Analysing starts with using the output of the scan phase and clustering your manifestations. In this blog I’ll help you find common themes and patterns amidst all your signals of change.

 

WHY CLUSTER?

Let’s first start with why you would want to cluster your findings. By clustering all the manifestations you’ve found you create groups of signals that tell the same story, known as ‘trends’. The goal of this clustering is to connect the dots and understand which of your spotted signals of change are similar.

You need to do this because any one observation is just that, a single observation. You now have to find out if your observations are signs of something bigger happening.

Clustering is about moving from information to meaning.

 

HOW TO CLUSTER

You will need to sift through your collection of signals and find the underlying values and needs related to each manifestation. Clustering is about combining an analytic mindset with a creative one; it goes beyond just keeping score and adding up types of new products or behaviours you have seen. You need to be able to see patterns in unexpected ways that might seem unrelated at first.

Finding patterns and relationships between manifestations is a continuous process. Professional trend researchers often analyse while scanning. They constantly compare a manifestation to others they have spotted before and have kept in the back of their mind.

 

Create an overview

For this exercise it is best to have all your spots together in one place. This could be on your computer, but I would recommend you to print them out. This way you can pick them up, easily move them around and put them aside when you want to. This is called a ‘trend evidence wall’.

As you can see on the image below, it allows you to physically walk into your database with manifestations. This will stimulate different parts of your brain while clustering then staring into a spreadsheet would.

You can also recreate this evidence wall feeling on digital online boards in Miro, Mural or any other whiteboard tool. As seen in the example below.

Cluster checklist

Be flexible and don’t pin yourself down to a fixed number of clusters beforehand; just see where a first round of clustering takes you. It can help to structure this process in a few stages. Consider the proposed steps in the cluster checklist below:

  • Prepare the session
  • Share manifestations
  • Cluster by values
  • Assign keywords
  • Reflect and adjust

Let’s go through every step by taking an in-person meeting as an example. You can also apply the same process for a live online session.

  • Prepare the session

When working in a team, reserve a time slot in every participant’s agenda. Assign a specific space for the cluster session where you can easily move furniture around and attach things on a surface. This space should be quiet, light and airy, and block out external stimuli. Make sure there is enough food and drink for the duration of the session. When working in a team you can assign participants homework, for example “bring printouts of five signals of change to the session”.

  • Share manifestations

When everybody is in the room, start out by collecting everyone’s signals and creating a display. Look at all the manifestations in silence individually and have everyone write down their initial thoughts on paper, sticky notes or any way you prefer.

  • Cluster by values

Discuss the signs amongst each other and try to find connections between them. Get a stream of thought going. Which signals are from different sectors or regions but are telling the same story? Move the evidence around to get inspiration from unexpected combinations.

Remember that the manifestations are representations of values and needs people feel are worth pursuing. This means thinking on a more abstract level than just combining signs into features or categories like “technology” or “clothing”. Cluster them based on values and needs, such as “control” or “connection”. Take another look at the leftover manifestations: are there more signals that belong to the trend clusters you have just created?

  • Assign keywords

Describe your clusters in a couple of keywords or a sentence. Use words that indicate what change in feelings, attitudes and values lie behind this cluster. Try to find consensus about the words assigned to the clusters, but also remember this is the first attempt to create them and you can still change them later.

  • Reflect and adjust

Do not rush the clustering process! It is important to take some time and space to reflect on it. During sessions you can take a break and go for a walk to clear your mind. After the first session, let it all sink in and reflect on it the next day or next week.

The output of clustering are drafts of trend clusters to be researched further in the next step of the trend analysis process: validate. To record the work you’ve done while clustering, save materials like sticky notes, and flipcharts, take photos of walls, and write down the clusters.

To wrap it up:

Most people who analyse trends for the first time feel a bit uneasy about where to start. It helps to set up your analysis in a structured manner. This also makes it easier to explain to others what you have been doing and communicate about your analysis in a transparent way. Clustering trends by using the checklist of steps guides you through the process of going from manifestations to meaning.

 

NEED HELP WITH UNDERSTANDING TRENDS?

Read my book for more in-depth information or contact me to for trend training or trend coaching sessions.