Most organizations do not have a survey problem. They have evidence scattered across forms, files, systems, sites, languages, and time. Start here to connect it around a real decision.
Connected Data Intelligence joins evidence from forms, spreadsheets, CRMs, documents, notes, sites, languages, and reporting periods so quantitative, qualitative, and longitudinal evidence can be read together.
The earlier course name focused on surveys and feedback. Those lessons remain useful: attrition, pre/post analysis, open-ended responses, multilingual feedback, demographic subgroups, and change over time still matter. But a survey is only one input. Most organizations have evidence distributed across too many places, with no reliable way to know which records describe the same person, organization, program, site, award, or reporting period.
This is especially visible in membership organizations, chapter networks, federated programs, multi-site nonprofits, and funders working with many grantees. Each part may collect sensible data locally. The organization still cannot answer a shared question without weeks of reconciliation.
A connected evidence layer makes four relationships explicit:
Once those relationships are governed, a team can read numeric change beside the words that explain it, compare sites without silently mixing definitions, and distinguish a missing outcome from a missing data source.
| Impact Measurement asks | Connected Data Intelligence asks |
|---|---|
| Are we measuring the right change? | Can the required evidence be connected correctly? |
| How do organization and funder priorities align? | Where do those requirements live across files and systems? |
| What definition and disaggregation should govern each metric? | Which existing fields match that definition, and which do not? |
| What decision should the evidence support? | Is the evidence complete, current, traceable, and comparable enough? |
Impact Measurement defines the evidence logic. Connected Data Intelligence makes the evidence usable. Case, Grant, and Portfolio Intelligence place both inside a real operating workflow.
Bring one decision question and the actual materials needed to answer it. Do not begin by drawing a perfect future architecture. Begin by mapping evidence already created during work.
| Move | What to do | Artifact |
|---|---|---|
| Map | Inventory forms, files, systems, documents, owners, cadence, and identifiers. | Source map and one decision question |
| Connect | Define persistent identities and relationships across people, organizations, programs, sites, and time. | Identity and relationship rules |
| Clean | Standardize definitions, missing values, categories, units, dates, and open text without erasing the source. | Transformation log and data dictionary |
| Read | Analyze quant + qual, cohorts, attrition, change, duration, sentiment, and subgroup patterns together. | Evidence review with limitations |
| Prove | Connect claims back to records and produce a cited narrative for the intended audience. | Traceable result and visible gaps |
For one decision question, create a table with these columns:
The map often reveals that most required evidence already exists and the missing piece is the relationship between sources. In other cases, it reveals that a claimed outcome—such as job placement or sustained employment—has no source at all. Both findings are valuable.
A training provider may have applications in one form, baseline data in a spreadsheet, attendance in an LMS, mentor notes in documents, exit responses in a survey platform, and employment follow-up in a CRM. A report may describe completion and confidence change yet still fail to prove placement because employer requisitions and placement records were never connected.
Connected Data Intelligence does not hide that gap. It shows which claims are supportable, partial, unsupported, or blocked by identity.
For the decision question below, create a source map across all available forms, spreadsheets, CRMs, documents, notes, surveys, and reporting files. Identify the unit of analysis, persistent identifier, time period, quantitative fields, qualitative fields, owner, refresh cadence, and known quality risks. Then classify each required claim as supported, partially supported, unsupported, or blocked by identity. Do not propose a dashboard until identity and definition problems are clear. Decision question: [insert one real question]
Use the result to choose the next lesson. If identity is broken, work on connection. If open text is unusable, clean it. If waves cannot be compared, work on longitudinal structure. If the evidence is complete but the narrative is untraceable, move to cited reporting.
It connects evidence across systems, files, sites, and time so quantitative, qualitative, and longitudinal data can be analyzed together with source traceability.
Those terms describe one input. The larger problem is connecting many evidence sources around a persistent identity and a shared decision.
No. Start with one decision and connect the minimum sources required to answer it responsibly.
A warehouse centralizes infrastructure. Connected Data Intelligence focuses on governed identity, meaning, sequence, and evidence use across workflows.
Use Impact Measurement to define what evidence should mean; use Case, Grant, or Portfolio Intelligence to embed it where evidence is created.
Next: Clean open-ended evidence → · or Define what the evidence should prove →
ChatGPT, Claude, and Gemini are fine for a quick test — but not for an answer you'll put in front of a funder or board. When it has to hold up, run it in Sopact Sense.
Connect one evidence workflow in Sopact →
