Simplify AI decisions with this quick framework

By now, you should have a list of ideas for how to use AI in your business. (And if not, you can revisit part 1 and part 2.)

But GenAI misuse brings with it some real risks. Things like: 

• Exposure of sensitive or confidential information. 
• Unintentional distribution of misinformation. 
• Jeopardizing intellectual property. 
• Irreversible damage to brand reputation and client trust.

Some of these can be avoided with a solid company AI policy. Others depend on how your organization adopts AI: the use cases you select, how AI-literate your team is, which technologies you use.

These risks can be enough to hamstring conversations before your company ever gives AI a chance. You may have already experienced this first-hand — a brainstorming session quickly spirals into: 

• What about data privacy? 
• How much will this cost? 
• What if the AI hallucinates?
• What are the legal risks? 
• Is our team properly trained?

Hold up. 🖐️ Today we’ll give you a straightforward framework to redirect these conversations — an easy way to evaluate use cases to minimize AI risk and maximize AI reward. 

You’re reading part 3 of a 4-part series on AI use cases. We’re exploring frameworks and strategies to think about: 

• Part 1: What knowledge workers do.
• Part 2: What AI can do.
• Part 3: What AI should do.
• Part 4: How to mitigate risk when adopting AI.

AI Use Cases in Business: How to Choose the Best Ideas for Your Company

Step 1. Brainstorm Potential AI Use Cases

Our Knowledge Work Matrix ignores risk by design, and for that same reason we’ve separated “What can AI do in our company?” from “What should AI do?”: 

The brainstorming phase should be free of the whens, wheres, hows, and what ifs. Ideally, they’re two separate conversations at two separate meetings on two separate days. Because your mission during each of these discussions is quite different. 

First, create your list of use cases. Then, ruthlessly evaluate them to find the right use cases. Remember that even a handful of the right ideas, executed well, can be massively valuable. 

Step 2. Evaluate and Prioritize AI Use Cases

We’ve found this simple 2x2 framework to be most helpful in evaluating and prioritizing AI use cases:

Ask yourself: 

• Value: How valuable would automating this activity be? In general, how valuable is this task to my company? 
• Risk: What are the risks of automating this task? How concerned are we about them?  

Below are suggestions to help evaluate value and risk, but you can use this framework to facilitate a discussion using whatever criteria best fit your company.

How to Think About Value 

• Alignment with goals for AI adoption
• Alignment with business goals / strategic fit 
• Revenue potential
• Productivity potential (i.e. increased output)
• Efficiency potential (i.e. hours saved per month)
• Future-proofing core offerings 
• Innovation impact (differentiation / market positioning)

How to Think About Risk

• Industry regulations
• Input risks (confidential information, sensitive data) 
• Output risks (exposure, reputation) 
• Lost opportunities and trade-offs (inability to copyright outputs*)
• Investment (time and effort to train AI and upskill team) 

*We’re not lawyers, but recent reports suggest that under current U.S. law, work produced by AI is not protected by copyright.

Your company’s ability to evaluate AI’s capabilities — and recognize its risks — is directly proportional to your team’s understanding of AI.

Each use case should fall into one of four quadrants: high value-low risk, high value-high risk, low value-low risk, or low value-high risk.

Here’s how we think about each:

• High value, low risk: Automate. These are ideal AI use cases to pilot and have massive potential to drive productivity, efficiency, and/or performance in your company.
• High value, high risk: Mitigate. The use case you’re currently proposing is risky. But is it possible to lower the risk? 
• Low value, high risk: Delegate. This task is not a good fit for AI, but it might still be important to your company. (For example: Meeting face-to-face with customers or participating in media interviews.) Remove from your ideas bank and allow people to own 100%.
• Low value, low risk: Eliminate. Remove from your list. (And if the task has little value in general, consider eliminating it altogether.)    

It’s important to note that your company’s ability to evaluate AI’s capabilities — and recognize its risks — is directly proportional to your team’s understanding of AI. 

You cannot adequately assess AI risks on the horizon if your vantage point is full of blind spots. Your company’s AI decision-makers must be at the forefront of learning about and experimenting with this technology.  

Next week, we’ll dig into that second quadrant — mitigate, including three specific ways to lower the risk of AI adoption.

Meanwhile…

• Consider the different ways AI could bring value to your company — increasing output, saving time, augmenting current capabilities, optimizing performance, etc. Then rank them in order of priority to you. 
• List the biggest risks your company could face by adopting AI.  
• Plot a few of your use case ideas on the matrix, and look for quick wins that land in the “Automate” quadrant.

Question(s) to Consider: 

🪄 AI has handed your company a magic wand — What would you choose to do with it first? Save time? Drive revenue? Expand into new services or capabilities? Something else? (And why?) 

You're Reading a Preview. Want More?

Welcome! This is a preview of our newsletter. Join dozens of marketing and business leaders as we explore frameworks, templates, and advice to navigate the challenges that come with growth. Sign up to receive new insights every other week.