---
title: "Data-Driven Decisions Canvas"
description: "Participants sort project questions onto a two-by-two grid of discovery versus optimization and qualitative versus quantitative data. Groups identify inquiry gaps and assign concrete research methods to key questions."
last_updated: "2026-09-07"
canonical: "https://unitedwetransform.com/exercises/data-driven-decisions/"
---

# Data-Driven Decisions Canvas

Participants sort project questions onto a two-by-two grid of discovery versus optimization and qualitative versus quantitative data. Groups identify inquiry gaps and assign concrete research methods to key questions.

**Time:** 30 minutes. Source recommends 15 to 30 minutes; planned for 30 minutes to fit subgroup placement, method matching, and a debrief discussion.
**People:** 4 to 24. Subgroups of 3 to 6 people at separate canvases. Groups of 4 to 6 work as a single team. For 7 participants, split into one group of 3 and one group of 4, or work as a single group of 7 around one large canvas.

## Choose this exercise

**Purpose:** Classify project questions across inquiry types and data forms to select balanced research methods.
**Use it when:** Use during project kickoff, research design, or quarterly planning when a team needs to choose which data to collect next.
**Skip it when:** Skip when research questions, metrics, and data collection protocols are already fixed by compliance or contract.

## Materials and tools

- One large 2x2 grid canvas per group (flip chart paper or whiteboard)
- Sticky notes in two colors per group (yellow for questions, blue for research methods)
- Fine-tip markers (one per participant)
- One Research Methods Reference Guide per table

## Before participants arrive

1. Draw a large 2x2 grid on each group's flip chart or whiteboard.
2. Label the horizontal axis: Left = Unknown (Discovery), Right = Known (Optimization).
3. Label the vertical axis: Top = Qualitative (Thick Data), Bottom = Quantitative (Big Data).
4. Place markers, sticky notes, and printed method guides at each table.

## Say this to open

To make solid decisions, we need both broad numbers and deep context. We will map our project questions across these two axes to see what we are asking and what methods will answer them best. Start by drafting individual questions silently.

## How to run it

### 1. Draft individual questions (3 minutes)

Deliver the opening script. Participants write project questions silently on yellow sticky notes, aiming for two or three notes with one question each. Prompt for both exploratory questions and specific performance questions. Anyone wishing not to write may dictate questions to a partner to transcribe onto a note. Notes can remain unsigned and anonymous.

### 2. Introduce the canvas axes (3 minutes)

Explain the axes. The horizontal axis separates Discovery (exploring the unknown) from Optimization (improving an existing process). The vertical axis separates Qualitative or Thick Data (why or how) from Quantitative or Big Data (how many or how often). Give neutral examples for all four quadrants: 'What values guide customer choices?' (Qual/Discovery); 'Why do users abandon signup?' (Qual/Optimization); 'Which new regions show unexpected traffic?' (Quant/Discovery); and 'What is our churn rate?' (Quant/Optimization).

### 3. Map questions to quadrants (8 minutes)

In subgroups, participants pool notes. Taking turns, each member places a note on the canvas. A participant may read their note aloud, place it silently without speaking, or ask a peer to place it for them. The table discusses briefly to agree on the quadrant that best fits the inquiry.

### 4. Assign research methods (8 minutes)

Subgroups review their mapped questions. Using blue sticky notes, participants tag at least one research method beside each key question cluster, consulting the table reference guide (for example, ethnographic interviews, exploratory analysis, or A/B testing). Groups circle any quadrants that have no questions.

### 5. Debrief and reflect (8 minutes)

Reconvene the room. If multiple groups exist, ask a spokesperson from one or two subgroups to name their most crowded quadrant and one empty quadrant. For a single group, review these areas together. Then pose the debrief questions to the full room. Take three or four brief responses from volunteers across the room to close.


## Debrief questions

- Which quadrant holds most of your team's current focus, and what blind spots does that create?
- Which questions require qualitative stories rather than numeric metrics to answer well?
- What is the first research method your team will run next week based on this map?

**Finished output:** A mapped 2x2 matrix per subgroup pairing project questions with specific qualitative and quantitative investigation methods.

## Remote

Pre-build 2x2 canvases on a digital whiteboard for each subgroup of 3 to 6 people. Provide digital sticky notes in two colors. Keep participants in the main room for drafting and the axis walk-through. Open breakout rooms for mapping and method selection. Bring all participants back to the main room for the final debrief.

## Hybrid

Form separate all-remote and all-in-person subgroups if numbers allow. If a mixed group is necessary, assign an in-person partner to transcribe spoken questions and place physical sticky notes on behalf of the remote participant while sharing a table camera view of the physical canvas.

## Access and participation choices

- Participants may dictate questions verbally to a partner to write on sticky notes.
- Participants may place sticky notes silently onto the canvas or leave them in a shared pool without reading them aloud.
- Spokesperson reporting during plenary debrief is entirely voluntary and focuses on group patterns rather than individual authorship.

## If the session gets stuck

**Participants write proposed solutions or tasks instead of questions.**

Ask the writer to rephrase the task into what they must verify first, beginning the note with Why, How, or How many.

**All notes cluster exclusively in Quantitative/Optimization.**

Pause the subgroup and ask what they still do not understand about user motivations. Prompt them to draft two qualitative discovery questions to test the opposing quadrant.


## Terms used here

- Thick Data: Qualitative context and stories explaining why and how people behave.
- Big Data: Quantitative metrics and logs showing scale, rates, and frequency.

## Participant material: Research Methods Reference Guide

Use these suggested methods when matching questions to quadrants:

1. Qualitative / Discovery (Top-Left: Thick Data + Unknown)
- Ethnographic interviews: In-depth conversations in the user's natural environment.
- Empathy mapping: Documenting observed needs, frustrations, and thoughts.
- User narratives: Collecting lived stories to find unexpected patterns.

2. Qualitative / Optimization (Top-Right: Thick Data + Known)
- Usability testing: Observing people complete specific tasks to find friction.
- Targeted feedback interviews: Asking users why they abandoned a known process.
- Focus groups: Structured discussions with defined user segments on existing features.

3. Quantitative / Discovery (Bottom-Left: Big Data + Unknown)
- Exploratory data analysis: Mining system data to spot emerging behavioral patterns.
- Trend and cross-dataset correlation: Analyzing shifts across time or market segments.
- Demographic segmentation: Identifying clusters with high unmet activity.

4. Quantitative / Optimization (Bottom-Right: Big Data + Known)
- A/B testing: Controlled experiments comparing two versions of a live feature.
- Funnel metrics tracking: Measuring drop-off rates at each defined transaction step.
- Churn and retention analysis: Calculating duration and exit rates over specific periods.

## Sources and adaptation

Observed in the Participatory Methods repository, which attributes the exercise to the Integrated Data Thinking framework developed by Sudden Compass.

Retained the 2x2 matrix structure, the discovery versus optimization horizontal axis, the thick versus big data vertical axis, and the research method pairings from the source. Added explicit neutral examples across all four quadrants during step two. Adjusted grouping instructions to resolve arithmetic edge cases for 7 participants, allowing either a single canvas or split subgrouping. Extended duration to 30 minutes to provide time for method tagging and debriefing.

[Participatory Methods](https://participatorymethods.com/activities/data)

[Participatory Methods](https://participatorymethods.com/activities/data-driven-decisions/)

Current run sheet: https://unitedwetransform.com/exercises/data-driven-decisions/