Using ChatGPT to complete Product Management tasks

Akhilesh Goyal
9 min readJan 24, 2023

As Product managers, we are constantly gathering information from various sources to make decisions. So far, Google has been the way to go, but here comes the ChatGPT, its conversational interface makes it significantly efficient, and the answers it has produced to so many random and abstract questions are mind-boggling.

You must have seen ChatGPT writing Poems, letters and answers to random questions, but can it be useful with serious real-life business-related questions? Can a Product manager rely on it to make key decisions for a product?

Let's find out.

What to expect from this Blog

Instead of just trying random queries on ChatGPT and posting results, I tried to use it as a tool for different tasks which Product managers do. In this blog, I am sharing the results of this activity and my take on it. I am sure I have not utilized the full potential of ChatGPT yet, but feel free to take these examples as inspiration and use them with your imagination.

Task 1: Extensive User research

User research is always done in two ways, Quantitative and Qualitative research. Assuming ChatGPT has learned from multiple sources, we can easily use it for Quantitative research.

In the below example, I was trying to understand the Data Engineer persona which is one of the Primary personas in the Product I am currently working on. I have a basic understanding of what Data Engineers do, I started with a very simple question:

What I liked here is that ChatGPT summarized the answer quite nicely, I could find this with a simple google search, but I might have to read at least 2 or 3 articles to have an overall understanding of this.

Then, I dig a little deep and to understand the challenges faced by this persona

and even further deep:

After understanding the challenges, I tried to understand how Data engineers are solving them

I must say the answer was complex and not specific, so it could be confusing for some who are learning about it first-hand.

Next, I tried to gain more contextual information; Data engineers in Retail Industry

Here, ChatGPT did an exceptional job of answering the question within the context of the Retail industry. In fact, I searched these keywords on google and I could not find anything on the first 3 pages.

Finally, I asked ChatGPT to generate the persona of a Data Engineer

Task Summary: Overall as a Product Manager, you can develop a basic understanding of the target user persona with ChatGPT very efficiently, further, you can validate it while doing Qualitative research i.e. actual user interviews.

Task 2: Roadmap planning:

On a high level, it seems too much for ChatGPT to develop an actionable Roadmap for the product, especially with all the different contexts we need to take into consideration for the proper plan. But I think it can help us create a skeleton of the plan which we can use and fill in the details and prioritise stuff based on our understanding.

To build a Roadmap for the product, first, we should try to understand various themes related to the Product. This will help us with high-level Prioritisation.

The themes suggested by it might require further division or clubbing based on the Company’s focus and capacity, but this could be treated as a comprehensive list of themes from which we can take inspiration.

Once we have identified the themes, we need to prioritize based on the different factors, I asked ChatGPT what it thinks about it:

We could ask ChatGPT to simply give us prioritization but this would be too much of a risk given that ChatGPT will not have the full context of the Product and target customers. I would suggest we should do this ourselves. Anyway, I gave it a try

Even ChatGPT is saying prioritization will depend on the specific needs of the platform and its users.

Once we have themes, we would go deep down on each theme and try to figure out on a high level what different initiatives we can work on these themes:

Again, I and ChatGPT would recommend prioritizing these initiatives based on the specific needs.

Further, we can ask about different features which should be built into a particular Initiative

In a nutshell, we get a comprehensive list of features which we can use as inspiration to identify features for the initiative

Task Summary: Defining a roadmap entails taking many critical decisions which will impact the future of the Product. You cannot simply ask ChatGPT to make these decisions for you. What you can do instead, is take inspiration from it. You can get a comprehensive list of the features, initiatives and themes and feed it into the road map after careful consideration and taking feedback from customers.

Task 3: Creating Backlog: User stories and Acceptance criteria

Creating a backlog is all about capturing all the necessary details for the development team. This is one of the most important outcomes coming from product managers. Every activity we product managers do such as user research, vision setting, roadmap planning, customer interviews etc. feeds into the Backlog.

Let us see how ChatGPT can help us in this regards

Different companies follow different processes for creating the backlog. In this example, for the sake of simplicity, I start by listing all the user stories needed for the feature to be developed.

In reality, PM discusses the feature in detail with the Engineering and design team and breaks user stories based on their inputs. This output from ChatGPT can not be simply converted into the backlog. But this could be a good place to understand all the different use cases which should be supported by the feature.

Once we have the list of all the user stories, we can use ChatGPT to create a first draft of a description of these user stories

In this example, I used the one suggested by ChatGPT, but you can write your own and ask ChatGPT to write the description. In real-life scenarios, you would add many more details, mocks and usecases based on the product. But you can use this as a template to create user stories.

You can also ask ChatGPT to write acceptance criteria for the user story

Task Summary: We cannot simply use ChatGPT to automate this, we will need to do lots of context setting, but it can make this process quite efficient. We can use ChatGPT to Create a template/structure for the Backlog, and thus you will not have to start from scratch. This could save a significant amount of time for Product managers.

Task 4: Product Marketing

A product marketing strategy is all about positioning and promoting your product, both before and after launch. Messaging in these campaigns sometimes becomes tricky and it can take a significant amount of time to perfect it. I sincerely think with ChatGPT we can have several initial ideas, which can help us speed up the full process.

So let's see how ChatGPT help us with this

I started with identifying the different target persons for the marketing campaign

Chat GPT did a great job of identifying the different groups. Although the groups made are not mutually exclusive, it should not be a problem for us. Let's see how it does with the messaging for each group

This is too generic for my liking, but can it differentiate while in action, i.e. designing the messaging for each of these groups

I think for a marketing campaign, we might have to be a little more creative than this. Although all of these messages make sense, you will find these lines in the marketing campaign of every other AI company.

Let's say we target one user group, and we don't need so much creativity, then:

I like how it assumed the different necessary features in the Platform without me prompting it. I am sure, if I would have provided more context about the platform, this message would be more tailored.

In the next step, I specifically provided a little more context and ask it to be creative :

This is great, I sincerely think we can use this to create custom messaging and send tailored messages to thousands of customers from hundreds of different industries.

Finally, I asked ChatGPT to design the full product marketing campaign:

Overall, It summarized different components very well. We can use this as a checklist to make sure we are reaching target customers in every manner possible.

Task Summary: ChatGPT can be a great tool for designing product marketing campaigns by taking inspiration from it to develop messaging. It can be a game-changer in scaling the campaign to become tailored for each target user. But, to be distinctive amongst competitors, you have to be a little bit more creative than this.

My take on using ChatGPT

ChatGPT is sincerely producing some great results which can be directly incorporated into day-to-day tasks. It could be a great source of information for everyone and will significantly decrease the time needed to research various things.
From a technology point of view it is great, but in terms of context setting it is still far from what an actual human can do. So I don’t think it will replace the Job of Product manager anytime soon. But it can make the life of product managers much easier. As Product managers, we should embrace it and learn how to use it.

Some tips which can help you use ChatGPT

  1. Context setting: ChatGPT has all the answers given that it has the right context; Whatever you need to ask ChatGPT make sure you set all the necessary context for it.
  2. Breakdown the Problem: If you are having difficulties framing what exactly you want to ask, break down the problem for it. You can follow the same structure which you will use with a human who doesn't have the context of the problem
  3. Take it with a Grain of Salt: The latest model of ChatGPT is trained on data collected before June 2021, so it might not consider all the new information. Also, we don't know all the sources utilized, especially for niche topics

Let me know what you think about this in the comments section

--

--