AI Product Management P1: How do you know if your organization is ready to support AI?

Stella Liu
Oct 13 · 5 min read

Learn how to set yourself up for success in the long term by enabling cross-functional teams and executives on AI.

This article is part of a series that breaks down AI product management into 5 distinct phases. The introduction to these series starts here.

Introduction

This article will dive into the most important phase of all: Should you even use AI? Where should you apply AI and why? During this time, it’s critical to determine if the cross-functional teams around you are ready for AI. You’ll be best served to enable them before you even begin so you are set up for success later.

Consideration 1/2:

Should I even use AI?

Short answer:

Not necessarily.

Long answer:

It can be easy to jump straight into the most complicated AI implementation, but you should take a step back. Look at the problem you want to solve, break it apart, and then deliberate if AI is best suited for the job. You may find that traditional programming could solve your problem.

When is AI better than traditional programming? AI tends to out-shine traditional programming in two main areas. One, when you are dealing with unstructured data (i.e. not in a database). And two, when you are dealing with fluctuating data inputs.

“Unstructured data” is data that isn’t organized in a database like photos, images, video, and sound. AI allows you to tap into this data by ‘structuring’ it into a usable format for a machine to understand.

Additionally, AI can help deal with fluctuating data points better than traditional programming. Traditional programming techniques used to organize unstructured data require developing a long list of rules that need manual hand-tuning. AI techniques are better suited here to simplify code and increase performance.

On the flip side, sometimes programming techniques can be more efficient than AI. For example, if you want to find the word “dog” in text, you could use a keyword search for that. But if you want to determine sentiment from text, then you are pushing into AI territory. Your data scientists will understand that nuance. As the product manager, you should strive for the simpler method first and be open to discussing trade-offs with your data science team.

However, it is not a binary AI v. traditional programming. Often-times, many applications use AI and programming techniques together. Once a machine understands unstructured data, traditional programming can then take over.

Let’s think about an example of a self-driving car. When AI models can take in unstructured data and identify a potential obstruction, you can program the system to warn the human driver. AI and traditional programming often work together to actualize an end to end scenario for the user. In fact, you can think about defining an AI solution as AI = ML + traditional programming.

Traditional programming and AI together can also help manage the cost of errors. For instance, if you are building a loan rating AI model in finance, you have to account for the legal risk of calling a “poor” loan “good.” You can program a rule that alerts a human to do further expert review on loans that had out of usual range inputs.

Consideration 2/2:

Are people ready?

Short answer:

Think beyond your immediate team and secure budget for the long term.

Long answer:

You should enable the teams around you. The earlier the better. Your executives must understand that AI is a long term investment because of the ongoing iteration that will have to occur to get it right.

Here are some key considerations for the teams around you:

Support team:

How will the support team differentiate between defects related to the AI model versus the software? Who should they route the request to — the data science team or the developers — and when?

Lab services team:

In the B2B context, it is common to have a “lab services” team to help enterprise customers deploy and customize their solutions. What happens if customers want to retrain the model for their specific needs? How do you enable your lab services team to handle these requests? How do you enable your lab services team to be ready for that? This will involve up-skilling and creating new business processes. It will help you to have a ‘champion’ in the organization to work with you to develop these changes.

Sales team:

On the sales side, they’ll run across a mix of enterprise customers out in the field. Some customers will be experts in data science. They will want to know the ins and outs of the algorithm and what the model’s accuracy metric is. You’ll have to equip the sales team to be able to answer those questions with confidence. On the flip side, the sellers may run into customers who can be too enthusiastic about AI and what it can do. The sellers need to learn how to manage high expectations while still show the solution’s value to the customer. You’ll have to equip the sales team with the right FAQ documents, training sessions, and positioning so they can manage this wide range of customers.

Marketing team:

For marketing, there is a fine line between wanting to tell a good story and overselling. How will the marketing team describe the AI product in a way that’s clear on what it does today vs tomorrow? In the beginning, it might be helpful to have a strict QA process with them. I wasn’t afraid to review the different messaging that comes out and have SMEs sanity check the messaging before it’s published.

Executive team:

It’s important to start socializing early that AI product management is a ‘cycle’. You’ll need to plan for the long term and secure data science resources that can help you in later phases even after the product releases. You don’t want to ask for short-term resources.

Data science team:

How do you respond to feedback on the model and improve it over time? AI is very probabilistic — if the data changes the model needs to change — so there needs to be a way to capture feedback. Having a team on standby that can review the feedback and enhance the model over time will serve you well.

IT team:

Your IT team will need to deploy the model and they will need to know how to deploy it and if there are frequent iterations.

Test team:

They need to understand how to deal with probabilistic errors.

Overall Tips

  • Err on simplicity. The simpler the method you can use to solve it, the better.

Stay tuned! There will be more articles released in the next couple of weeks that will provide some best practices on how to set priorities and train your AI model.


Stella Liu was a Product Manager at IBM Watson IoT where she helped build her team’s first AI-based product at scale. She loves to talk about AI, product management and environmental sustainability.

Please reach out to her at LinkedIn for any questions or comments!

IBM Watson

AI Platform for the Enterprise

Stella Liu

Written by

I like to demystify complex tech. PM @ VC-backed Indigo Agriculture. Former IBM Watson PM & Fulbright Scholar. https://www.linkedin.com/in/stellaliu93/

IBM Watson

AI Platform for the Enterprise

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade