SST: System Storyboarding Techniques

by

Richard A. Zahniser, CASELab, Inc.

Storyboarding was invented by Walt Disney in the 30's, as a technique for roughing out his cartoons^[1]. After his death, many of his top people left the company, and began to apply the technique to a wide variety of problem solving situations, always with positive results^[2]. The storyboard is "low-tech groupware"; enabling groups to work more effectively together anywhere without a big investment in computer hardware and software.

1.     Basic Storyboarding

Storyboarding consists of writing ideas or sketches on index cards, Post-its^TM or sheets, and arranging them on a vertical board. If the board is porous, the ideas are fastened with pins; if the board is smooth and/or hard (like a whiteboard or "FoamCore"^[3]) Post-its are an ideal choice.

The Fundamentals

Room Setup. The storyboard is the focus of the group; chairs and tables should be arranged to aid this focus. Figure 1 shows the basic setup for a small group. A space between the two tables makes it easy for additional helpers to step up and work on rearranging ideas on the board. It's important that group is facing the storyboard rather than each other -- this is a significant step toward consensus.

Supplies. Everyone should have a fresh yellow 3x5 Post-It pad, and a black Sharpie pen or equivalent. All pens should be the same color and weight, and black is most legible. Ball-point or fineline felt pens will not be legible at a distance. 

Procedure for Creation

1.   The facilitator stands at the storyboard.

2.   Someone writes the problem or topic on a Post-It and the facilitator puts that at the top center of the storyboard.

3.   When a player has an idea, they write it down on a Post-it, read it aloud, and present it to the facilitator. If they don't read it, the facilitator reads it before s/he puts it on the board.

4.   Everyone writes concurrently.

 

Sidebar

Rules. There are not too many rules.

Rule 1. No criticism

Simple, but broader than you think. Criticism of a person's writing style (sloppy, too light, etc.) will usually cause that contributor to dry up for ten minutes or longer. Discussion of any single idea, even if it is positive, will break the flow.

Tips. There will be a lull when you have exhausted the initial rush of ideas. The facilitator should stop, and wait silently for more ideas. They will come, more slowly, but they will come. In a larger group (8 to 12) the ideas will come so fast that you need two people putting them on the storyboard. With a very large group, you should start in the upper left corner and work across and down, putting Post-its close together. Otherwise, you will have to close them up later. You can expect a large group to generate 400 ideas in a half hour if it's the right group and you've stated the topic clearly!

Warm-up

The group should warm up first. My favorite warm-up topic is

"Foods that begin with 'B'."

I put "Foods with B" on a Post-It and put it at the top of a smallish (4x5) storyboard. Then I wait. Silently. I would wait forever, but it usually only takes about 10 seconds.

After we've created four or five ideas per person, we reorganize them by categorizing them and ranking one or two of the categories. 

Strategies for Reorganization

Categorize First. Put together columns of categories. Ask the group to think of categories for reorganizing. If no global categories are immediately apparent, you may wish to cluster ideas without stating why they are clusters. The explanation, if you need one is, "they look like they belong together". To cluster, you can get two or three people working on the board, and encourage the others to kibitz.

Out of Scope. If you see that some things don't belong on this board at all, move them over to another board, or even an available wall space or window. If you're sure they are really out of scope, mark them with an "O" in a different color. Postpone thinking about them for now. Don't throw them away!

Name the Clusters. If it hasn't done so yet, the group should step back, look at the clusters and come up with a set of names for them. This will clarify the rationale for clustering. Make different colored labels and put them above the categories. It's not productive to argue about the names. They don't need to be "right", right now, because they are easy to change.

Rank or Sequence. Within each category, you can rank or sequence things top to bottom. Don't struggle too much with this, you can do it more than once.

2.     Advanced Evaluation Techniques

After you've tried some basic storyboarding, study the techniques below to improve your quality and productivity.

Categories

Sometimes categories can be used to test the group's previous thinking about some of the individual ideas.

System Models. Nature has provided a number of models which follow the pattern:

                        Beginning          Middle               End

We frequently exploit these in building organizations or systems. You may wish to use this as the rationale to cluster cards on the left, center, and right portions of your storyboard. Do this as a first cut at clustering. Then step back and look for a few minutes without discussion. You may see some ready-made generic categories appear, such as:

                        Setup                Run                  Support

                        Input                 Process            Output

  Setup              Input                 Process            Output              Interface

 

Org Charts. The typical organization is broken into functional areas. If you are working with functions, these ready-made functional areas are a good thing to use, since they're familiar. For example, in a library system, the areas might be

            Acquisition        Circulation        Serials              Patron

                                                                                    Manage­ment                                        

A word of caution about using these functional areas; the very fact that they are familiar should make them suspect. They probably will not lead to a brand new systems approach. To ensure that, I recommend the event-driven approach developed by McMenamin & Palmer.^[4] Storyboarding is an excellent way to generate both business and system events.

Don't Force It. Don't try to force categories. They should arrive out of a group consensus, and this takes a little time (a good allotment is 10 or 15 minutes.) If people get possessive about their own categories, take a break or go to lunch!

Ranking

Once you have things in columns, you can order them by importance. This may be a sequential process. If so, the group should go back to one facilitator, taking the group to consensus on each ordering.

Important vs. Trivial. Start by putting obviously important things at the top and obviously trivial things at the bottom. You can usually get fast consensus about these.

Compare Post-Its. Take two items and compare them. Actually take the Post-Its off the board and hold them in front of the group. Hold them one over the other, and see if you get consensus (nodding heads) on that ranking. If not, reverse them. As a handy heuristic, a superior should be twice as important as a subordi­nate. If you can't get agreement on ranking them, put the two items side by side on the storyboard.

Sequencing     

Some items form a natural sequence. Do this, then do this. Step 1, 2, 3. Take advantage of this. Steps can go from top to bottom, and you can number them with smaller (1.5") Post-Its.

Stimulus - Response. A common sequence proceeds from stimulus to response. For instance, in dealing with events in the system space, we have a chain of events that look like this.

 

External      System       Activity       Activity         Response          External

 Event         Stimulus                                                                 Activity

 

Omissions. As you sequence, omissions will show up. For example, you may not have a ready-made response for a stimulus which you thought of during the creation phase. This constant work-in-process review is one of the key reasons that SST produces a quality product; it is essential that you fill in the holes as you go along.

Stratifying

Levels or layers may appear as you cluster. When a rationale for the layers becomes apparent, discuss it briefly and then make up labels to attach to the left or right edge of the storyboard.

Levels of Abstraction. A natural way to level things uses "levels of abstraction." Everything at a specific level has the same level of detail. Samples: 

            Logical                   or                            Strategic

            Physical                                                Tactical

                                                                        Operational

You can separate levels with a black line, made from a strip of black 1/2" masking tape.

3.     Capturing SST Results

The SST session is fast and furious, and produces a lot of output which is immediately visible to the group, but people outside the group need to see that output as well.

Capture

It is tempting to use a laptop computer to capture results as you go along. We have tried this, putting the operator "front row center." It works, but the board is in continual flux during the SST, and the intermediate results are usually not that important. Do it this way:

1.   Define a capture team. Two or more typing members of the team who are familiar with the laptop(s) they will use.

2.   Tape Post-Its together. When you are through moving Post-Its, tape them together with long strips of transparent tape. Otherwise, you may see your work disintegrate before your eyes!

3.   Capture on breaks. Move either board or laptop so that you are directly in front of the storyboard. Capture quickly, because you may need the board for something else.

4.   Build an outline. Build your capture document as an outline, with major numbers for columns and decimal sub-divisions to enumerate top-down details. 

5.   Formalize the output. Use bold and italic headings to emphasize categories, rankings, and to generally enhance the output.

Reviews

The group has invested considerable time and energy in the products which are produced by SST, and it's important that these results don't "plop."   To ensure that group members and outsiders don't trivialize this work:

1.   As soon as possible, the group should review (individually first, and then as a group) the material and make corrections as required.

2.   The group formulates conclusions and recommendations based on their work

3.   Everything is presented to upper management, preferably in a group setting.

4.     Our Experience with Storyboarding

I've been storyboarding since Jacquie Vierling (GE) introduced me to it in 1988. Almost without exception, it has worked astoundingly well, based on the following criteria for group success.

     It elicits almost total participation

     It produces lots of high-quality output in a short period of time.

     It uses the group as a resource to create more synergy than any other technique I've seen.

     It creates a feeling of consensus among group members without extended discussion. In turn, that leads to greater team productivity

     The group can easily "self-facilitate" the session (because of the lack of discussion and conflict).

     It's fun.

Applications

Here are products that our collaborators have produced in record time using SST.

   Organization / Activity Charts, Functional Decompositions, Work Breakdowns

   Mission Statements, Goals & Objectives, Critical Success Factors

   Problem Analysis, Benefits & Costs, Quality Factors

   Events & Stimulus-Response Chains

   Objects, Entities, Relationships, Normalized Attributes

   Module Charts, Object Life-Cycles

   Plans, Time lines, Gantt Charts

   Seminar Content and Outlines

   Test Scripts

Caveats

Sometimes, storyboarding does not fit the nature of the task. There are also some traps to watch for.

Networks. Storyboarding works best when the items elicited are related hierarchically, rather than when relationships form a network. Building such a network requires a smaller group (two or three) and more time. I suggest breaking the large group into smaller sub-groups to build data flows, E-R diagrams or state-transition diagrams.

Detail Work. Very detailed work -- such as programming, writing documentation, putting items into a data dictionary -- is sequential and is best done solo or in pairs.^·[5]

Definitions. It may appear that you need to "define your terms" before you start storyboarding. Do not do this in the new group. It is a "black hole" of discussion and delay!

Commitment. Participants should be at the session when it begins. Do not try to catch someone up. They cannot come in later. Furthermore, they should commit to stay until the end, and to participate fully.

Problem People. Initially, some people will have a different agenda and will hold back. If they do not participate, the end product will suffer. Most will join in eventually, however, and group pressure will help with this.

5.     In Conclusion

Using Post-Its is more profound than it first appears. Frequently, delegates to a group have an initial position to uphold. Post-Its fragment this position into individual ideas, which must stand or fall on their own. The group builds (and owns) a new consensus position from these ideas. Multiple storyboards can help the group to develop multiple perspectives on the problem or system.

SST is low-tech groupware, so your initial investment is small. Try it if you're looking for a better way to improve group productivity and product quality.     

##

Figures

Figure Sidebar-1.

 

Figure 1. Room layout for small group storyboarding.

About the Author (updated)

During the 60's, Mr. Zahniser built complex systems for ITT and IBM. In the 70's, he was a consultant, team-leader and manager for CIBAR, Inc. and System Development Corporation, specializing in real-time and data-based systems. During the 80's, he wrote and taught professional seminars for CIBAR Systems Institute and Learning Group International (Structured Design & Programming, CASE Hands-On) and built a number of PC-based applications to apply new analysis, design and prototyping techniques and tools. In 1990 he founded CASELab, Inc. which focused on improving software team productivity. He was the publisher of LabNotes: a journal on group software productivity, and the architect of Design By Walking Around, an N-dimensional approach to system analysis and design, and JMPSSTART, a highly structured team approach to rapid software development. In 1999, he decided to watch the Turn of the Century from Belize, Central America.  You can see his further adventures at http://belizenorth.com.

Author's Present Address: 15 Eighth Avenue, Corozal Town, Belize, Central America.  Telephone 501-402-0300, Email rickz@usermail.com.

About this Article

This article originally appeared in the September 1993 issue of American Programmer.  Permission to copy is granted, provided that the article is presented in its entirety, with suitable credits.