How ChatGPT AI can help product managers day to day

In the last post, I talked about what product managers need to know about AI for 2023. We covered how AI works, things it's good at, bad at, and some high-level areas of opportunity. As promised, in this post I'll dig a bit deeper into AI (ChatGPT specifically) and see how this incredible technology can help product managers in their day to day duties. TLDR: It can help a lot to scaffold out the main ideas and then edit to refinement.

We'll see how well ChatGPT can handle:

  • Writing a PRD for a feature
  • Scaffolding out a plan of work for multiple teams, with inter-team dependencies.
  • PR announcements
  • Writing an email to a client asking for feedback

Using AI to generate a Product Requirements Document

For the purposes of this test, we'll assume a feature for a restaurant app where users can refer their friends to a restaurant and get rewarded. I started with a pretty simple one-liner - to "Create a PRD for a feature regarding referring friends to a restaurant app". It generated the below, which, is honestly a bit mind-boggling.

Adding screenshot here for proof as I'm not sure I'd believe it otherwise

Here's what it came back with, on the first try:

Overview:
This feature will allow users of the restaurant app to refer their friends to the app and earn rewards for successful referrals. The referred friends will also receive a reward for signing up and making their first purchase on the app.

Objective:
The main objective of this feature is to increase user acquisition and retention for the restaurant app by incentivizing current users to refer their friends and family. This will also help to increase customer loyalty and overall engagement with the app.

User Value Proposition:

  • Current users can earn rewards for referring their friends to the app, such as discounts on their next purchase or free items.
  • Referred friends will receive a reward for signing up and making their first purchase, encouraging them to try out the app.
  • Referring friends to the app is a simple and easy process, requiring only a few clicks.

User Experience:

  • The referral feature will be accessible from the main menu of the app.
  • Users can enter the email or phone number of the person they want to refer and a referral link will be sent to them.
  • Users can track their referral progress and see their rewards in a referral dashboard within the app.
  • Referred friends will receive an email or text message with a referral link and instructions on how to sign up and redeem their reward.

Functional Requirements:

  • The referral feature will be available to all users of the app who have made at least one purchase.
  • Users can refer as many friends as they want, but will only receive rewards for the first successful referral from each friend.
  • Referred friends must sign up for the app and make their first purchase within a certain time period (e.g. 30 days) to be eligible for the reward.
  • The rewards for both the referrer and referred friend will be automatically credited to their account upon completion of the referral process.

Implementation Considerations:

  • The referral feature will need to be integrated with the app's existing authentication and payment systems.
  • The referral process and rewards will need to be clearly communicated to users through in-app messaging and notifications.
  • The referral feature will need to be tested to ensure it is functioning correctly and providing a smooth user experience.

While not fully comprehensive, that is a much better starting point than a blank page. Having this as a starting point shortcuts at least some of the brainpower that would otherwise be required.

Using AI to generate a project plan spanning multiple teams with dependencies

The first try, I didn't specify that platform is a blocking dependency, but this is how fast it generates your plans for you

So now that we have our PRD, let's ask it to generate a project plan. The below is a 2nd attempt after I clarified that there are multiple teams, and the platform team is the blocker for the other consumer teams to start work.

I asked: "ok, now provide a timeline, in table format, if we assume that the API team needs to do all of its work first, before other teams can start. Also, separate the design and implementation tasks into concurrent bodies of work. For the implementation part, break it down into implementing referral button on checkout, referral area in user's profile section, and then finally push notifications to alert users of relevant steps in the referral process, like a friend signed up, account credit updated, etc."

It responded with:

Here is a revised timeline for the referral feature development, with the assumption that the platform API team needs to complete all of their work before the other teams can start:

WeekiOS TeamAndroid TeamWeb TeamPlatform API Team
1Integrate referral feature with authentication and payment systems
2Test and debug referral feature integration
3Design referral feature UIDesign referral feature UIDesign referral feature UI
4Implement referral button on checkoutImplement referral button on checkoutImplement referral button on checkout
5Implement referral area in user's profile sectionImplement referral area in user's profile sectionImplement referral area in user's profile section
6Implement push notifications for referral processImplement push notifications for referral processImplement push notifications for referral process
7Test and debug referral featureTest and debug referral featureTest and debug referral feature
8Release referral feature to beta testersRelease referral feature to beta testersRelease referral feature to beta testers
9Collect and address feedback from beta testersCollect and address feedback from beta testersCollect and address feedback from beta testers
10Finalize and release referral featureFinalize and release referral featureFinalize and release referral feature

This timeline allows the platform API team to complete their work before the other teams start on the design and implementation of the referral feature. The implementation tasks have also been broken down into concurrent bodies of work, with the referral button on checkout and referral area in the user's profile section being completed in parallel, followed by the implementation of push notifications. The timeline can be adjusted based on the specific needs and resources of the development teams.


A bit like our PRD before. This is not perfect, but is a great starting point. The API team may need more than 1 week to implement the feature, but aside from moving the times around, it's a logical flow of work for multiple teams. Not bad.

Using AI to generate a press release for a feature

Now that we have our PRD and a project plan, what about the marketing side of things. I asked it to "Write a press release for the feature, assuming the company name is Deluxe Apps"

It responded with:

Deluxe Apps Introduces Referral Feature to Restaurant App

[City, Country], [Date] – Deluxe Apps is excited to announce the launch of a new referral feature on its popular restaurant app. This feature allows users to refer their friends to the app and earn rewards for successful referrals, while referred friends will also receive a reward for signing up and making their first purchase.

"We are always looking for ways to increase user acquisition and retention for our app," said [Name], CEO of Deluxe Apps. "By incentivizing current users to refer their friends and family, we hope to not only attract new users but also increase customer loyalty and overall engagement with the app."

The referral feature is easy to use and requires only a few clicks. Users can enter the email or phone number of the person they want to refer and a referral link will be sent to them. Users can track their referral progress and see their rewards in a referral dashboard within the app. Referred friends will receive an email or text message with a referral link and instructions on how to sign up and redeem their reward.

"We are confident that this new feature will be a hit with our users and help us to continue growing our community," said [Name], Director of Marketing at Deluxe Apps.

The referral feature is available to all users of the app who have made at least one purchase, and rewards for both the referrer and referred friend will be automatically credited to their account upon completion of the referral process.

For more information about the referral feature, visit the Deluxe Apps website or download the app on the App Store or Google Play.


Aside from the semantic error that perhaps "purchase" should say "order", this is pretty damn good. Like the others, this is a great starting point which a PM could simply refine to perfection, and would be much quicker than starting from scratch.

Using AI to write an email to a client

Some people struggle with communication, so can AI help? I wanted to see if it could generate an email that handled an awkward customer who recently had an issue with our system, and we're trying to make them feel special by giving them early access to this feature.

I asked: "Write an email to a client asking them if they'd like early access to this feature. The contact's name is Jane and her business's name is Bob's Salads. Our company's name is Deluxe apps. Jane complained recently due to a bug so is not our biggest fan, but the email should make her feel special, that she's getting early access to the feature before her competitors and that we especially value her opinion and feedback. To enable the feature she'll need to log in to our web portal and select it from experimental features."

This is what it came back with:

Hi Jane,

I hope this email finds you well. We at Deluxe Apps value your business and want to apologize for any inconvenience caused by the recent bug. As a gesture of goodwill, we would like to offer you early access to a new feature we are launching on our restaurant app.

We are confident that this feature, which allows users to refer their friends to the app and earn rewards for successful referrals, will be a hit with our users. We value your opinion and would love to get your feedback on this feature before we release it to the general public.

To enable the feature, simply log in to our web portal and select it from the experimental features section. We believe this feature will give you a competitive edge and we are excited to see how it performs for your business.

Thank you for your continued support and we look forward to hearing your thoughts on this new feature.

Best regards, [Your Name] Deluxe Apps


That is impressive. I might edit one or two bits to make it a little more personal and less corporate, but again as a starting point, this is much better than a blank page. The email length, structure and content is almost perfect in my opinion.

In conclusion

A recurring theme is that the AI can't get it perfect, at least not yet, likely because it does not understand the context of your company and product. But what if it did? What if this was a personal, company-wide assistant that could help everyone plan features, products, marketing and client communications, to name but a few use cases.

What do you think? Will you use this in your daily work? If you do, just be careful you're not sharing any proprietary company information with third parties.

Stay tuned for more AI updates, specifically:

  • Understanding the GDPR implications of using AI with your user's data
  • Answering the question "Do I have to bring the AI in-house or are there APIs I can use?"

PS: don't forget to share this with your team, colleagues or friends!

Chris

You've successfully subscribed to Chris Dermody
Great! Next, complete checkout to get full access to all premium content.
Error! Could not sign up. invalid link.
Welcome back! You've successfully signed in.
Error! Could not sign in. Please try again.
Success! Your account is fully activated, you now have access to all content.
Error! Stripe checkout failed.
Success! Your billing info is updated.
Error! Billing info update failed.