A vendor demo is only useful if you arrive knowing what to measure. Without a pre-specified scorecard, the most compelling demo wins, and the most compelling demo is not necessarily from the best scanner for your laboratory. This buyer’s guide gives you the scorecard.

Most lab managers book a digital pathology demo to see the technology in action. That is the right instinct, but it is only half the purpose a demo should serve. The other half, the part that separates a structured evaluation from a purchasing experience shaped by the vendor’s presentation skills, is the set of questions you bring with you.

A vendor demo is a controlled environment. The scanner runs on optimal slides, chosen by the vendor. The demonstration workflow shows the platform at its best. That is not deceptive. It is a reasonable way for a vendor to present their product. But it means that if you do not ask the questions that reveal how the platform performs outside those optimal conditions, in your environment, at your volume, with your tissue types and your staffing levels, you will leave the demo with an impression rather than an evaluation.

This digital pathology buyer’s guide is structured around seven sections of evaluation questions, organized from the operational criteria that determine daily laboratory performance to the clinical and regulatory evidence criteria that determine whether a platform can be deployed for primary diagnosis. Each section includes a prespecified scorecard item and the diagnostic question that reveals whether a platform meets it. Bring this guide to every demo. The vendors who answer it well are the vendors worth evaluating further.

Before You Walk In: The Pre-Demo Checklist

A structured demo evaluation begins before you arrive. Three preparation steps ensure that the session produces evidence rather than impressions:

irst: bring your own slides. Specifically, bring slides from the diagnostic categories your laboratory finds most challenging, such as fine nuclear morphology cases, borderline dysplasia, high mitotic count assessments, tissue types with thin sections or significant background staining. These are the categories where image quality differences between platforms become visible in practice. A vendor will scan their own slides expertly. What you need to know is how the scanner performs on yours.

Second: bring your volume profile. Your current daily average, your documented peak day, your annual slide count, your projection for years 3 and 5 assuming 5–10% annual volume growth (Munari et al., 2024). You need the vendor to respond to your actual demand, not a generic case study.

Third: bring this guide. Not to read from it, but to use the questions as a structured scorecard. Record the vendor’s answer to each question immediately. You will be evaluating multiple platforms, and the differences between vendor answers become clear only when they are recorded consistently.

This section determines whether the scanner can physically handle your volume profile in a live clinical environment, not just in theory.

The loading architecture question is the most important in this section and the one most vendors will not volunteer. Published operational data shows that scanner throughput in routine clinical practice consistently falls below rated specifications, with loading overhead as a primary cause (Schwen et al., 2023). A scanner that requires staff-attended batch transitions three times per shift at your volume profile is consuming staff time and scanner capacity that a continuous-loading architecture eliminates.

This section determines how much of your scan team’s time will be consumed by quality exception management.

Published benchmarks identify 1 FTE per 3–4 WSI scanners as the staffing requirement for image QC in a well-automated digital pathology operation (Schwen et al., 2023; Ardon et al., 2023). The fraction of that FTE consumed by rescan management depends on automation tier. At 5% rescan rate across 500 slides per day, a Tier 1 or Tier 2 system generates approximately 25 manual interventions per day, or 6,250 per year. Tier 3 autonomous QC eliminates focus-failure interventions entirely.

This section tests whether the vendor’s quality claims are supported by evidence from sources other than the vendor.

Independent quality evidence is the criterion that distinguishes a platform backed by regulatory review and third-party benchmarking from one supported primarily by manufacturer claims. As of March 2025, six WSI platforms hold FDA 510(k) clearance for primary diagnosis. The first peer-reviewed comparative analysis of all six platforms (Kuhlman et al., 2025) shows a more than three-fold range in major discordance rates. The EMPAIA ISB 2025 provided independent multi-laboratory quality evaluation by a panel with no commercial affiliation. Both evidence sources are publicly available. Independent benchmark data can provide an additional point of comparison when available.

This section uses your own tissue to evaluate what the vendor cannot control.

The CAP 2022 guideline requires that local validation studies use cases representative of the intended clinical application (Evans et al., 2022). This means your challenging cases, not curated easy cases, are the ones that will determine whether your local validation succeeds. A platform that performs well on vendor-selected tissue but struggles with your actual diagnostic workload will produce a difficult validation study and ongoing diagnostic confidence issues in clinical deployment.

This section determines whether the scanner integrates with your existing infrastructure or creates a parallel workflow.

IT integration is the hidden implementation risk in digital pathology deployments. Published operational data from high-volume implementations identifies LIS integration gaps and undersized network infrastructure as the most commonly reported post-go-live performance problems (Schwen et al., 2023). A demo conducted on a vendor’s own network and LIS does not test integration with your specific IT environment. Reference sites can provide valuable insight into real-world integration performance.

This section determines the operational resilience of the platform in a live clinical environment.

Published operational requirements for clinical digital pathology specify scanner uptime targets of 99.9% for a platform treating scanning as a critical clinical infrastructure function (Schwen et al., 2023). At 500 slides per day, even 0.1% unplanned downtime represents approximately 12 hours of scan capacity lost per year — a manageable figure if service response is fast and a serious operational problem if it is not. The service model and MTTR commitment are as important as the scanner specification for a high volume clinical deployment.

Scoring Your Demo: The Evaluation Matrix

After completing a demo using this guide, score each vendor against the seven checklist sections on a simple three-point scale: strong (specific, independently verifiable answer with documented evidence), adequate (plausible answer without independent verification), or weak (unable to answer or deflected to ‘it depends’). A platform that scores strong on all seven sections has demonstrated that its performance claims are independently verifiable across every operational dimension that determines daily lab outcomes.

The sections are not equally weighted for every laboratory. For a high-volume lab primarily focused on throughput and staffing efficiency, Sections 1 and 2 carry the most operational weight. For a lab preparing for a CAP inspection cycle, Section 3 and the FDA clearance questions carry the most regulatory weight. Apply the guide to your institution’s specific priorities and score accordingly.

Conclusion: The Demo Is a Test, Not a Presentation

A well-run demo evaluation using this guide produces a structured evidence base for platform comparison rather than a collection of vendor impressions. It produces specific, recorded answers to specific, operational questions and answers that can be verified against independent evidence, compared across platforms, and defended in a capital request to a finance committee or a platform selection decision to clinical colleagues.

The vendors who deliver the strongest answers to these questions are the vendors that can support their claims with documented evidence across clinical, operational, and technical criteria and not for the controlled environment of a product demonstration. That is the distinction this buyer’s guide is designed to surface. Book the demos. Bring this guide. The answers will tell you what the specification sheet cannot.

Key Takeaways: Running a Structured Demo Evaluation

Learn More

References

1. Schwen, L. O., Kiehl, T. R., Carvalho, R., Zerbe, N., & Homeyer, A. (2023). Digitization of pathology labs: A review of lessons learned. Laboratory Investigation, 103(11), 100244. Read more.
2. Evans, A. J., Brown, R. W., Bui, M. M., Chlipala, E. A., Lacchetti, C., Milner, D. A., Pantanowitz, L., Parwani, A. V., Reid, K., Riben, M. W., Reuter, V. E., Stephens, L., Stewart, L., & Thomas, N. E. (2022). Validating whole slide imaging systems for diagnostic purposes in pathology: Guideline update from the College of American Pathologists. Archives of Pathology & Laboratory Medicine, 146(4), 440–450. Read more. 
3. Kuhlman, B., & Bedi, R. (2025). Comparative analysis of FDA-cleared whole slide imaging systems for primary diagnosis. International Journal of Pathology and Clinical Research, 11(1), 165. Read more.
4. EMPAIA (Ecosystem for Pathology Diagnostics with AI Assistance). (2025). 2025 International Scanner Benchmark: Award results. 37th European Congress of Pathology, Vienna. Read more.
5. US Food and Drug Administration. (2025). 510(k) premarket notification K241717: Epredia E1000 Dx Digital Pathology Solution. FDA CDRH. Read more.
6. Ardon, O., Klein, E., Manzo, A., Corsale, L., England, C., Mazzella, A., Geneslaw, L., Philip, J., Ntiamoah, P., Wright, J., Sirintrapun, S. J., Lin, O., Elenitoba-Johnson, K., Reuter, V. E., Hameed, M. R., & Hanna, M. G. (2023). Digital pathology operations at a tertiary cancer center: Infrastructure requirements and operational cost. Journal of Pathology Informatics, 14, 100318. Read more.
7. McGenity, C., Clarke, E., Jennings, C., Matthews, G., Cartlidge, C., Freduah-Agyemang, H., Randell, R., Treanor, D., & Bankhead, P. (2024). AI in pathology: A systematic review of clinical performance across cancer types. Lancet Digital Health, 6(8), e545–e555. Read more.
8. Munari, E., Zamboni, G., Cecchini, M. J., Zamo, A., Sommaggio, M., & Brunelli, M. (2024). Pathology in motion: Automation from specimen to report. Annals of Diagnostic Pathology. Read more.
9. Hanna, M. G., Reuter, V. E., Samboy, J., England, C., Corsale, L., Fine, S. W., Klimstra, D. S., & Travis, W. D. (2020). Implementation of digital pathology offers clinical and operational increase in efficiency and cost savings. Archives of Pathology & Laboratory Medicine, 143(12), 1545–1555. Read more.