Developed by Dan Roam (The Back of the Napkin, 2008).
Matches the kind of question a problem poses (who, how much, where, when, how, or why) to one of six simple pictures, then sketches it to think and to explain.
Seven questions about lost sales deals, each matched to the simple picture that answers it
A sales team is losing more deals than last year, and each question below asks something different about why. The first is placed as an example. Select each other question, then the picture that would answer it: a portrait, a chart, a map, a timeline, a flowchart, or a plot of two variables. Illustrative.
- Not placedPlacedPortraitChartMapTimelineFlowchartPlot
- Not placedPlacedPortraitChartMapTimelineFlowchartPlot
- Not placedPlacedPortraitChartMapTimelineFlowchartPlot
- Not placedPlacedPortraitChartMapTimelineFlowchartPlot
- Not placedPlacedPortraitChartMapTimelineFlowchartPlot
- Not placedPlacedPortraitChartMapTimelineFlowchartPlot
- Not placedPlacedPortraitChartMapTimelineFlowchartPlot
The picture shows
The question asks
Your sort matched 0 of 7.Your sort matched 1 of 7.Your sort matched 2 of 7.Your sort matched 3 of 7.Your sort matched 4 of 7.Your sort matched 5 of 7.Your sort matched 6 of 7.Your sort matched 7 of 7.
- Portrait
- Who or what: draw the thing itself, such as a sketch of the typical lost account with its size, industry, and buyer.
- Chart
- How much: compare quantities side by side, such as bars of lost deals by quarter.
- Map
- Where: place things in relation to each other, such as territories shaded by loss rate.
- Timeline
- When: put events in order along time, such as the stages of a deal with the days spent in each.
- Flowchart
- How: show steps and the paths between them, such as who touches a deal from demo to signature.
- Plot
- Why: set two or more variables against each other, such as discount against deal size, lost deals marked, to see what drives the loss.
The sort
- When do deals stall?: Timeline. A question of when: the deal's stages along a line, with days spent in each, shows where time piles up.
- Which accounts are lost?: Portrait. A question of who: a sketch of the typical lost account makes the pattern visible at a glance.
- How many lost a quarter?: Chart. A question of how much: bars by quarter compare the amounts.
- Which regions lose most?: Map. A question of where: territories shaded by loss rate show the clusters.
- How does a deal close?: Flowchart. A question of how: the approval steps and handoffs, drawn as a flow, show the extra loop legal added.
- Is it price or fit?: Plot. A question of why: discount against fit score, lost deals marked, shows which variable separates wins from losses.
- Who signs for the buyer?: Portrait. Another who: a picture of the buyer's team, with the signer circled.
An example
A regional sales team was losing more deals than the year before, and three meetings with the vice president had gone in circles. Every chart in the deck was a bar chart answering how much, and the questions in the room were about something else.
The sales operations manager wrote each question on a card and named its type. Which accounts do we lose was a who question, so she sketched the typical lost account: mid-sized, food processing, a new plant manager as buyer. When do deals stall was a when question, so she drew the deal's stages along a line with the days spent in each; almost all the extra time sat between the technical review and legal. How a deal gets signed was a how question, and the flowchart showed a second legal loop added that year. Is it price or fit was a why question, and a plot of discount against fit score showed the lost deals clustered at low fit, whatever the discount. The next meeting took twenty minutes and changed the legal review.
Why it works
A problem that cannot be seen tends to get argued in circles, because each person holds a different picture of it in their head. Drawing it puts one picture on the table that everyone can point at and correct. The trouble is that people reach for the same picture every time, usually a bar chart, whatever the question.
Matching the picture to the kind of question fixes that. Who and what questions are answered by a portrait of the thing; how much, by a chart; where, by a map; when, by a timeline; how, by a flowchart; and why, by a plot that sets variables against each other. Once the question's type is named, the right picture is nearly obvious, and a simple sketch of the right kind often answers what a polished slide of the wrong kind never could.
The thinking behind it
Dan Roam Dan Roam, a creative director, set out the method in The Back of the Napkin: Solving Problems and Selling Ideas with Pictures (2008), one of his popular books on visual thinking. The books teach people to solve and explain business problems with simple drawings: start from the kind of question being asked, choose the picture that answers that kind, and sketch it before making anything finished. It is a popular method book, built from practice, without research behind it, and its test is whether the people in the room understand the problem faster.
How to run it
- Write the question in one sentence. State what the group is trying to understand, such as why deals are being lost this year.
- Split it into its questions. Most problems hold several smaller questions; write each on its own card so none hides inside another.
- Name each question's type. Label each sentence by what it asks about: a person or thing, an amount, a place, a time, a process, or a cause.
- Pick the matching picture. Use a portrait, a chart, a map, a timeline, a flowchart, or a plot of two variables.
- Sketch a thumbnail first. Draw it small and rough on paper or a whiteboard before making anything polished.
- Show it and redraw. Watch where people point, ask, or frown, and redraw the picture there.
Where it goes wrong
The common failure is drawing the habitual picture. Teams with dashboards reach for bar charts, and a bar chart answers how much even when the question is why. If a picture prompts the same debate as the words did, check the question type again.
The second is polishing too early. A finished slide invites comments on fonts and colors, and people hesitate to scribble over it. Keep the sketch rough until the picture itself is agreed.
The third is one picture for a problem that holds several questions. A single crowded diagram trying to show who, when, and why at once shows none clearly. Draw one small picture per question and lay them side by side.
When to use something else
- Tufte's Data Display Principles. When the picture is chosen and the data display itself needs to be honest, dense, and clear.
- Concept Map. When the aim is to lay out how many ideas relate to each other, beyond a single question.
- Issue Tree. When the problem needs breaking into parts that do not overlap before any picture is drawn.
Where it sits in SOLVE eX
Go to the source
- The Back of the Napkin: Solving Problems and Selling Ideas with Pictures, 2008