decision making for AI projects

Decision Making for AI projects: Potential vs Feasibility Matrix

In one of our previous articles Launching an AI project: first steps checklist, we explained the importance of planning in AI projects, mentioning the potential vs feasibility 2×2 matrix.

Used mainly by business consultants, this matrix is an effective way of presenting and interpreting information regarding the value of a new project or idea.

In this process, you should involve all business departments and team leaders to have a 360 degree understanding of the pain points of the business and what to prioritize. Every department of your business has a different agenda and will seek your attention. You’ll need the support from the right tools to cut through the noise and be able to identify your business’ priority based on short term visibility. This is where the Potential vs Feasibility Matrix comes in.

Potential vs Feasibility Matrix


From the BCG to the Ansoff matrix, decision-makers can use these tools to frame their products status and their potential for growth. When it comes to AI projects, seeking help from AI consultants will allow you to:

→ Assess the feasibility of the project in terms of time, technology (including data) and team. 

→ Assess the potential of the new project to bring real value to your business, including financial payback, problem solving and profit. 


Based on these two key factors you can build a matrix and find your priority project. (We suggest you read this blog and know if are you ready for AI?).

The matrix is built upon two independent and continuous variables: potential and feasibility. On these two axes your matrix is divided into four blocks, or segments, in which is described the action to take for each “option” (project) within the segment. The options are mapped out in the matrix and as a result, you’ll be able to identify the opportune strategy:

  1. Improve
  2. Divest
  3. Scale back
  4. Invest

The resulting matrix should look like this:

potential vs feasibility matrix for decision-making


Improve.
Projects allocated in this box have a high potential for the business but are less feasible mostly due to internal factors. AI is a real help and source of profit, however, there is a gap between the skills and technologies implied. You could potentially overcome this gap by investing in your own R&D team, but realistically this solution will be costly in terms of both time and money. AI projects in this box are more likely to move right into the “Invest” box as soon as you decide on a solution (i.e. outsourcing AI development). Read more about this in our previous article: Two things to consider before starting an AI project: People and Data.

Divest.
It’s quite obvious that projects that end up in this box just won’t work. These AI projects will take a longer time to become feasible and generate the potential for your business.
Note: things can change! Tech is evolving, you are evolving too and so projects that end up here can grow in potential and feasibility. Don’t get rid of them yet, leave them somewhere you can go back to when the time is ready. 

Scale back.
Not all great ideas work. Although you have the capability to develop it, the world out there isn’t ready for it. When the potential for your business success is missing, external factors that are out of your control are involved (i.e. shifts in trends, new technology introduced). Generally, these projects are set up for a long term win (or no win at all).
Perhaps you just need to reduce the size of the project, cut out the unnecessary to enjoy a greater benefit. It took a while for people to adapt to chatbots; businesses were investing a small amount on developing bots, testing them on their customers, letting them get used to the bot and slowly improving it by adding investment in time (training), and resources.

Invest.
BINGO! These AI projects are feasible and have great potential. Get final buy-in from other business departments and start work on your AI roadmap. You’ll soon discover that AI projects can bring benefits in more than one department, solving more than one of your teams problems. 

Once you have mapped out your AI project ideas into the matrix, it would be clear what AI project you should pursue. It is critical that you involve a diverse pool of people in this process, including external AI experts who can help to accelerate the development of your idea. 

Alternatively, this matrix can also support your decision on current AI projects. If you feel that your projects aren’t taking the expected turns, I suggest you come back to this matrix and reevaluate your projects current state; which project to improve? Which project to make more feasible (i.e. you need more data or more people involved)? Which project you should stop investing in as it’s not bringing any value? Which project to reduce in size, resources or amount of time spent on it? And finally, which project has the right balance between potential and feasibility?


Need more reading before you get started? Here is an article that explains how to avoid getting stuck in the middle between a Proof of Value and Production: How to manage AI project.

Blogs that might interest you:

Our mission

Skim’s mission is to empower people to use data more effectively and to demystify artificial intelligence. Rather than holding up the common narrative of machines replacing humans, we see how machines can help humans to have easier lives and better businesses.

Supported by

Contact

London office
27 Finsbury Circus,
London EC2M 5NT

Portugal office

+44 207 129 7497
sales@skimtechnologies.com

skim-logo