Aftercare helps you run surveys with AI-powered follow-up questions to dig deeper for insights—it’s like having a skilled interviewer for every survey. All your responses are automatically categorized and summarized into key insights using AI, saving you time.
This sounds like an exciting launch—congratulations to you, @justinmay, and @anandnanduri ! Aftercare’s focus on blending the scale of surveys with the nuance of intelligent follow-ups feels like a game-changer for collecting meaningful insights. I especially love how it bridges the gap between raw data and actionable insights, which is a massive pain point for many teams.
This feels like a tool built for today’s AI-driven insights world. Kudos to you and your team! Wishing you tons of upvotes and interest on Product Hunt! 🚀
Report
Great job guys! How does the AI determine which responses need follow-up vs. which are already comprehensive enough?
@ar_12 Thanks! We evaluate if the answer to the question is 1) relevant enough to the context provided in the follow-up feature and 2) whether or not it answered the question completely enough. We've set up LLM generated scores and thresholds to evaluate this. If the answer is not relevant or the question is not answered completely, we ask another follow-up.
Report
Great product - congrats on the launch, Aftercare team! Can't wait to see where you take it. Woohoo!
Saw you guys recently added a 'create with AI' option to build a survey. What was the inspiration behind that new feature? I like that there are multiple ways to get started.
@davidhu Before we had the "create with AI" option, we had reached out to some early users who started building surveys but never finished. They had expected the survey to be built for them automatically after entering in a description. We figured this feature would help with activation and time-to-value, so we built it 🙂
Hey @aidanlee@justinmay@anandnanduri love what you've built! Can we customize how the data is categorized? Sometimes our team needs very specific groupings.
Thanks @hhendrix! Not at the moment -- right now, open-ended data categorization happens based on the topic question that you ask and GPT-4o will determine what categories should be created based on the question. This is something we're considering adding soon.
@aidanlee Congratulations on launching Aftercare! The concept is fantastic, it’s creating an entirely new type of research, and I’ve already started using it. The potential here is immense!
After quick product exploring, I’ve put together a few suggestions that might help enhance the experience:
1. Clarity on AI Followups
It’s a bit hard to predict how AI followups will behave, which can make the experience less intuitive:
- The segmented control for switching between editor AI modes feels like selecting a specific regime. Placing "AI Followups" directly under the "Open Ended Editor" could make the connection clearer. It might seem a bit visually crowded, but with the right design, it can still look clean.
- The AI indication in the block diagram is subtle and not immediately clear. Showing follow-up questions as a separate block in the diagram would make it much easier to understand.
2. Improving Question Formulation
Writing great questions is an art, and there’s room for the system to assist more here:
- Questions could often be much shorter. If context is essential, splitting it into a short main question and a subsidiary explanatory block could keep things concise but clear.
- The current questions are very straightforward, they mirror what the user initially wrote during survey creation. Reformulating them to prompt deeper, more reflective answers could improve the quality of the responses right from the start.
3. Interface Usability
There are a few small interface issues that could be addressed to make things smoother:
- Always show block names, because it’s tough to set up conditions without them.
- An undo button for the block view would be quite helpful.
- The progress bar finishes faster than the survey itself, which can be confusing for respondents.
- Errors during survey creation don’t block preview functionality, but a preview attempt triggers another error.
- The “Multiple Choice” question type is misleading and should be renamed “Single Choice” to avoid confusion.
Despite all the comments, I’m truly impressed with what you’ve built so far and excited to see how Aftercare evolves. Great work, and best of luck!
@anton_osipov Wow, this is incredible feedback, and thank you so much for taking the time to use it and write up these improvements!
Some of these changes can be implemented fairly quickly (especially for AI follow-up clarity and usability) and should have significant impact to the user experience.
We'll get to work on these. Would love to keep you updated on your suggested changes and will be in touch!!
Congrats on shipping! 🚀 Love the auto-categorization feature and the branding is very cute. What are my options for exporting the data? Do you guys integrate with any other tools?
Thanks @raymond_huang5! We have a csv download if any data needs to be exported. We don't have any integrations yet, but plan on building some in the future
Milestone Content Studio
Aftercare
Aftercare
Magic Hour
Aftercare
Respan
Aftercare
TRDATA
Aftercare
TRDATA
Respan
Aftercare