The goal of the Triastorm technique is to find novel ideas, to solve problems. It relies on the scientific study of problem solving, but Triastorm defines a pragmatic and simple process practictioners can use for business or everyday life, in group or individually.

In the future, software will be made available to support the triastorm technique.

The Triastorm breaks down the problem solving process into four processes:

• Generate ideas: the problem solver starts with five or more ideas. It is important that these are as diverse as possible. They may be partially adequate or totally defunct. During later cycles of this process, the problem solver will combine solutions to fill up the solution space.
• Elicit dimensions: This process helps in exploring the solution space, thereby guiding the problem solver in his search for adequate solutions.
• Rank dimensions: Establishing the relative weights of the dimensions found during elicit dimensions is useful to explore the solution space efficiently. Efficient exploration can be done by visualising the two or three most important dimensions, then indicating the part of the solution space most likely to yield good solutions.
• Position ideas: This process places the ideas in the solution space, thereby discovering gaps and thus potential for new ideas. It also helps in discovering which parts of the solution space are unlikely to generate adequate ideas.

At the end of each cycle, the problem solver chooses the process used in the next cycle, until the problem is solved or time runs out, in which case the best ideas may be used as input for a new Triastorm session.

The name of the Triastorm technique derives from two important aspects used:

• Triadic elicitation to explore the solution space
• Brainstorming to generate a wide range of ideas

Before you start a Triastorm session

To use the Triastorm technique with success, it is necessary to define the problem as precisely as possible. Don’t confuse your goal with the means to reach it: this way you constrain yourself and make novel solutions less probable.

For example, suppose you want to become famous. It is better to state as problem description “how to become famous” than “how to become famous by writing a book”. If your problem is that you want to write a book but don’t know how to begin, state “how to write a book” as your problem description.

Use GERP as your first cycle: first generate five or more ideas, then elicit one or two dimensions, then rank the dimensios, then position the ideas on the dimensions.

You need a few index cards, a piece of paper, and a pen. In case you have the Triastorm software, that and a computer is all you need.

Process 1: Generate ideas

When you first cycle through generate ideas, quickly write five or more ideas down—one per index card, and number them starting from 1. These ideas can be

• Solutions you already tried (but which failed)
• Ideas that seem fruitful but lack a few things to make them perfect
• Outright silly or bad ideas

It is more important to have a rich set of ideas than to search for the perfect idea in this stage. The ideas generated in this stage will largely serve to construct the solution space.

When you cycle through generate ideas in a later phase of the Triastorm session, combine existing ideas to generate new ones. Do keep looking out for totally new ideas, however.

Also, take a look at the visual solution space and try to fill in gaps. For example, in figure 2 the lower right quadrant is still empty. You can try to

• Contrast: find ideas that are the opposite of the empty quadrant (in this case, the opposite of idea 2)
• Combination: combine ideas from neighbouring quadrants (in this case, 4 or 6 combined with 1, 3 or 5).

When you have already gone throuhg elicit dimensions, consider to perform position ideas with the newly generated ideas now.

Process 2: Elicit dimensions

Knowing which dimensions structure the solution space is important because this will guide your search: you won’t lose time by summing up ideas that won’t work and don’t bring any new insights. 

Triastorm uses triadic elicitation: select three ideas in a random way (by turning the index cards and shuffling them). When you have just performed generate ideas, you are allowed to select one new idea instead of a random idea from the list.

Now select the one out of three which is different from the others and state the difference.

For example, suppose the three items selected are “write a paper”, “give a speech” and “convince a local newspaper to interview me”. You could then select the last one to be different from the others, because in that case, you don’t compose the text yourself. The dimension would then be expressed as “do it yourself” versus “get help from others”.

Some dimensions can be expressed as a single term (for example, “amount of work”), others are more conveniently described with two opposite terms. Both ways are fine.

From the moment you have two dimensions or more, select the two most important ones with rank dimensions and visualise them by drawing a quadrant like figure 2.

In each case, try to find new dimensions. Don’t get stuck on the ones you already found. If you can’t find a new dimension because of the three ideas you selected, skip this step.

Process 3: Rank dimensions

After you’ve gone through elicit dimensions, you can start rank dimensions. That way, you will concentrate on the most important dimensions when searching for solutions.

Ranking can be a simple matter of putting them in a list, the most important ones at the top. Important here means that the dimension is likely to make a difference between good and bad ideas.

Typically, you will keep two dimensions to focus on (which can change during the Triastorm session). Visualisation of two dimensions is easy, visualizing more is more difficult unless you use a computer program.

A second activity in this process is constraining the solution space by delimiting dimensions.

On the quadrant diagram, use color to indicate delimiters of dimensions so you can effectively visualize the part of the solution space most likely to yield useful ideas. If your idea list contains the current, nonworking solution, that part of the solution space is likely to be in the nondelimited zone.

Don’t delimit a dimension if that is not necessary: if the value of an idea on the dimension does not matter for its quality, leave the dimension unconstrained.

Example of delimiters on a dimension: if you discover that the solution has to be cheap to carry out, delimit the dimension “price” by coloring the “cheap” part of the solution space.

If multiple dimensions have delimiters set, the solution space becomes narrower and focused. This guides the problem solver to that part of the solution space that contains the best ideas.

Process 4: Position ideas

The fourth and last Triastorm process is position ideas on the dimensions you elicited. You can choose a scale that is convenient for you, though usually you don’t need more than five marks on each dimension (–, -, 0, +, ++).

The importance of this process is to indicate the gaps in the solution space: those parts that are delimited but still empty. It can also help to generate new ideas by indicating which ideas need to be combined or contrasted.

Use this tool when brainstorming individually. Use small pieces of paper to note ideas and dimensions and place them on this printed sheet.

When generating new ideas, be sure to position them regularly. Also, after each cycle of elicit dimensions, position the current ideas on the newly found dimensions.

Triastorm in practice: an example

The problem

As an example, we’ll tackle a well-known problem: “How to make business meetings start on time in our organisation?” You might find this exercise fun and useful to repeat. The outcome will differ because it is specific to your organisation.

First cycle: Generate ideas

• (1) Be angry at latecomers
• (2) Offer a drink at the beginning of the meeting
• (3) Discuss the most important things first
• (4) Send extra invitations before the meeting
• (5) Send an SMS to latecomers at the time of the meeting

Second cycle: Elicit dimensions

These three ideas were chosen randomly to extract the first dimension:

• (1) Be angry at latecomers
• (3) Discuss the most important things first
• (5) Send an SMS to latecomers at the time of the meeting

I chose to group 1 and 5 on the one hand, 3 on the other hand. 1 and 5 both confront the latecomers directly, while 3 does so indirectly. The dimension is called “directness”.

A second choice of three ideas resulted in

• (2) Offer a drink at the beginning of the meeting
• (4) Send extra invitations before the meeting
• (5) Send an SMS to latecomers at the time of the meeting

Here I group 4 and 5 on the one hand, 2 on the other. I’ll call the dimension “positive stimulation” (that’s exactly what 2 does).

Third cycle: Rank dimensions

At this time, I have a slight preference for the dimension “positive stimulation” over “directness” because of our corporate culture. However, I won’t delimit either dimension now.

Fourth cycle: Position ideas

Now that we have two dimensions and several ideas, we can represent them visually:

Fifth cycle: Generate ideas

We see that one of the four quadrants is empty. If we are looking for innovative ideas, we can search for ideas in this quadrant. It’s hallmark is high directness and high positive stimulation. What action can we take towards latecomers, but such that it is positive? For example, we could

• (6) take time to welcome latecomers extensively

While this might look like a bad solution, it actually isn’t. Still, if it was outright bad, we would include it in our list of ideas because it could help us discover dimensions of the solution space we didn’t think of.

Sixth cycle: Position ideas

We just put idea (6) in the empty quadrant and continue with the seventh cycle.

Seventh cycle: Elicit dimensions

Let’s keep our new idea (6) and randomly select two more from the list:

• (1) Be angry at latecomers
• (3) Discuss the most important things first
• (6) Take time to welcome latecomers extensively

It’s obvious we could group (1) and (6) versus (3) because (1) and (6) confront latecomers directly and (3) doesn’t, but that would not lead to a new dimension. Therefore it’s more fruitful to look for other similarities.

What if we combine (1) and (3)? Do they have something in common that (6) doesn’t, apart from the dimension “positive stimulation”? Let’s try to combine (3) and (6) then. Both can be part of the meeting agenda, while (1) can’t. So let’s call the dimension “agendability”. I interpret this dimension as: “can the idea be part of the meeting agenda and thus of the meeting notes”?

Eight cycle: Rank dimensions

For our corporate culture, I would rank the dimensions found so far as follows:

1. Positive stimulation
2. Agendability
3. Directness

Neither dimension needs to be delimited.

That means we’ll have to adapt the diagram in the next cycle.

Ninth cycle: Position ideas

Again we have a part of the solution space that is empty: the part that is “positive stimulation” but can be part of the agenda and the meeting notes. Time for a tenth cycle.

Tenth cycle: Generate ideas

In search for positive action that can be in the meeting notes, we come up with

• (7) make a short endorsment for each meeting participant that shows up on time.
• (8) record a “timeliness indicator” in the notes of each meeting, start with “good” no matter how late the meeting actually starts.

Both ideas seem to be worth of further investigation. I especially like (7) because it works on the need of every human being to hear something positive about himself. Now that I think of it, I’m going to try this.