rachellaw

I think they're all pretty bad tbh. They aren't very good at guessing age, and guessing gender is harder especially if you're young or have rounded features.

thanks! We were wondering if it was worth applying.

We went from zero revenue to generating revenue after our old accelerator and was looking into YC to grow now that we had a solid, validated product-market fit.

How much is the bias against startups that have been to other accelerators? What are the benchmarks for progress in those cases?

One the reasons why it took so long (2 months to approval) is there isn't a clear set of rules for success.

When you submit to Play store or Apple store, there's a list of guidelines and checkbox reasons why you get rejected.

With Slack, it's more like having a conversation about making it better, not necessary a fixed set of rules.

It probably didn't help that we submitted in mid-December when most people were already on holiday.

I like the penguin it's cute, but I can't help but think of Microsoft clippy

Yes, that's what we thought as well when we got rejected. With such a large group of applicants, the core reason is probably "you're on the wrong side of bell curve"

At the same time I wish they were more transparent. Dropping bits and pieces in no particular order like "don't do a previous accelerator" or "get a recommendation from another alum" makes their process seem irrational.

one of the things that really struck me when fundraising:

"...that 25k he just gave you, could've gone to his kids' college fund or investing in blue chip stocks instead. It's amazing that you can raise money at all."

Angels give less money than VCs, some angels give low amounts (15-25k) especially in an early seed round. They're doing it for you. Screwing them over for more equity or YC makes you a shitty irresponsible human being and imho, should be considered a type of fraud.

Honestly I'm surprised at Sam's stance. Unless it's really justifiable, this seems like bad advice.

one of the key reasons why people join accelerators is for network and connections though, not just advice.

Without that conversations, that sounding board, all that written advice just makes another investopedia/venturehacks guidebook. Most startup founders already do the google legwork and required readings. It's great, but not the main reason why people would apply for YC.

It's more of a natural evolution. We had one idea on how to use the technology, but the user feedback preferred a smaller aspect that we overlooked. So we narrowed down and turned that one small aspect into the company focus instead. The technology that powers it all, is still the same

I guess you can call it baggage, but nothing works right out of the box. Youtube was a dating site before it became a video site, but they didn't start fresh/reincorporate. They just grew from what their users wanted.

We've already taken investment pre-pivot from angels, screwing investors over to be a 'fresh startup' seems to be a type of fraud.

It's different pitch, but a natural and logical progression. Most of the time ideas need refining, especially with user feedback. It doesn't make sense to start with a new company since the technology that powers is still the same.

It seems to unfairly penalize startups who've pivoted since after the accelerator.

We've been to one. It was early stage, and within our geography. When we spoke to YC/500startups in 2013, it was all about having traction or earning revenue and breaking even. We didn't have that. We were academia dropouts with some patents and an idea.

Going through the 1st accelerator helped us because we didn't even know how to put together a pitch deck. We've since pivoted because ideas change, become more focused and polished. Our idea, business model and pitch is completely different.

YC would not have accepted us in 2013, someone else did. Now if we want to apply to YC post-pivot, we get penalized? That seems incredibly unfair.

I think it's in-combo with the whole AI/automated personalization as well. It used to be very difficult to do intelligent bots, but now you have tensorflow/caffe/word2vec frameworks that come built-in with NLP

The walled garden thing has always been an issue though. I mean Facebook threw a fit when India turned down their "facebook-only" internet: http://www.economist.com/news/business-and-finance/21685292-...

What's fascinating is the bot-bandwagon is mirroring the early app market.

With apps, you downloaded things to do things. With bots, you integrate them into things, so they'll do it for you. (Save extra step laziness)

I think you have to be a successful founder i.e. huge revenue or huge user growth or a VC in order to give expert advice and feedback.

For user testing, the most important metric is not how many people give us feedback but whether it's coming from a credible and relevant source.

We were rejected, despite being in the YC RFS' category (A.I.)

Kip is a deep learning search for fashion in IRL stores around you: https://kipsearch.com

I don't think there was a particular reason why they rejected us, most likely that in a bell curve we just weren't as good compared to other applicants.

It was great fun doing the application, and we learned a lot! More importantly, we closed a lead investor/partnership the day before, so even though we were rejected it wasn't a big disappointment.

Thanks! I'm curious to know how Lyst would handle outlier datapoints. ANN seems to be a variation of regression techniques (achieving nearest result) instead of creating a new mutual exclusion (unsupervised learning)