ProductHow it WorksBrief TypesUse CasesPricingFAQ

Use caseI know the options

Compare the options without hiding the evidence gaps

Turn vendor pages, reviews and threads into one structure where every criterion says what the evidence covers — and where it doesn’t.

Free plan · No card required

Known options, shared criteria

Options you already know

  • Linear
  • Jira
  • Shortcut

Shared criteria

  • Setup effort for a small team
  • Cost at scale
  • Migration effort from Jira
  • Reporting depth

One comparison, criterion by criterion

  • Setup effort for a small teamOption advantage · Linear

    Limited evidence · Shortcut

  • Cost at scaleUnresolved

    Limited evidence · Jira · No source coverage · Shortcut

  • Migration effort from JiraComparable
  • Reporting depthContext dependent
Not produced: Overall winnerNo such output

Coverage is marked only where it is limited or missing. Each criterion keeps its own takeaway; nothing adds them up.

The situation

The comparison usually happens in someone’s head, then gets defended in a meeting

  • Comparing tabs by hand

    Vendor pages answer the questions they want to answer. By the seventh tab you’re comparing whichever features happened to be described, and you can’t remember which claim came from whom.

  • Asking a chatbot for “the best tool”

    You get a confident answer with no way to see which criterion decided it, or whether the losing option lost on evidence or on absence of evidence.

  • Collapsing it into one number

    A weighted total hides the one thing a reviewer will challenge: which criterion actually mattered, and whether it was the best-evidenced one.

What a defensible comparison needs

The same criteria for every option

Including the cells nobody has evidence for yet.

Every cell traceable to a source

So a surprising assessment gets checked instead of argued about.

Thin spots marked as thin

So the conversation is about the trade-off, not about whose summary to trust.

Start here

Name the options. Your wording steers the criteria.

Web app · Comparison

“Compare Linear, Jira and Shortcut for a six-person product team moving off Jira”

Find sources & createMode · Comparison

The link fills in the prompt only — choose Comparison in the composer before you create.

Route in
  1. Options you name
  2. Comparison
  3. Comparison Brief

Say who the comparison is for — “a six-person team”, “moving off Jira” — because criteria come from your wording and the evidence. There is no criteria picker.

Known options, or not yet

Comparison lines up options. Whether you’re ready to compare is a separate question.

Comparison · “Compare Linear, Jira and Shortcut for a six-person product team moving off Jira”
Comparison Brief
Comparison · “Where do these pages agree and disagree about migration effort?”
Source Comparison Brief
Auto · “Find the best issue trackers for my team”
Research Brief

With no options named, Comparison compares whichever options the gathered sources establish — it doesn’t research which ones belong on your list. If you can’t name them yet, start in Auto and research first.

What you get

A row you can argue with, not a verdict you have to accept

One criterion, as the brief renders it. What’s inside a Comparison Brief

  • One criterion of four

    Cost at scale, one criterion of the comparison
    LinearJiraShortcut
    Evidence gap: Cost at scale
    		Supported: Per-seat pricing published; no independent cost analysis in this set source 1.Limited evidenceThe review and the pricing page give incompatible figures for the same tier source 2 source 5.No source coverageNo eligible source gives pricing for Shortcut.
    Coverage is incomplete and the two sources that do speak conflict. No option leads here, and the absence of evidence is not read as weakness.
    Cost at scale
    Coverage is incomplete and the two sources that do speak conflict. No option leads here, and the absence of evidence is not read as weakness.

    Each cell carries its sources and its coverage; the sentence under the row is the brief’s takeaway for this criterion alone.

How SourceFrame handles the evidence

Vendor claims don’t quietly become facts

Every cell is cited

Open a citation to read the extracted evidence, or the source in context, before a cell goes into a slide.

Perspective is marked

Vendor-published material is marked single-perspective, so you can see how much of the set is a company describing its own product.

Conflict stays visible

When a review contradicts a pricing page, both stay, and the criterion’s takeaway says the evidence conflicts.

AI chats stay context

A captured ChatGPT or Claude conversation can be in the set, cited separately as AI-generated context rather than as independent evidence.

A worked run

Tracker choice for a six-person team

  1. The prompt

    “Compare Linear, Jira and Shortcut for a six-person product team moving off Jira”

    Find sources & createMode · Comparison

    The link fills in the prompt only — choose Comparison in the composer before you create.

    “Six-person” and “moving off Jira” are the words that steer the criteria toward setup effort and migration effort.

  2. What came back

    • Option advantageLinear leads on setup effort
    • ComparableMigration effort is comparable
    • Context dependentJira leads if reporting depth is required
    • UnresolvedCost at scale — and no pricing source for Shortcut

    Four criteria, four different kinds of takeaway. The category names are for teaching; the brief writes each takeaway as a sentence under its criterion.

  3. The next meaningful action

    1. Add sources
    2. variant 2

    Find pricing sources for Shortcut, because a column with no evidence can’t be judged fairly. The better-evidenced comparison arrives as a variant beside the original, which stays Primary until you choose otherwise.

When someone has to sign off

The call is a separate brief, and you start it

From this brief
  1. Comparison Brief
each move below starts here
    1. Turn into…
    2. Recommendation Brief

    The comparison never makes the overall call, however clear it looks. When a decision has to be made, you turn it into a Recommendation Brief, which carries these criteria into Factors considered and states whether the call is confident, conditional, or made on insufficient evidence.

A different comparison

Options as the subjects, not sources

Comparison Brief asks

“How do these options compare on the criteria that matter?”

Versus

Source Comparison Brief asks

“What do these sources agree and disagree about?”

If the disagreement you care about is between your sources rather than between your candidates, compare what the sources say instead.

Where you can go next

  • Primary brief type

    Comparison Brief

    The matrix, coverage marks and criterion takeaways in full — and why the artifact has no verdict field.

  • Can lead to

    Recommendation Brief

    Where the overall call belongs, with its status stated. It carries this comparison’s criteria into Factors considered.

  • Too early if

    Research an open question

    You can’t name the options with confidence. Comparing an invented shortlist just moves the guesswork upstream.

  • Different comparison

    Compare what sources say

    When the disagreement is between your sources rather than between your candidates.

Put the options on the same criteria.

Free plan · No card required