
Hypersonal
Generative AI for shopify product pages
It's been a few months my app has been on the app store, but I haven't done much anything for it. I think the main thing blocking me from investing more in it is that I'm not sure I understand the pain point I'm solving. It feels convoluted. This feels like a tool looking for a problem, and personally I get demotivated when I don't feel like I'm helping people, and the flip side intuition is that I don't think this would sell easily if it was just a tool looking for a problem.
I'm at a crossroads because I only want to work on one project and it will be Hypersonal or Jopro. I'm going to become much more time constrained soon. I'm more intrinsically motivated with Jopro, I find the space more interesting personally and have found it easier to develop Jopro than Hypersonal. However Hypersonal has better economic potential in my opinion, and a clearer distribution strategy. I just don't think Hypersonal is valuable, if I can excited about Hypersonal actually making an impact then I think Hypersonal would make sense to choose.
I think the best thing I can do is talk to potential customers to figure out if Hypersonal could solve any of their potential top of mind issues.
After a few months of developing and testing, I think Hypersonal is ready to see the light of day! It's nowhere near where I want it to be, but I think that's okay. I have no idea what features merchants will actually like, if any. I am worried that putting it out now will result in negative feedback, but I have to see what people want. If they end up wanting something else entirely I can always pivot it under a new app itself.
1 Like
1 Comment
1 Comment
-
2
Hi Ameer - best of luck with your lauch. Regarding your concern, this is a big part of putting something out there - the only way you get feedback, i.e. the ingredient that makes your product better is to put it out in the market.
Enjoy the journey and take pride in your achievements!!
If you want someone to test it pre-launch, I have access to Test Pilot and I can give it a look so please feel free to PM me.
I'll have another update soon as I actually launch but an interesting realization today is that in writing my getting started docs is that things are really coming together. It feels very real and there is actually a lot of cool stuff there for people to use.
I've given up on the soft launch approach, after about a month without traction. It's unclear if the product is not interesting or if people don't want to sign up to get notified about a shopify app launch. But the clear path now is to launch to the app store. I've been focused on making things more polished for the merchant experience over the last few days, cleaning up setting options, dashboard, etc. and should be able to apply for listing in the next few days.
1 Like
Comment
I finished coding product chat today. It's a great feature that lets you ask questions about the product listing and also has context (as in a hypersonal context, e.g. your search term, or leading ad, or referring page, etc.) aware suggested questions. Once I finish with the product recommendation front end I can make a demo video! This will round out my landing page, which isn't getting much response as is today.
The problem with my product and landing page today is that the problem is not clear, not singular, and not understood as a "problem". Finding PMF is proving tricky. Something I do need to do is try out ~5 different landing pages and figure out what things resonate.
I spent about a month foregoing this problem, and instead tried to demonstrate Hypersonal's value through a series of contrived experiments and surveys which ended with mixed results. The blind comparison surveys always showed favor to what I was building. The real case ended up being so contrived that the value of the product was reduced to color matching the product from an ad I ran for a product that didn't sell because it was dropshipped and poorly differentiated. And in that case somehow had a negative impact for my small sample size. 🤷♂️ I ended up getting embroiled in trying to build out this secondary brand and product. Oh well, back to reality, time to stop procrastinating and let people tell me it's crap instead of doing everything I can to avoid hearing it from someone else.
Once I have a demo video I will look for other distribution channels for getting initial customers (in the worst case just putting it out on the store as a listed app). Managing the initial customer reviews is critical for things that thrive or die on a marketplace. I promise it's not an excuse for not putting it out there, I'm very consciously trying to launch with feedback first so that reviews that hit the public store profile are good. I might try talking to people who make a lot of apps and see if that's just a way I'm tricking myself. Anyway, my goal is to get it into the hands of a real customer as soon as possible to start getting feedback. I will try to ask for the help of others where I can.
I will also move on to testing 3 more unrelated ideas before the end of the year, in very simple landing page form.
1 Like
Comment
Product development has been going good. I have dynamic editorial text blocks, and variant recommendations or tentatively called option recommendations instead. These are both really cool! I'm working on a v2 of the generation pipeline that's based on some additional best practices and langchain. Already seeing some great improvements in quality, especially for the relevancy filtering subtask.
We should be launching soon and seeing if we can get any response. I am working this week l in building out the v2 product that has all the primary features that may entice the user. I've cut a lot of stuff from the plan though, it's now just what's at the heart of the content to be generated and a table in the admin UI for editing and deleting the content.
1 Like
Comment
About
I think generative AI has a strong position to revolutionize personalization in ecommerce, and elevate it beyond sales stats.


Comment