What would you like AI help with most for your product management jobs?

Janna Bastow
11 replies
We’ve just launched our AI tools to help product managers automate the grunt work and win back time to spend on the important stuff like discovery and talking to customers. As part of our own discovery when planning these features, we asked thousands of product folks what it was that they’d most like to get AI help with. The two main winners were ‘Getting off the blank page’ when it came to writing tasks with 23% of people selecting this, and ‘summarizing long documents’ with a whooping 64%. 

What would your answer be? What job do you really wish took less of your time?

Replies

Simon Cast
Tedious tasks like writing user stories.
Roger Swannell
I get that there's benefits in PM using AI copilots for things like writing user stories, but I think the real value is upstream in analysing large datasets to spot patterns humans don't see. To me, that's where AI has something unique to offer.
Simon Cast
@roger_swannell yep helping process large volumes of feedback is key to help spot emerging trends. ProdPad has a beta feature called signals that is doing this.
Russell Ashby
Sharing some thoughts. We shouldn't only think about minimizing the time it takes to do jobs - this is the go-to use case that everyone is jumping on and probably won't end up being that much of a differentiator. (and hey, I can just go straight to ChatGPT for my writing tasks. Prompts for writing likely won't turn out to be valuable IP). There are other outcomes like, "reduce the risk of", or "increase the likelihood of". For example, we could be thinking about how Product Management tools using AI can increase the likelihood of sign-off on a big investment. Or, could a Product Management tool using AI help to reduce the risk of being caught out by a new compliance regulation coming down the pipe that we weren't aware of? Or, could a Product Management tool using AI help reduce the risk of bad prioritisation through flawed thinking/data? (stop me from making mistakes)
Mohamed Aboshihata
@russell_ashby I've built multiple agents to simulate impact of decisions, generate options, help with walking backwards on strategy...etc. An experienced PM with custom Ai agents will help will improve decision making. Ideally, You would have a mix of PM's on your staff, peers, and agents to help navigate the next era.
Rotimi Ojomo
Using AI to analyze data would be top of my list. Improved data analytics ,I think that will help build better products and product features.
Justin Woods
For me I'd like: * User request / feedback sentiment analysis * Identifying common trends in data and text * Automatic executive summaries * Ability to translate a single concept / body of text into meaningful versions for different audiences * Summarise large bodies of text * Expand / elaborate small bodies of text * Suggesting a framework / technique to help create a document / body of text (Not like Clippy from Microsoft though!) * Creating messaging (at least a first draft)
Shane Williamson
Anything that frees up my time to spend more on customer research, interviews, discovery etc. The big impact for me currently is assisting with building training & user manual materials.
Mars Escobin
Reaching out and scheduling calls with users - would love if AI can automate that 🤩 Workflow in mind: - We get a sign up (or any other trigger) - Email goes out with a custom message specific to user's profile and what they were trying to do in the app
David Coyle
Early 1 pager drafts and PRDs Survey review and writing Interrogation of documents based on personas Persona drafts Potential partner overviews Etc
Adrian Parker
Market sizing work. ProdPad, what is the TAM for (product idea). in (geography), serving (segment)? What data points were most relevant to your calculation? What marketing methods would reach the people in that segment? Show your working. It's always wild stab in the dark back of a napkin stuff when the team and I try to answer those sorts of questions, and it feels like an AI could help reach data points we don't have the time to find.