A BETTER WAY TO SHARE THE WORK
Build a vendor comparison that helps people choose
A wall of checkmarks can look decisive while hiding the most important questions. Compare vendors against your needs, and make missing evidence impossible to mistake for a missing feature.
New to creating with an agent? The request will guide it through the task. Bring your notes; you don’t need to write HTML.
How do I create a trustworthy vendor comparison with an AI agent?
Start with your requirements and approved source material. Ask the agent to compare each vendor on the same criteria, cite evidence with dates, and label unknowns explicitly. Share a reviewed HTML comparison that explains the tradeoffs and the questions to resolve before choosing.
01 / SEE THE DIFFERENCE
Unknown is not the same as no
An illustrative team compares fictional Vendor A and Vendor B for an internal service. Their mandatory requirement is a specific access-management capability.
A comparison gives Vendor A a green checkmark and Vendor B a red cross because the evaluator found documentation for A but not B.
- Requirement
- Confirm the required access-management capability on the plan the team would purchase.
- Vendor A
- Documented on the enterprise plan in the supplied source. Availability on the proposed plan still needs confirmation.
- Vendor B
- Unknown from the supplied evidence. No documentation found is not evidence that the capability is unavailable.
- Decision status
- Neither option has yet satisfied the requirement for the actual purchase scenario.
- Next step
- Request written plan-specific answers from both vendors and update the comparison with the answers and dates.
02 / MAKE IT YOURSELF
Start with the evidence. Shape the page around the reader.
Separate requirements from preferences
Write down what would rule out an option before comparing appealing features. A mandatory access control is not interchangeable with a nicer interface. Ask the agent to preserve this distinction and identify requirements that are too vague to test. Avoid weighting criteria simply to manufacture a single numerical winner.
Make every comparison traceable
Give each cell a status and evidence: confirmed, conditional, not supported by the cited source, or unknown. Include the relevant plan, date, and limitations. 'We did not find documentation' should remain unknown unless stronger evidence establishes absence. Keep observed trial behavior separate from a claim on a marketing page.
Compare the same purchase scenario
Use the same number of users, billing period, required capabilities, and assumptions for each option. Mark quotes and published list prices distinctly. Note excluded implementation or usage costs rather than silently treating them as zero. If information is stale or missing, make verification the next step instead of offering a confident total.
Show the decision and the remaining work
Create static HTML with a short fit summary, a readable comparison, and questions to resolve. Use your saved style if available and meaningful text labels alongside color. Review claims and current terms before sharing. Frame any recommendation around the stated use case, not as a universal best-vendor ranking.
Adapt this to your work
Keep the same approach. Change the inputs and review checks for the job you need.
A seller-authored product comparison
Disclose who wrote the comparison and any commercial relationship. Start with the buyer's requirements, apply the same criteria to all options, and preserve limitations of your own product alongside competitors' strengths.
Fictional source notes: Fictional seller A writes the comparison. The buyer requires export on the proposed plan. A supplied document confirms A's export only on its enterprise plan; B's export is unverified. No trial or plan-specific confirmation is supplied for either proposed purchase.
A clearer result: Disclose seller A's authorship. Mark A conditional on plan confirmation and B unknown. Neither is a confirmed fit yet. Ask both vendors the same plan-specific question instead of displaying a green check for the author and a red cross for the competitor.
The matching instructions are included in the creation request below; your agent should use them only when relevant.
03 / TRY THIS PROMPT
Copy this into your agent. Add your source material.
Use the conversation where you already did the work, or start one with your notes. Replace the placeholders and supply any missing context; only attach information you are allowed to share with your agent.
Get the ready-to-copy creation request for guided questions, this task, and a draft review before anything is published.
Create a vendor comparison for [use case] using my requirements and supplied evidence. Separate mandatory requirements from preferences. Compare every option on the same criteria and purchase assumptions. For each material claim, retain its source, date, applicable plan, and limitations. Distinguish documented capability, observed trial behavior, vendor claims, and unknowns. Never convert missing evidence into a 'no' or invent prices, features, scores, or weights. Show the main tradeoffs and unresolved questions before making a use-case-specific recommendation. Flag information that needs current verification. Build readable static HTML using my saved artifact style if available, with text labels as well as color and no scripts or forms. Present the page for evidence, accuracy, confidentiality, and recipient-access review before publishing. For a seller-authored product comparison, disclose the author and supplied commercial relationships. Apply buyer-approved criteria consistently; include limitations of the author's offering and strengths of alternatives. Never turn unverified competitor information into a negative capability claim.
04 / BEFORE YOU SEND
A polished page still needs your judgment.
- Does each material capability or price claim have a dated source and the correct plan or purchase assumptions?
- Are missing evidence, confirmed limitations, conditional availability, and trial observations clearly distinguished?
- Are mandatory requirements visible, and are all options evaluated against the same criteria?
- Have changing terms and unresolved questions been checked before relying on the comparison for a purchase?
When HTML fits
Use an HTML comparison when a group needs a shared view of requirements, evidence, and tradeoffs. Short summaries plus a structured matrix can make an asynchronous evaluation easier to discuss.
When another format fits better
Keep formal procurement approvals, contract negotiation, and security review in their established processes. A static comparison is a dated decision aid, not live pricing, a product warranty, or an automated purchasing recommendation.
FROM DRAFT TO A LINK
Make it clear. Make it yours. Then share it.
Keep creating and revising in the agent you already use. With a compatible connection, share/artifacts publishes the reviewed HTML so you do not have to rebuild it in another editor. Ask your agent to use your saved artifact style for headings, color, and tone; review the result before sending it.
Choose who should be able to open the page and verify the returned link. For corrections to the same deliverable, update the existing artifact without changing its link. Publish a separate artifact when the old edition needs to remain available.
These are static pages, not live apps: no scripts, forms, or automatic data refresh. Source collection and scheduling depend on your agent and its available tools, not on share/artifacts.
Make this with my agentQuestions before you start
Should I use a score to choose the winning vendor?
Only when the criteria, scale, and weights represent the decision-makers' priorities. Keep mandatory requirements separate from a total score. A high score should not conceal a disqualifying gap, and unknown information should not receive a made-up rating.
How do I handle conflicting vendor information?
Show the conflict with the dates, sources, and relevant plans. Ask for clarification or test the requirement where appropriate. Do not silently choose the source that favors an option, and do not treat a general marketing claim as a plan-specific commitment.
Can I reuse the comparison for another team?
Reuse the structure, but review the requirements, assumptions, and sources. Another team may need different capabilities or have different constraints. A recommendation only makes sense in the context of the use case and evidence that produced it.