recurring cost that was never modeled |
Data | How much data per run? What formats are exported, raw and processed? Is the raw data accessible without proprietary software? | Storage cost surprises and format lock-in |
Software | What is licensed, perpetual or subscription? How many seats? What happens to access if we stop paying? Are analysis tools included or extra? | Recurring software cost and stranded data |
Installation | What are the power, environmental, and vibration requirements? What site work is our responsibility? What is the realistic time from order to first data? | Facility work discovered after the purchase order |
Service | What is included in warranty and what does the contract cost after? What is the guaranteed response time in our region? Where is the nearest engineer? | Downtime with no recourse |
Training | How many staff are trained, for how long, and where? What does additional training cost? What happens when trained staff leave? | An instrument only one person can run |
Roadmap | Is a chemistry or platform generation change expected? Will current consumables remain supported? What is the end-of-life policy? | Buying immediately before a generational change |
Table 2. Vendor question set. Require written answers and keep them, since they become the specification the vendor is held to.
Three of these are routinely under-asked. On data, ask specifically whether raw output is accessible without proprietary software, and whether export conforms to any open standard such as the OME next-generation file format, because format lock-in constrains analysis choices for the life of the instrument. On software, establish what happens to your access and your existing data if a subscription lapses. And on roadmap, ask directly rather than guessing, since vendors will often say more than you expect under a confidentiality agreement, and buying weeks before a generational change is a genuinely avoidable outcome. Broader guidance on evaluating suppliers, negotiating, and factoring installation timelines is in this guide to balancing cost and readiness in lab equipment purchasing. Vendor-specific due diligence is covered further in Evaluating Spatial Biology Vendors: The Questions to Ask Before You Buy.
Lab Management Certificate
The Lab Management certificate is more than training—it’s a professional advantage.
Gain critical skills and IACET-approved CEUs that make a measurable difference.
Running a Demonstration That Tells You Something
A demonstration on the vendor’s reference tissue tells you the instrument works, which was never in doubt. Insist instead that the demonstration runs on your tissue, prepared by your people where possible. Vendors resist this because it introduces risk to their sales process, and that resistance is itself informative.
Specify the demonstration in advance and treat it as an experiment with a defined outcome.
- Supply your own sections, including at least one from your most difficult or oldest sample type rather than only your best material.
- Ask to observe the full workflow, including sample loading and the hands-on steps, not only the finished result. You are assessing what your staff will have to do daily.
- Require the raw output files, not a curated figure. Then have whoever will actually analyze the data open them before you decide.
- Record the real timings, from sections in to data out, including any steps the vendor performs off-site.
- Talk to a reference customer the vendor did not choose, ideally one running your sample type at your scale. Ask what they would do differently.
The third step is the one most often skipped and the most valuable. An instrument that produces data your analyst cannot work with has not passed evaluation, however good the images look in the demonstration. Live demonstrations and prompt evaluation once approval is in place are also emphasized in these three keys to acquiring new instruments.
Total Cost of Ownership
Purchase price is the smallest number in this decision and the only one most approvers see. Build the comparison on total cost of ownership across the same horizon for every shortlisted option, because the ranking frequently changes once recurring cost is included. A cheaper instrument with proprietary consumables and subscription analysis software can cost more over five years than a more expensive one with open formats and perpetual licensing.
Interested in life sciences?
Include all of the following for each option, over the same period.
- Capital, plus any required ancillary equipment and workstations.
- Facility modification, scoped and quoted rather than estimated.
- Consumables at your realistic annual volume, including a repeat rate.
- Sequencing, where the workflow requires it, priced separately as cores do.
- Data storage across your retention period, not just the project length.
- Software licensing, distinguishing perpetual from subscription.
- Service contract from the end of warranty onward.
- Staff time for operation and analysis, costed at your loaded rate.
The full cost category breakdown, with published core facility rates as reference points and an embedded estimator, is in the spatial biology cost guide. For framing the request around lifetime cost rather than purchase price when it reaches an approver, this analysis of total cost of ownership as a tool for securing lab equipment is directly applicable. Footprint, throughput, and site requirements in operational detail are covered in Instrument Footprint, Throughput, and Sample Requirements: Assessing Operational Fit.
How Should You Score the Decision?
Agree the criteria and their weights before you see the vendor responses. Weighting afterward is how a decision already made gets dressed as an evaluation, and experienced approvers can tell. Score each shortlisted option from 1 to 5 against each criterion, multiply by the weight, and total.
Criterion | Suggested Weight | Score 1 to 5 On |
Sample type compatibility | 20% | Validated support for the tissue you will actually run |
Fit to the scientific question | 15% | Target scope and spatial scale against your stated requirement |
Total cost of ownership | 15% | Five-year modeled cost at your volume, not purchase price |
Data and format openness | 10% | Raw data accessibility, export standards, absence of lock-in |
Analysis burden | 10% | Whether your analyst can work with the output as delivered |
Throughput and turnaround | 10% | Realized sections per week against your requirement |
Operational fit | 5% | Footprint, services, and whether installation is straightforward |
Service and support | 5% | Regional coverage, response times, engineer proximity |
Training and staffing | 5% | Depth of training and resilience to staff turnover |
Roadmap and longevity | 5% | Generational risk and consumable support commitments |
Table 3. Weighted evaluation scorecard. Weights shown are a defensible starting point. Adjust them to your situation, but agree them before scoring.
Adjust the weights honestly rather than diplomatically. A core facility serving many groups should weight throughput and service higher. A lab with one irreplaceable retrospective cohort should push sample compatibility above 20% and accept that it may decide the outcome alone. A lab with no dedicated analyst should raise analysis burden sharply, because an instrument whose output nobody can process will not be used regardless of how it scores elsewhere.
Keep the completed scorecard. It becomes evidence in the business case; it answers the question of why this option rather than another, and it gives you a documented baseline to measure against once the platform is running, which is the subject of Measuring ROI on Spatial Biology Investments. For the operational picture across the whole workflow, from procurement through data infrastructure and staffing, see the Manager's Guide to Evaluating, Implementing, and Scaling Spatial Technologies.
This article was produced under Lab Manager's AI Editorial Guidelines.
