Trust

How OutcomeLens works in your org

The people in your data trusted your program with their answers. Here's how OutcomeLens treats that trust, in plain terms.

Your data stays in your org.

There's no outside copy. Forms, answers, and scores are records in your Salesforce org, alongside your contacts. Reports are built from those records when you open them.

Only invitations leave. They go out by email, or by text through your own texting provider.

Salesforce reviewed it. OutcomeLens passed Salesforce's security review on June 16, 2026, before it could be listed on AgentExchange.

For your admin

OutcomeLens is a managed package in the MJRSCO namespace: Apex classes, Lightning Web Components, and custom objects, installed from AgentExchange. It runs on Nonprofit Cloud, NPSP, and Health Cloud, and doesn't need OmniStudio.

Templates, forms, responses, scores, and snapshots are all records in your org. There is no outside database holding a copy.

The people in your data are protected.

Consent comes first. Before any link is created for a text invitation, OutcomeLens checks the do-not-text field on the person's record. If they've opted out, or the field can't be read, nothing is sent, and the reason is logged so your staff can see why.

No license, no sign-in. Participants answer on their own device through your Participant Site. They don't need a Salesforce license of their own, and the link opens only their form.

Small groups stay private. In reports, any demographic group with fewer than 5 participants shows as "<5", with no exact count or percent.

For your admin

Participant Site forms use Experience Cloud guest access, and each link carries a token that opens only that person's form. Texts are handed to your texting provider through a delivery Flow.

Staff can also give the same forms inside a screen Flow, for example during intake. The scoring and the records are the same either way.

Your org needs a Participant Site license for self-service forms; see pricing.

Every number shows its work.

Every rate says who it's out of. A count and its total, never a bare percent.

Every number says where it came from. The instrument, its scale, and the dates it covers.

Thin data is flagged. A result that rests on thin data is marked "interpret with caution", and measures built to stay apart are never averaged together.

Two kinds of evidence, labeled apart. Validated instruments and forms your program built are labeled differently, in the app and in printed reports.

For your admin

OutcomeLens scores three ways: score-based (points roll up into categories and classify into tiers your admin defines), value-based (a raw number, such as income, classifies directly), and experience measures (NPS, CSAT, and CES on a shared percent-favorable scale).

Polarity is set per category, so on an instrument where lower is better, a falling score reads as improvement. It ships with no default classifications.

Every form is compared with the participant's baseline and their previous form, and goals and outcome targets track goal against actual.

Questions about any of this? Talk with us.