Getting feedback - on Product Hunt and elsewhere!

Nik Hazell
11 replies
Product Hunt seems pretty unique in terms of getting feedback from the community. We launched https://www.producthunt.com/posts/zappi-ad-predictor this morning, and we've already got some actionable feedback, but of course we want more! 🙏🏻 Specifically, I'm super interested in hearing whether you care more about "Image/static" marketing ads, than video ads - we need to know where to train our model next! But, where next for feedback? Does anyone have any suggestions, or is this the pinnacle of community feedback? 😂 TIA!

Replies

Eugene Hauptmann
Hey @nik_hazell great product, just upvoted! I'm no expert in the ads market, but I'd love to see A/B testing comparison between several different ads. And yes, +1 for images. You can deal with the video later :)
Nik Hazell
Zappi Ad Predictor
Zappi Ad Predictor
@eugenehp thanks! Weirdly - we actually built the video predictor first! Our dataset of video ads was already really well tagged, so it was straightforward to build a good model. I think we'll get to work tagging and cleaning our image ad database! Thanks for your feedback, Nik
Tarek Dajani
100% include images, to be honest I didn't use the first time because it had no images and then used a ready made video from canva that I made in 2 minutes just to check it out. Seems really nice by the way! Congratulations.
Nik Hazell
Zappi Ad Predictor
Zappi Ad Predictor
@tdajani Thanks Tarek! Appreciate the feedback.
Jeremy Bradley
Encrypted Image Filtering App
Encrypted Image Filtering App
A little late to the convo, but I concur with the others that image ads should be the priority, if only because the vast majority of ads that companies serve are static. Video ads are awesome too, though, and you've done a great job especially with regards to analytics.
Felicitas Hilge
Oh yeah, you should include image ads as well ;)
Eugene Hauptmann
@felicitashilge @nik_hazell I'd love to read more about your engineering endeavors once you're done with your ML fun!
Nik Hazell
Zappi Ad Predictor
Zappi Ad Predictor
@felicitashilge Yesss we should get to work! Lots of tagging to be done - but more importantly, cleaning the data once it's tagged!
Nik Hazell
Zappi Ad Predictor
Zappi Ad Predictor
@felicitashilge @eugenehp Brilliant - I'll share them when (/if!) they're successful!
Dawn Veltri
First, I definitely agree with static images. That was my first thought this morning as most companies have a higher number of static ads than video ads because of production costs. To answer the feedback question I would say to start hustling. Reach out to marketers individually on here, Twitter, LinkedIn, and other social media platforms and ask them to quickly give it a try. The nice thing is it's free and quick, so it shouldn't be a hard sell. If you see a video ad online reach out to the company and ask nicely if they will test your product and provide feedback. Put a quick quiz on your site with the scorecard asking people for quick immediate feedback. Make it easy to provide feedback and people will.
Daria Novikova
@dawn_veltri1 great insights! I agree about social media use - that really works, but only when your messages are personalized