How to use AI as a Product Owner

A Product Owner is responsible for maximizing the value of a product by defining its vision, managing the product backlog, and acting as the key link between stakeholders and development teams.

We will dive into all the ways you can use AI to become more efficient as a Product Owner (PO).

Creating the vision

As a Product Owner, an important part of your job is creating and adjusting your vision of the product you are responsible for. A vision is created by talking to customers, suppliers and colleagues, but also by looking into the data your product is collecting. AI can help to streamline this.

When using AI, it is good to follow best practices. For example, OpenAI explains how to best prompt GPT-5.

Talk to users

When talking to users, create a format you can reuse for each conversation. Not only will it help you structure the conversation, it will make it easier for AI to analyse the data.

For example, before the conversation:

• Define the goal: what are you trying to learn?
• Know the user: review the customer's profile, look at their usage data and potential previous feedback

Then, create a set of open ended questions you can use during the conversation. Only use this as a guide and dive deeper into specific topics if you deem fit.

Either take notes during the conversation, or ask the user whether you can record the conversation. If allowed, you can use a voice-to-text AI model to transcribe the conversation. There are many websites where you can transcribe for a small fee.

Then, put use the following prompt to analyse the conversations. We also provide an example result, so the AI better understands what we are expecting.

Find out what your competitors are doing

Another way to get ideas what to build, is looking at competitors. Although you should never copy a competitor's functionality one-on-one (as everyone's context is different), it can give you ideas.

ChatGPT's new agent mode can really help to quickly perform market/competitor research. Find the competitors you are most interested in, open up chatgpt and select Agent mode.

I use the following prompt structure for these use cases:

I am working at [name and url of your company], we help [explain what your company is doing]. We have a couple of main competitors, like [list your competitors]. Find out [mention what you are interested in]. Summarize your findings and write down all the takeaways for my company.

I also often look at competitor's open vacancies, as companies often tell future employees what they will be working on/what their vision is. You can include this in the prompt as well:

I am working at [name and url of your company], we help [explain what your company is doing]. We have a couple of main competitors, like [list your competitors]. Find out [mention what you are interested in]. Summarize your findings and write down all the takeaways for my company. Also search through the companies open vacancies to find interesting information. 

Turn your vision into a Proof of Concept

As a Product Owner, you are not creating the designs: you usually work together with a designer. However, it is important to be able to have clear requirements and to communicate these clearly to the rest of the team. AI can help refine your ideas and make them visual.

Creating entire User Interfaces (UI)

AI allows you to quickly generate entire user interfaces, with tools like V0.app, Google's Stitch or Loveable. In this example, we uploaded a screenshot of a finance app. The app just shows the current price of a stock. We asked the AI to add a new functionality: to show the user the current value of their portfolio. Therefore, the user needs to add how many shares they own of individual companies. Then, the app should display the current value.

Within 5 minutes, the AI recreates (a basic version) of the screenshot we uploaded. On top of it, it added the functionality we requested. This gives you as Product Owner a feeling how your feature could work. If you don't like a certain way the User Experience is implemented, you can prompt it to change it to your liking.

Within minutes, you can validate and try out your requirements and adjust them if needed.

Tips for prompting: make sure describe your requirements in detail. Do not assume the AI will understand what you have in mind. In our experience, the better you describe what you have in mind, the better the results will be.

Creating individual media assets

AI also allows you to quickly generate individual media assets, whether it is logos, icons or entire images. Nowadays, it is also extremely easy to edit images by describing what you want.

For example, we started off with the following prompt in Google's Nano Banana (this is the name of Google's image editing model):

Generate a minimalist, isometric line art illustration of a house with solar panels. Use a clean, desaturated color palette with dominant light grey, subtle light blue-grey for shading/reflections, and a soft green for accents. Ensure crisp lines, rounded edges on objects, and a simple, geometric representation for all details. The background should be a solid, uniform light blue-grey. Include subtle drop shadows beneath the objects.

The result:

Next, we asked "now, add a car in front", and the result came back within seconds:

Lastly, we asked "Zoom out and make an entire neighborhood":

This allows to quickly turn your ideas into something tangible.

Creating tickets with AI

As a Product Owner, your main task is creating tickets for developers to work on. Tickets are short descriptions and requirements of features you wish to add to your product. Many Product Owners find this a tedious task. Luckily, AI can help you create detailed tickets.

We use AI to first create all user stories. Then, we use a second prompt to write individual user stories.

Overview of all user stories

Nowadays, most AI models are "multi modal". This means that they understand text, but also images, videos and voice. We can use this to generate all relevant user stories/tickets to develop new features.

First, we need to get all the designs created by the designers. Then, we upload these designs in our favorite LLM (we use ChatGPT) and provide the following prompt:

The response will be a list of many user stories. Not all might be exactly you were looking at, and the AI might have over looked other as well. But overall, this should cover most of the tickets you need to create.

Write individual user stories

Then, per user story, we use the following prompt to generate individual tickets. Don't forget to re-upload all relevant designs, so the AI knows the context. 

Now, the AI will write the user stories. In our experience, the AI writes about 90% of the user story correctly. You will still need to make some adjustments to finalize the ticket.

Testing is usually performed by either the Product Owner, or a dedicated Quality Assurance (QA) role. It is important to know whether the features delivered by the developers are actually working.

Testing can be quite a complicated activity. Luckily, AI agents can help out.

What are AI agents?

AI agents are autonomous software systems that use AI to perform tasks. There are many companies offering AI agents, but today, we will be using ChatGPT's agent mode.

How you can use AI agents to test tickets

In our experience, agents work quite well to test features, but it takes quite a bit of time. If a developer finished a small feature, you will probably be better off testing it yourself. However, agents excel when you want to test multiple features at once, or do a final testing round before going live.

We provide the agent with a task to perform some end to end test and report the results. The prompt we used is:

Now, the agent is kicking off, and you can follow live what the agent is doing: it opens the website, clicks around and notes down its findings.

The agent is performing work for minutes at a time. 

Other use cases

There are plenty other ways you can use ChatGPT Agent to test tickets.

Onboarding / form testing

Many companies have onboarding forms. The longer these forms, the more people will stop filling the form in which leads to fewer people signing up on your website. ChatGPT agent can test your lead form and find potential bugs, unclarities and improvements.

Behind a login

For many features, customers need to login in first. Luckily, ChatGPT Agent can also handle these scenarios. You can provide ChatGPT Agent with the credentials to login. When encountering the login screen, it will use these details to login and continue the test. It is important to only provide your password to a dummy account, as OpenAI will know the password. It is good practice to change your password afterwards.