Join our daily and weekly newsletter for the latest updates and exclusive content on industry-leading AI coverage. learn more
Although companies know that AI cannot be ignored, the real question is not when it comes to building it. What can AI do? And what’s more important is where do you start?
In this article, we present a framework that helps businesses prioritize AI opportunities. Inspired by project management frameworks like Rice Scoring Model for Prioritization, it balances business value, time to market, scalability, and risk, and helps you choose your first AI project.
Today, AI is successful
AI doesn’t write novels or run a business yet, but where it’s successful is still worth it. It does not expand and replace human efforts.
In coding, AI tools improve task completion speed by 55% and code quality by 82%. Throughout the industry, AI automates repetitive tasks, including email, reporting, and data analysis.
This effect is not easy. All AI issues are data issues. Many companies struggle to ensure that AI works because their data is stuck in silos and they are inadequate integration or simply not ready for AI. It takes effort to make your data accessible and available. So it’s important to start small.
Generated AI works best as a collaborator rather than as an alternative. By drafting emails, summarizing reports, improving code, and more, AI can reduce the load and unlock productivity. The key is to start small, solve real problems and build from there.
A framework for determining where to start with the generation AI
Everyone is aware of the possibilities of AI, but when it comes to making decisions about where to start, they often feel paralyzed by the vast number of options.
Therefore, it is essential that there is a clear framework for assessing and prioritizing opportunities. It provides structure to the decision-making process and helps businesses balance trade-offs between business value, time to market, risk and scalability.
This framework draws on what we learn in collaboration with business leaders and combines actionable insights with proven approaches such as rice scoring and cost-benefit analysis to help businesses focus on what’s really important. Provides results without unnecessary complexity.
Why a new framework?
Would you like to use an existing framework like rice?
It’s useful, but it doesn’t fully explain the probabilistic nature of AI. Unlike traditional products with predictable results, AI is inherently uncertain. “AI Magic” fades quickly when it fails, producing bad results, strengthening bias or misleading intentions. So it’s important that markets and risks are important. This framework helps prioritize projects with bias against failure, achievable success and manageable risks.
By adjusting the decision-making process to explain these factors, you can set realistic expectations, prioritize effectively, and avoid the pitfalls of chasing ambitious projects. In the next section, we will analyze how the framework works and how it can be applied to your business.
Framework: 4 core dimensions
Business Value: What is the impact? Start by identifying potential values for your application. Do you increase revenue, reduce costs, or increase efficiency? Are you consistent with your strategic priorities? High Value Projects directly address core business needs and provide measurable results. Time to Market: How quickly can this project be implemented? Evaluate how fast you can move from an idea to an deployment. Do you have the data, tools, and expertise you need? Is the technology mature enough to execute efficiently? Fastest implementation reduces risk and delivers value faster. Risk: What’s wrong? : Assess the risk of failure or negative outcomes. This includes technical risks (do AI produce reliable results?), recruitment risk (do users adopt tools?), and compliance risk (do they have data privacy or regulatory concerns?). Low-risk projects are more suitable for initial efforts. Ask yourself if you can only achieve 80% accuracy, is that okay? Scalability: Can solutions grow with the business? Assess whether your application meets future business needs or can handle higher demand. Consider the long-term feasibility of maintaining and evolving solutions as requirements grow or change.
Scores and prioritization
Each potential project is scored on these four dimensions using a simple 1-5 scale.
Business Value: To what extent will this project be affected? Time to market: How realistic and quick is it to implement? Risk: How well is risk manageable? (Low risk scores are excellent.) Scalability: Can applications grow and evolve to meet future needs?
For simplicity, you can use t-shirt sizing (small, medium, large) to get dimensions rather than numbers.
Calculating Prioritization Score
Once each project has sized or scored in four dimensions, you can calculate a prioritization score.
Here, using α (risk weight parameter) allows you to adjust how risk affects your score significantly.
α=1 (standard risk tolerance): Risk is weighted equally to other dimensions. This is perfect for organizations with an AI experience or are willing to balance risk and reward. α> (Risk Aversion Organization): Risk has more influence and punishes higher risk projects more strongly. This makes AI suitable for new organizations, and is also suitable for environments where operating in regulated industries or where obstacles can have serious consequences. Recommended values: α= 1.5 to α=2α<1 (high risk, high reward approach): Risk has little impact, supports ambitious and high reward projects. This is for businesses that are happy with experiments and potential failures. Recommended value: α=0.5~α=0.9
Alpha allows you to adjust the prioritization formula to match your organization’s risk tolerance and strategic goals.
This formula ensures that projects with high business value, reasonable time to market, scalability, but manageable risks will rise to the top of the list.
Applying frameworks: practical examples
Let’s explain how your business can use this framework to decide which Gen AI projects to start. Imagine you are a medium-sized e-commerce company trying to leverage AI to improve operations and customer experience.
Step 1: Brainstorming opportunities
Identify both internal and external inefficiencies and automation opportunities. The output from the brainstorming session is as follows:
Internal Opportunities: Internal Meeting Overview and Action Item Automation. Generates a product description for your new inventory. Optimize inventory replenishment forecasts. Perform sentiment analysis and automatic scoring for customer reviews. External Opportunities: Creating personalized marketing email campaigns. Implementation of chatbots for customer service inquiries. Generates automated responses for customer reviews.
Step 2: Build a decision matrix
Evaluate each opportunity using four aspects: business value, time to market, risk, and scalability. In this example, we assume a risk weight value of α=1. Assign a score (1-5) or use the size of your t-shirt (small, medium, large) to convert them to numbers.
Step 3: Verification with Stakeholders
Sharing decision matrix with key stakeholders to tailor priorities. This may include marketing, operations and customer support leaders. Include inputs to ensure that the selected project matches the business goals and is buy-in.
Step 4: Implementation and experiment
Starting small is important, but success depends on defining clear metrics from the start. Without them, we cannot measure values or identify where adjustments are needed.
Start small: Start with a proof of concept (POC) to generate product descriptions. Use existing product data to train your model or take advantage of pre-built tools. Pre-define your success criteria – time savings, content quality, new product launch speed, and more. Measure results: Track important metrics to your goals. In this example, we will focus on the focus. Efficiency: How much time does the content team save with manual work? Quality: Is the product description consistent, accurate and attractive? Impact on your business: Will speed and quality increase sales performance or customer engagement? Monitoring and Verification: Track metrics such as ROI, adoption rates, error rates and more regularly. Verify that the POC results match expectations and make adjustments as needed. If a particular area is poorly performing, refine the model or adjust the workflow to address those gaps. Iteration: Use lessons learned from POC to improve your approach. For example, if your product description project works well, scale your solution to handle seasonal campaigns and related marketing content. By gradually expanding, you can continue to provide value while minimizing risk.
Step 5: Build your expertise
Few companies start with deep AI expertise. That’s fine. It is built by experimenting. Many companies start with small internal tools and test in low-risk environments before scaling.
This step-by-step approach is important because there are often hurdles of trust in the company that must be overcome. Teams need to trust that AI is reliable, accurate and truly beneficial before it can invest more deeply or use it on a large scale. By starting small and demonstrating incremental value, you reduce the risk of overcommitting to large, unproven initiatives while building that trust.
Each success will help your team develop the expertise and confidence they need to tackle bigger and more complex AI initiatives in the future.
I’ll summarize
There’s no need to boil the ocean with AI. Just like with cloud adoption, we begin small experiments and scales to reveal value.
AI must follow the same approach. Small, learn and expand. Focus on projects that minimize risk and bring quick wins. Use these successes to build expertise and confidence before expanding into more ambitious efforts.
Gen AI has the potential to transform your business, but it takes time to succeed. Thoughtful prioritization, experimentation and iteration can help you gain momentum and create lasting value.
Sean Falconer is a resident AI entrepreneur of Confluent.
