Making Money with Mediocre AI Models: A Guide for Business Stakeholders

Making profit out of poor-performing models

In the world of AI, it’s easy to assume that only the most accurate models can bring value to your business. However, this is far from the truth. In fact, even mediocre models can be transformed into money-making machines with the right strategies. In this article, we’ll explore three real-life examples of how we turned poor-performing models into profitable assets, and share three strategies to help you do the same.

Can a Mediocre Model Really Boost My Business?

Yes, it can! Even if your model is only slightly better than random, it can still bring value. Let’s look at an example. We once built a model to predict if a certain product was going to be sold in the next week. The model’s overall predictive performance was mediocre at best, achieving only 60% balanced accuracy. However, the model showed high predictive performance when we looked at the top 10% and bottom 10% predictions. These “tails” were already optimizing millions of dollars. So, despite its overall poor performance, the model was moved to production and generated a positive cash flow.

What if My Model is Only Slightly Better Than Random?

In the healthcare domain, we had a model with 80% ROC AUC, which is reasonable but not great. However, when we looked at the extremes (cases where the model was quite sure a patient had a disease or was healthy), the predictive performance was over 90%. This allowed us to prioritize patients in severe conditions and save resources, optimizing a lot of value despite the model’s overall mediocre performance.

Can a RecSys That Fails for Newcomers Still Be Profitable?

Absolutely! We once built a recommendation system that was good at understanding the behaviors and purchase trends of recurring customers, but terrible for newcomers. Instead of trying to improve the model for the entire population, we decided to profit from the segment of recurring customers. This approach allowed us to generate additional revenue and produce a cash flow to keep working on solving the larger case.

Do you want to further discuss this idea?

Book a meeting with Kelwin Fernandes

Meet Kelwin Learn More

How Can I Transform My Mediocre Model into a Money-Making Machine?

Here are three strategies:

Strategy 1: Look at the tails

If your model isn’t as good as you’d like, look at the top and bottom cases as we have shown in this blog post. This can help you optimize value.

Strategy 2: Retarget your model

If your model is good at predicting a certain segment, retarget your model to that segment. This can bring good performance and cash flow to fund the project’s long-term vision.

Strategy 3: Ignore technical KPIs and focus on business KPIs

Instead of focusing on accuracy, precision, etc., focus on how much money you’re producing, how many customers you’re saving, or how many errors you’re preventing. This will help you see the value your model is already bringing.

Course, Templates

Data Ignite

Learn how to move from technical aspects of your AI models into a successful business implementation.

Learn More

In conclusion, don’t be discouraged if your AI model isn’t perfect. Even mediocre models can bring significant value to your business with the right strategies. At NILG.AI, we’re here to help you navigate the world of AI and make the most of your models, no matter their performance.

Like this story?

Subscribe to Our Newsletter

Special offers, latest news and quality content in your inbox once per month.

Signup single post

This field is for validation purposes and should be left unchanged.

Recommended Articles

Can Your Business Optimize AI Predictive Models?

Predictive models are transforming the AI landscape. They can forecast future events, identify past occurrences, and even predict present situations. However, building a successful predictive model is not as simple as it seems. To achieve an effective predictive model, you need to consider three crucial moments: the prediction time, the prediction window, and the data […]

Read More
Is Your Business Ready for Generative AI Risks?

Generative AI is a powerful tool that many companies are rushing to incorporate into their operations. However, it’s crucial to understand the possible risks associated with this technology. In this article, we’ll discuss the top nine risks that could impact your business’s readiness for AI integration. Stay ahead of the curve, and make sure you’re […]

Read More
Can the STAR Framework Streamline Your AI Projects?

As a manager dealing with AI projects, you may often find yourself overwhelmed. The constant addition of promising projects to the backlog can lead to a mounting technical debt within your team, forcing you to neglect the core aspects of your business. Here at NILG.AI, we have a solution for this challenge: the STAR framework. […]

Read More