Skip to the guide

JSM cross-project analysis framework

Compare Customer Demand Across JSM Projects, Not Raw Scores

Compare Wezy customer demand across JSM projects without adding incompatible scores or double-counting organizations. Build a scoped, explainable review worksheet.

Wesy November 11, 2026 5 min read

In short

Bring project-level evidence into one review, but preserve each project's scope, formula, observation time and permissions. Recalculate a common analytical view only when inputs are compatible, and do not sum organization counts as unique customers.

Runs on Atlassian
Company-managed JSM Cloud Existing JSM permissions apply
In this guide

Do not combine Wezy demand scores from different JSM projects until you know they use comparable inputs and scoring policies. A larger number can reflect a larger coefficient, a different eligible audience or more accumulated participation—not greater demand for the same product outcome.

This guide describes a team-managed analytical worksheet built from authorized project evidence. It does not claim that Wezy includes a global portfolio dashboard, cross-project vote merging or a site-wide customer identity export. The documented app provides project-level voting and demand analysis in company-managed JSM Cloud projects.

Define the question and the unit of comparison

“Which request should the product team investigate next?” compares request-level evidence. “How many distinct customer organizations support this capability?” requires deduplication across a defined set of requests. “Which service team is more successful?” is not answered by either count alone.

Write the decision before collecting files. Keep the request key, project, observed outcome/problem, observation time, V, O, R30, applicable W and active formula/version. Include the exported scope and coverage limitations. The CSV reconciliation guide explains how to validate each input file before joining it to another.

Use one row per request for current metrics, unless the actual source schema requires a more detailed grain. Do not silently add repeated rows from a historical or organization-level export.

Prove why the headline scores may be incompatible

Here is a fictional comparison using only documented Wezy formula components:

Request V O R30 Active policy Displayed score
Project A request 5 3 2 Standard: V + 3O + 2R30 18
Project B request 6 2 1 Advanced: 2V + 6O + 4R30 28

Sorting the displayed scores places B first. But B’s policy is exactly twice the Standard formula. Under a common offline Standard-formula comparison, B would be 6 + 6 + 2 = 14, while A remains 18.

That does not prove A should be delivered first. It proves that the displayed 28-versus-18 comparison mixed scales. Label a recalculated worksheet column clearly as an analytical comparison; do not overwrite Wezy’s app-managed fields or present it as the score actually active in Project B.

The scoring guide and calibration walkthrough provide the formula definitions. Historical scores retain their original policy context; a retrospective analysis with a new formula must not masquerade as the original historical output.

Keep organization breadth separate from summed counts

Suppose one request represents organizations Alpha, Beta and Gamma, while another represents Beta and Delta. Their O values are 3 and 2. The sum is 5 request–organization relationships, but the union contains 4 distinct organizations.

These names are illustrative. In real analysis, use an authorized, stable identifier in a consistent identity scope when one is actually available. Do not merge organizations solely because their display names look alike; names can change, and identifiers from different Jira sites do not automatically establish the same real-world customer.

When you only have aggregate O values and no authorized matching evidence, you cannot calculate the exact union. For the observed sets in this two-request example, it is bounded between 3 and 5. Report that the unique count is unavailable rather than picking a convenient value.

The same warning applies to V. One account may support several requests; summing request-level votes counts support relationships, not necessarily distinct people. Optional attribution also means organization counts do not represent every active vote. See Wezy’s attribution definitions.

Control audience and timing differences

A project open to a broad customer audience and another restricted to a pilot group have different opportunities to accumulate support. Without a defensible eligible-audience denominator, “votes per project” is not a participation rate or a fair performance target for the teams.

Compare observations collected over an agreed interval and record remaining time differences. A current view and a historical comparison cannot be treated as interchangeable. Wezy’s historical periods require complete covered UTC dates; retention does not establish that missing observations exist. The history guide explains the boundary.

Organization weights can also encode different local policies. Two segments both named “Strategic” may have different definitions or overrides. Document the policy owner and actual coefficients rather than matching configurations by label alone.

Use a comparison contract before a joint review

Contract question Evidence required Action when missing
Are we comparing the same type of customer outcome? Human-reviewed problem statements Separate unlike requests
Is the record grain compatible? Request identity and actual export schema Normalize with a documented mapping
Are score scales comparable? Active coefficients, weights and version Compare components or label a common offline calculation
Are observed periods compatible? Observation time or complete covered UTC periods Keep separate or mark unavailable
Can we deduplicate organizations? Authorized stable identity evidence Do not claim an exact unique total
May this audience read the combined record? Project permissions and approved reporting scope Redact, aggregate appropriately or do not combine

Copying evidence into a spreadsheet can widen its audience even when Jira itself remains correctly restricted. Keep the shared decision record limited to what its readers are entitled to receive; avoid importing private request descriptions just to make a portfolio table look complete.

Keep a reusable comparison contract

Download the cross-project review worksheet. It records the source grain, active policy, observation time, permitted audience and identity assumptions before the ranking. It is a blank team template, not a native Wezy export or an integration.

Stop a numeric comparison when the source formula or record grain is unknown, or the required period is uncovered. Stop combining identifiable records when the audience is not authorized or the identity scope cannot be established. You can still summarize the unresolved requirements; you cannot repair them by treating missing values as zero, matching organizations by name or inventing a participation denominator.

Make a shortlist, then explain the decision

Use compatible demand components to identify questions worth investigating. Add assessed effort, feasibility, dependencies and obligations in separate columns owned by the relevant people. Do not hide those judgments inside an unexplained “combined customer score.”

If two projects describe related needs, record the relationship in the review worksheet without claiming votes were merged. The duplicate-request workflow explains why preserving private source requests and choosing a reference need is different from transferring support automatically.

A useful final statement is: “These requests were compared on a common analytical formula using the documented observation scope; distinct organization totals were not deduplicated where identity evidence was unavailable.” That is stronger than an attractive ranking whose units are inconsistent.

Wezy Customer Voting supplies project demand evidence. The cross-project interpretation remains your team’s responsibility. Start with the JSM prioritization framework when turning that evidence into an accountable product decision.

Published by the team behind Wezy Customer Voting. Product capabilities are linked to their official documentation; worked examples are illustrative.

Let customers vote on your JSM portal.

Install Wezy from Atlassian Marketplace, open one request for voting and see the first vote in Jira.

Runs on Atlassian Read the full documentation

The site works without optional cookies. Your choice is saved on this device for 180 days.

Read more about data and Google