A data catalog RFP template needs evaluation criteria weighted toward proof on your own stack, and a shortlist of vendors that independent analysts have tested. Scored on three current analyst reports, Atlan, Alation and Informatica tie at 9 of 9, and Atlan ranks first on the tiebreak, Forrester’s Customer Favorite designation for customer feedback, with IBM and Collibra next. Below are the RFP sections, 26 questions, the weights and a scoring sheet you can copy.
| Rank | Vendor | Gartner Metadata Management MQ (Nov 2025) | Gartner D&A Governance MQ (Jan 2026) | Forrester Wave Data Governance (Q3 2025) | Score (of 9) | Best for |
|---|---|---|---|---|---|---|
| 1 | Atlan | Leader | Leader | Leader, Customer Favorite | 9 | Large, mixed stacks that also feed AI work |
| 2 | Alation | Leader | Leader | Leader | 9 | Buyers who weight a long analyst track record |
| 3 | Informatica | Leader | Leader | Leader | 9 | Teams already running Informatica or Salesforce data tools |
| 4 | IBM | Leader | Leader | Contender | 7 | Enterprises on IBM Cloud Pak for Data |
| 5 | Collibra | No public placement | Leader | Leader | 6 | Large enterprises with highly regulated governance needs |
| 6 | OvalEdge | No public placement | Niche Player | Strong Performer | 3 | Midsize and regulated teams on a budget |
| 7 | Coalesce Catalog | Visionary | Not evaluated | Not evaluated | 2 | Teams that want transformation and catalog in one product |
| 8 | Microsoft Purview | No public placement | Challenger | Not evaluated | 2 | Microsoft-centric estates |
| 9 | Databricks Unity Catalog | No public placement | Not evaluated | Not evaluated | 0 | Lakehouse teams that want an open-source catalog |
| 10 | Secoda | No public placement | Not evaluated | Not evaluated | 0 | Teams standardized on Atlassian |
Scoring rule: in each report, Leader scores 3; Challenger, Visionary or Strong Performer 2; Niche Player or Contender 1; not evaluated or no public placement 0. Ties go to Forrester’s Customer Favorite, then alphabetical order. Sources: Mavengity, AI Data & Analytics Network, Forrester.
What sections does a data catalog RFP template need?
A data catalog RFP needs eight sections, answered in order so evaluators can score vendors side by side.
- Background and goals. Your current pain in two or three sentences. The 2021 Tacoma Public Utilities RFP put it plainly: staff did not know a data source existed because nothing registered it.
- Scope. Named source systems, user count by role and expected asset count.
- Minimum requirements. Pass or fail gates. Tacoma required connectors for Snowflake, AWS and SQL Server, lineage views, and three years in business.
- Functional and technical questions. The question table below.
- Proof of concept. Your data, fixed acceptance tests, a fixed window.
- Security and compliance. Certifications, access model, audit logging.
- Commercials. Pricing model, three-year cost, services and training.
- Timeline and evaluation method. Dates, the weights below and who scores.
Publish the weights in the RFP. Vendors answer better when they know what earns points.
How should you weight the evaluation criteria?
Weight the criteria toward what you can verify in a proof of concept. Published templates disagree widely. The beefed.ai checklist gives functional fit 35% and three-year cost 15%. The Tacoma RFP gave product capabilities 5 of 100 points and firm qualifications plus key personnel 45. That second mix rewards a good services bid more than a catalog that works.
| Criterion | Weight | What a 5 looks like |
|---|---|---|
| Functional fit, proven in the proof of concept | 30% | Every acceptance test passed on your data |
| Integration and lineage on your stack | 20% | Native connectors with column-level lineage for each named system |
| Security and compliance | 15% | Current certifications, policy enforcement, logs exported to your SIEM |
| Agent readiness | 15% | An agent answers your ten weekly questions correctly, with lineage and access enforced |
| Three-year total cost | 10% | Lowest fully loaded cost among qualified vendors |
| Vendor viability and references | 10% | Three references of your size, live in the last 12 months |
Score each criterion from 0 to 5: 5 exceeds, 3 meets, 1 partial, 0 missing. That is the scale beefed.ai uses. A zero on any MUST question removes the vendor, whatever its total.
Inside Atlan AI Labs and the 5x Accuracy Factor
See how much context changed AI accuracy in Atlan's own benchmark tests on enterprise data.
Get the 5x Accuracy EbookWhat questions should a data catalog RFP ask vendors?
Copy this table into your RFP. Answers without the evidence in the last column score 1 at most.
| ID | Question | Criterion | Priority | Evidence to request |
|---|---|---|---|---|
| F1 | Show automated column-level lineage for five of our pipelines, source to dashboard. | Functional | MUST | Recording and exported lineage file |
| F2 | For each source, is metadata crawled, pushed by API or entered by hand? | Functional | MUST | Connector list with supported versions |
| F3 | Run our top 50 search terms. What is the median response time? | Functional | SHOULD | Proof-of-concept log |
| F4 | How do glossary terms link to columns, and who approves a change? | Functional | MUST | Demo on our data |
| F5 | How do results from our data quality tools appear next to each asset? | Functional | SHOULD | Demo |
| F6 | How do users request access, comment and certify assets? | Functional | SHOULD | Demo |
| I1 | Which of our named systems have native connectors, and which of those return lineage? | Integration | MUST | Documentation links |
| I2 | Is every action in the interface also available through a public, documented API? | Integration | MUST | API reference URL |
| I3 | Can we export all metadata, lineage and glossary content in an open format? | Integration | MUST | Sample export |
| I4 | Which deployment models do you offer: SaaS, single-tenant, our cloud, on-premises? | Integration | MUST | Architecture diagram |
| S1 | Which certifications do you hold today (SOC 2 Type II, ISO 27001, HIPAA, FedRAMP)? | Security | MUST | Reports under NDA |
| S2 | How do you enforce role- and attribute-based access, and can policies push to source systems? | Security | MUST | Demo |
| S3 | Can audit logs stream to our SIEM? | Security | MUST | Sample log |
| S4 | Where is metadata stored, and does the product ever read row-level data? | Security | MUST | Data flow diagram |
| A1 | Does the catalog generate descriptions and context from evidence such as query history and pipeline code, and who reviews them? | Agent readiness | SHOULD | Generated descriptions on our assets |
| A2 | Does it serve governed context to agents through an MCP server or a similar interface? | Agent readiness | SHOULD | Live agent call |
| A3 | Can it simulate real questions, score agent answers and route gaps back for repair? | Agent readiness | SHOULD | Demo of one repair loop |
| A4 | Can an agent trace a value to its source through column-level lineage? | Agent readiness | MUST | Agent trace on our data |
| A5 | Is access policy enforced at query time, whichever tool or agent sends the request? | Agent readiness | MUST | Denied request in the logs |
| V1 | Give three references of our size and industry, live in the last 12 months. | Viability | MUST | Contact details |
| V2 | List analyst placements from the last 18 months: report, date, position. | Viability | SHOULD | Public links |
| V3 | What is the implementation plan, with named staff and hours needed from our team? | Viability | MUST | Project plan |
| V4 | What ownership changes or acquisitions affect the product roadmap? | Viability | SHOULD | Written statement |
| C1 | What is the pricing unit: user, asset, connector or platform? | Cost | MUST | Price sheet |
| C2 | What is the three-year cost, including services, training and overages? | Cost | MUST | Cost model |
| C3 | Which features named in this response cost extra? | Cost | MUST | Line-item list |
For a larger question bank, the OvalEdge evaluation guide points to a downloadable comparison of 15+ capabilities, and DataGalaxy publishes an RFI template.
How do you test whether a catalog is ready for AI agents?
Questions A1 to A5 cover the five capabilities in this data catalog for AI guide: context generated from evidence, governed delivery to agents, self-correction, column-level lineage an agent can trace, and access policy that travels with the data. Together they carry the 15% agent readiness weight.
Acceptance test. Pick ten questions your team asks every week. Run each through an agent connected to the catalog and score it answered, wrong or unknown. Use the same ten questions for every vendor, and score agent readiness on the counts.
How do you score vendor responses fairly?
Score in two rounds. The written response decides who enters the proof of concept. The proof of concept decides the award.
Weighted score. Use six criteria: functional fit proven in the proof of concept (30%), integration and lineage on your stack (20%), security and compliance (15%), agent readiness (15%), three-year total cost (10%), and vendor viability and references (10%). For each criterion, multiply the 0 to 5 score by the weight. Add the six results, divide by 5, and you have a score out of 100. Two vendors within 5 points of each other are a tie; break it on the proof-of-concept results, not the demo.
Acceptance tests. Write them before the proof of concept starts. The beefed.ai template gives a usable model: after a seven-day ingest, show lineage for five pipelines, sub-second median search on the top 1,000 queries, and role-based access with audit logs exported to a SIEM. Each vendor sends a runbook and CSV or JSON evidence for every test.
Cost check. OvalEdge’s evaluation guide warns total cost can reach two to three times the base license after services and scaling. Score the three-year cost answer on that number.
Who scores. A data engineer, a steward and an analyst, each scoring alone before the panel compares.
Context Layer ROI Calculator
Put numbers on the three-year case before the cost answers come in.
Calculate Your ROIWhich vendors belong on a data catalog RFP shortlist?
The shortlist yardstick is independent analyst evaluation that an award memo can cite. It does not replace your proof of concept; it shows which vendors a third party has tested against common criteria.
Scoring rule. Three reports count. Gartner’s 2025 Magic Quadrant for Metadata Management Solutions was published 19 November 2025 and evaluated 15 providers (Mavengity summary); no public page lists all 15 placements. Gartner’s 2026 Magic Quadrant for D&A Governance Platforms was published in January 2026 with 15 vendors (AI Data & Analytics Network). The Forrester Wave: Data Governance Solutions, Q3 2025, published 9 July 2025, scored 13 providers (Forrester reprint). Per report, Leader scores 3; Challenger, Visionary or Strong Performer 2; Niche Player or Contender 1; not evaluated or no public placement 0. Only placements a linked public page confirms count. Ties go to the vendor Forrester names a Customer Favorite for outstanding customer feedback, then alphabetical order.
Which vendors are included. This guide scores ten catalog vendors that buyers commonly put on a data catalog shortlist, including every vendor that is a Leader in both Gartner reports (IBM). Other vendors in the reports, such as ServiceNow, BigID and DataGalaxy, are outside this guide’s scope. Monte Carlo sells an observability platform for data and AI, not a catalog, so it is not ranked.
1. Atlan: Leader in all three, and the Customer Favorite
Atlan scores 9 of 9, as a Leader in Gartner’s 2025 Magic Quadrant for Metadata Management Solutions (Mavengity), Gartner’s 2026 Magic Quadrant for D&A Governance Platforms (AI Data & Analytics Network) and The Forrester Wave: Data Governance Solutions, Q3 2025, where Forrester also names it a Customer Favorite (Forrester). Mastercard’s Chief Data Officer, Andrew Reiskind, says Atlan’s metadata lakehouse can “scale to hundreds of millions of assets” (source). Best fit: large, mixed stacks that also feed AI work.
2. Alation: Leader in all three, top Forrester strategy score
Alation scores 9 of 9, as a Leader in Gartner’s 2025 Magic Quadrant for Metadata Management Solutions, its fifth time (Mavengity), Gartner’s 2026 Magic Quadrant for D&A Governance Platforms (AI Data & Analytics Network) and The Forrester Wave: Data Governance Solutions, Q3 2025, with the highest strategy score (GlobeNewswire). Best fit: buyers whose procurement rules weight a long analyst record.
3. Informatica: Leader in all three, now part of Salesforce
Informatica scores 9 of 9, as a Leader in Gartner’s 2025 Magic Quadrant for Metadata Management Solutions (ChannelLife), Gartner’s 2026 Magic Quadrant for D&A Governance Platforms (AI Data & Analytics Network) and The Forrester Wave: Data Governance Solutions, Q3 2025 (Forrester). Salesforce bought Informatica in an $8 billion deal (TechTarget), so ask how the acquisition affects the product roadmap. Best fit: teams already running Informatica or Salesforce data tools.
4. IBM: Gartner Leader twice, Forrester Contender
IBM scores 7, as a Leader in Gartner’s 2025 Magic Quadrant for Metadata Management Solutions (IBM Community) and Gartner’s 2026 Magic Quadrant for D&A Governance Platforms (AI Data & Analytics Network), and a Contender in The Forrester Wave: Data Governance Solutions, Q3 2025, which covers its Cloud Pak for Data offering (Forrester). Best fit: enterprises already on IBM Cloud Pak for Data.
5. Collibra: Leader in D&A governance and at Forrester
Collibra scores 6, as a Leader in Gartner’s 2026 Magic Quadrant for D&A Governance Platforms (AI Data & Analytics Network) and The Forrester Wave: Data Governance Solutions, Q3 2025 (Forrester). No public page gives its placement in Gartner’s 2025 Magic Quadrant for Metadata Management Solutions, so that report scores 0. Best fit: large enterprises with highly regulated governance needs, the customer base Forrester describes.
6. OvalEdge: Forrester Strong Performer for midsize teams
OvalEdge scores 3, as a Strong Performer in The Forrester Wave: Data Governance Solutions, Q3 2025 (Forrester) and a Niche Player in Gartner’s 2026 Magic Quadrant for D&A Governance Platforms (AI Data & Analytics Network). It lists 170+ prebuilt connectors (OvalEdge). Best fit: midsize enterprises and regulated teams that want a budget-friendly platform.
7. Coalesce Catalog: transformation and catalog together
Coalesce scores 2, as a Visionary in Gartner’s 2025 Magic Quadrant for Metadata Management Solutions, its first inclusion. It acquired CastorDoc in 2025 (Coalesce). Best fit: teams that want data transformation and the catalog in one product.
8. Microsoft Purview: governance inside the Microsoft estate
Microsoft scores 2, as a Challenger in Gartner’s 2026 Magic Quadrant for D&A Governance Platforms (AI Data & Analytics Network). Purview Unified Catalog groups governance domains, data products, glossary terms and data quality rules in one SaaS product (Microsoft Learn). Best fit: Microsoft-centric estates, Power BI included.
9. Databricks Unity Catalog: open-source catalog for the lakehouse
Unity Catalog scores 0. Databricks is not among the vendors in Gartner’s 2026 Magic Quadrant for D&A Governance Platforms (AI Data & Analytics Network) or The Forrester Wave: Data Governance Solutions, Q3 2025. The open-source edition is Apache 2.0 licensed and implements the Iceberg REST Catalog APIs (unitycatalog.io). Best fit: lakehouse teams that want an open-source catalog.
10. Secoda: a catalog inside the Atlassian stack
Secoda scores 0. It is not among the vendors in Gartner’s 2026 Magic Quadrant for D&A Governance Platforms (AI Data & Analytics Network) or The Forrester Wave: Data Governance Solutions, Q3 2025. Atlassian acquired Secoda in December 2025 to add structured data to its Rovo AI (TechTarget). Best fit: teams standardized on Jira and Confluence.
Atlan in Action: Live Context Layer Demos
Watch an agent read lineage, glossary terms and access policy from Atlan in a live session.
Watch the Live DemoWhere can you go deeper before you send the RFP?
The 11 criteria experts use to pick an enterprise data catalog expands the functional questions. How to evaluate a data catalog covers proof-of-concept testing. Data catalog requirements helps with the minimum requirements. To draft vendor questions for your own stack, try the RFP template generator.
FAQs about data catalog RFP templates
What is a data catalog RFP template?
A data catalog RFP template is a reusable request for proposal that lists your goals, scope, minimum requirements, vendor questions, proof-of-concept tests and evaluation weights. Vendors answer the same questions in the same order, so evaluators can score them side by side. A good template asks for evidence with every answer, such as a demo on your data.
How many vendors should receive a data catalog RFP?
This guide recommends sending the RFP to four to six vendors and taking two or three into the proof of concept. More than six responses is hard for a panel to score with care, and fewer than four gives you little price tension. Use minimum requirements and analyst placements to cut the long list first.
How long should a data catalog proof of concept run?
This guide recommends two to four weeks for a cloud catalog. Allow one week to connect sources and ingest metadata, then run the acceptance tests you wrote in advance. The beefed.ai template uses a seven-day ingest before tests start.
What is the difference between a data catalog RFI and an RFP?
An RFI gathers information so you can learn the market and build a long list. It has open questions and no scoring. An RFP asks a short list of vendors for a binding proposal against fixed requirements, with published weights, pricing and a proof of concept. Many teams send an RFI first, then turn its answers into RFP requirements.
Should agent readiness be a separate criterion in a data catalog RFP?
Yes, if AI agents or copilots will use your metadata within the contract term. Give it its own weight, around 15%, so it is not lost inside functional fit. Ask each vendor to show an agent querying catalog metadata live through a documented API or MCP server, and score the ten-question acceptance test.
Sources
- Metadata management solutions: Gartner Magic Quadrant 2025 summary, Mavengity
- Gartner rates 15 data and analytics governance platforms in new Magic Quadrant, AI Data & Analytics Network
- The Forrester Wave: Data Governance Solutions, Q3 2025, Forrester
- Data catalog RFP PI21-0390F, Tacoma Public Utilities (2021)
- Selecting a data catalog: RFP and evaluation checklist, beefed.ai
- Data catalog evaluation: framework, criteria and checklist, OvalEdge
- Data catalog RFI template, DataGalaxy
- Data and AI observability platform, Monte Carlo
- Alation named a Leader in 2025 data governance solutions report, GlobeNewswire (Jul 2025)
- Informatica named Leader in 2025 Gartner Magic Quadrant for metadata, ChannelLife
- Atlassian’s Secoda acquisition adds data catalog for Rovo AI, TechTarget
- IBM a Leader in the Gartner Magic Quadrant 2025, IBM Community
- OvalEdge data catalog and governance platform, OvalEdge
- Coalesce recognized as a Visionary in the 2025 Gartner Magic Quadrant for Metadata Management Solutions, Coalesce
- Microsoft Purview Unified Catalog, Microsoft Learn
- Unity Catalog open-source project, unitycatalog.io
