5 Steps to scoring and clustering concepts
Where to go first?
Analyzing innovation opportunities and clustering innovation concepts
This post covers a preliminary method I have been using to rate, rank, and cluster innovation opportunities for companies to decide…
What ideas do we move forward with?
I use this method after I interviewed users, identified insights, synthesized the insights, and created a list of ideas (opportunity areas a.k.a. groups of similar ideas) and I am overwhelmed. I have a wall of post-its clustered into groups and I feel like a hungry dolphin looking at a school of fish.
How do I move forward? Where do I even start?
- List the ideas or opportunity spaces in a spreadsheet
List the concepts down the page, one in each row. It helps if they are all at the same level of scope. Example: Creating ‘new employee incentives aligned with production’ and creating a ‘mobile application for operators’ are at a similar level. Adding ‘creating autonomous transportation methods’ is at a higher level concept because there are multiple types of transportation and levels of autonomy. Use your judgement to make sure the list is consistent.
2. Define your rating attributes
At the head of the columns put the attribute names. An attribute would be a variable used to rate the concept. Below are the attributes I used:
Desirability — Do users want this concept?
Feasibility — Can we build this concept?
Viability — Will the concept make money? (generate value)
Criticality — How critical is this concept?
Vitality — How strong is this concept?
Complexity — How complex is this concept?
Feel free to add others based on your interest and priorities. I have also added market size, obtainable market, net present value, and return on investment in the past. This also helps in prioritizing clusters of concepts.
3. Rate the concepts
For each of the attributes develop a ‘rating’ to score the concepts. The rating may be subjective in some cases, but take a shot at defining them. The output of the analysis is meant to be directional not prescriptive so forcing a rating when you are not sure is acceptable. It provides a preliminary direction of what concepts make the most sense to pursue together.
Sample rating system:
Desirability — 0 to 3(0 = no proof of desire from users, 3 = a paying user)
Feasibility — 0 to 3(0= no one has built it yet, 3 = we build these constantly)
Viability — 0 to 3(0 = no profit estimation, 3 = highly profitable)
Criticality — # of concepts that depend on this opportunity
Vitality — # of concepts that are strengthened by this opportunity
Complexity — # of concepts that precede this concept (pre-requisite concepts)
4. Run your calculations
There are many ways to run your calculations. I decided to group the individual concept ratings (desirability, feasibility, and viability) and the relationship related ratings (criticality, vitality, and complexity) separately. I simply added them up. I then sorted the concepts, first by the individual rating total and then the relationship rating total.
How the concepts are ranked depend on the organization’s priorities. If they want small, easy, ‘quick win’ opportunities, sorting by complexity and feasibility first may make the most sense. This will identify the easiest concepts to implement. My method considers the strength of the individual concept first before its relationship to others.
Also, make sure that you treat each category accordingly. Example, a higher number in complexity has a negative effect (meaning it should lower the ranking) while a higher number in vitality has a positive effect (meaning it should improve the ranking). Make sure the effect of each attribute rating is appropriately accounted for.
5. Analyze your list
After you sort your list, look at the groupings of concepts together. Do any make sense to implement as a group? Does a group that is in a row stand out? Do the concepts make sense together? How might you change the sorting of the concepts? Why? Do any new apparent groupings appear from resorting?
I added conditional formatting to show where ratings get lighter or darker (lighter = better and darker = worse). This helped me to visually see what concepts may group together. I recommend looking at the individual strength of the concept and then the relationship to other concepts, but that is not a die hard rule. Judgement is needed to identify logical groupings (or clusters) for the opportunities.
At the end of the analysis, the team will have several clusters of opportunities to consider and the list can help inform growth strategies for the company going forward.
