How rankings work
cro.directory ranks providers on one principle: how well they match your program. Here is how that works across search, shortlists, and sponsored inventory.
Not every page ranks the same way
cro.directory uses more than one ordering rule, depending on what you are doing. That is deliberate — browsing fifty profiles is not the same problem as scoring five candidates against a specific brief.
- Search and browse — keyword relevance to profile text and structured filters (services, modalities, regions, certifications, claim status). Default sort is relevance; you can switch to name (A–Z).
- Brief-driven shortlists — fit score (below), used when you submit a structured RFQ or request a ranked match against services, modalities, and indications.
- Explore and taxonomy hubs — Programmatic explore pages (service × modality × region) list matching CROs alphabetically. They are discovery hubs, not a single ranked recommendation for “best in region.”
The fit score
When you submit a structured brief — through an RFQ or a programmatic match request — we score each CRO against the capabilities you specify. The result is a fit score from 0 to 100, with a component breakdown you can inspect.
Fit scoring is fit-first. It weights structured capability overlap — what you need against what a provider publicly claims — not popularity, traffic, or sponsorship tier.
If your brief omits a dimension (for example, you specify services but not indications), that dimension does not count against anyone. Only the criteria you actually supply participate in the score, and their weights are rebalanced so the total still reflects your stated priorities.
What goes into the score
Baseline weights for a fully specified brief on cro.directory. Percentages shift when you leave a field empty.
Listing fit
~33%Whether the organization is published on cro.directory (or listed on both directories when a cross-domain brief allows it). A provider that does not serve your site context scores zero on this dimension.
Service overlap
~33%How many of the services in your brief appear on the provider profile. If you ask for three services and the provider lists two, that is a two-thirds match on this axis.
Modality overlap
~19%Same logic for modalities — ADC, mAb, cell therapy, and the rest of the taxonomy. Certain high-demand modalities can receive a modest bonus when both your brief and the profile align on them.
Indication overlap
~9%Therapeutic-area alignment when your brief includes indications. Like modalities, selected indication matches can receive a small bonus when overlap is strong.
Tie-breakers
When two providers earn the same fit score, ordering is deterministic:
- Domain-verified operators — providers who have completed Tier 2 domain verification (DNS or HTML proof) rank ahead of email-only claimants. Verification breaks ties; it is not a standing boost above a higher-scoring unverified profile.
- Alphabetical slug — stable, reproducible ordering when verification status is also equal.
Cross-domain briefs
Some programs span research and manufacturing. When a brief explicitly allows cross-domain matching, eligible providers from cdmo.directory can appear in results with a clear cross-listing marker. Their fit score uses the same overlap model; site context is one of the scored dimensions.
Sponsored placements
Paid sponsorship inventory appears in labeled Sponsored regions on home, search, and selected landing pages. Sponsors must still meet minimum profile and safety requirements, but sponsored ordering is separate from organic fit score.
We never interleave sponsored cards into an organic fit-ranked list. If you are reading a fit-scored shortlist, every row earned its position on capability match — not ad spend.
Sponsorship and listing tiers →Transparency
Fit score responses include a per-criterion breakdown: which terms matched, raw overlap on each axis, and the weight applied. Profile facts on cro.directory carry provenance links where enrichment applies, so you can trace a capability claim back to a public source.
We publish this methodology openly because outsourcing decisions are too important for opaque ranking. If a provider looks like a weak match, the breakdown should tell you why — and what data was missing.
Put the model to work
Filter CROs by capability on search, or submit a structured RFQ and build a fit-scored outreach list.