Reflections on startup life: Week 76 (Y Combinator Interview)

First part

============

I was expecting more intense heated conversation with lots of cut-off. But it wasn't. Besides PG and Trevor, the others were pretty relaxed. 80% questions coming from PG. I had the bipolar feeling in the middle of the interview that either they've decided we're in or we've already screwed the whole thing up.

We were expecting they asked about how we differ from pinboard.in or twitter or any other competitors we're having in mind since that was what came up in a prep phone call. But instead one question that has been brought up was how is trunk.ly different from delicious?

Looking back, I think we could do a better job delivering why we're better than delicious. I got side-tracked explaining why implicit was important and when I tried to explain why search can do a better job than tagging. Why remembering which tag you were using was such a big headache and content discovery by related tags didn't really work, ... I was cut off without being able to finish.

It's quite useful to see that how these group of people think of delicious. Delicious has done such a fantastic job associating itself with the problem "find my links" that the "tagging" concept has become the "only" solution to the problem. I believe we're a better solution for the same problem, but a different one. We shouldn't spend time talking about how "implicit" we want to be, not that it's not important, but that the key message should be "tagging is dying. search will replace it. " It was a bit sad to hear the "how you're different from delicious" from Trevor at the end of it, meaning we haven't "get the idea through".

A couple of highlights are 1) the growth chart; 2) user testimonial; 3) demo using YC alumni for topics the group are interested in.

Tim has done a fantastic job charting out the growth chart, monthly growth which is showing a clear increase and the retention rate, which is pretty stable at 30%. Those three pieces of papers were passed and looked at by everyone in the meeting. PG even stood up, holding the paper up. Tim said there was someone videoing the whole thing but I didn't notice that.

We were asked what our users thought of us. Tim handed out the print-out with twitter messages about trunk.ly on it. PG took that paper and had a quick read, which was good. While Tim was talking, I notice Robert was looking at the other page of testimonials. Not sure whether he wanted to read it or he was just bored. We also talked about 3% of our users uses the API and roughly 10 developers are writing software using our API. Tim saw some spark in someone's eyes by then.

At roughly 3 minutes into the interview. PG raised the question, "what does social mean". I can feel the word became too fuzzy and abstract for him. So I proposed we gave a demo of the site. Tim has created a YCDemo account on trunk.ly, following all YC alumnis so that any search using "from:friends" are going to show interesting links the alumni were talking about. We showed anybots, smtp and founders at work. I wasn't quite sure how it goes by watching the facial expressions. However, the question about "what do you mean by social" never came up again. I feel really grateful that Tim spent so much time preparing a demo. It was just hard to describe something in word.

Thinking back, now I can see that they were trying to figure out what's special about your service? PG and Harj have already heard about Rand Fishkin from SEOMoz is a big evanglist. So at one stage PG raised the question why Rand likes this so much. Tim's question was great: Rand is the link guy and he has many moments when he needs to recall a link. So I think it's a sure thing that we definitely describe the need right. Both PG and Trevor said they have the same needs.

At the very end of it, PG raised the question "what's your business model? why can't you guys just like pinboard.in, charge for usage?". I explained the benefit of why having network effect is crucial if we want to move on with the "social search" direction: "10M users will surely make the search result a lot different/better. " But PG's answer was: "Once you got 10M users, you don't need to worry about business model anymore.

Second part

===============

Around 7:15, Tim and I came back from supermarket. There was one email from PG waiting for us. Hope it's ok for me to paste the email here.

<quote> I'm sorry to say we decided not to fund you guys. It was a difficult choice because you seem like smart guys and we were impressed with the progress you'd made so far. Ultimately what deterred us was that we just couldn't figure out a way to make money from it-- at least on the scale a startup requires. </quote>

We have high respect to YC people. So my first reaction when we saw this email was "Let's focus on getting some basic numbers worked out. It's at least a good way to remind us besides user acquisition, we want to get the business model sorted out. " Having said that, I am still not sure on whether the decision was "made" on what understanding of the service. Would it be because delicious has no business model and we failed to figure one out? Would the result be different if we get the message across? Or maybe the message did get delivered?

That's it. Sad we didn't get in but it's just another bump we need to get over.

I have no vested interest at all. I'm not an investor, not an advisor, just a user and a fan. Trunkly's become an essential tool for me, it has traction, it has a product that 100s of thousands would find useful today and millions could find even more useful in time.

Maybe I'm blinded by my own selection bias (because I find it so useful and everyone who joins makes it more and more useful for me), but I'm going to have some very high expectations for the startups that do get selected :-)

I have a take on this. You, "Thurston" have just been stood up by a selective "match-making" agency. Have you blown your shot at getting access to the hot date? Maybe, maybe not. The agency is very picky. They have a captive market and a select clientèle. There is also a big mis-match between information the agency has over the candidate. The agency knows things you "Thurston" and other candidates do not - the intimate needs of the client, the market and other candidates.

The agency look at you and see something promising, but they have a suspicion - no a big question mark - is Thurston better in a right kind of way compared to a previous hot candidate who we'll call "Dave". Dave was made of the right stuff and quickly got snatched up but never really clicked. Something wasn't right. Dave was liked by the right people, lots of people and made all the right moves but something was lacking? The client liked Dave but Dave never pulled in the big bucks and now the middle of a messy divorce. Everyone lost, except perhaps Dave :)

So when the agency look at you, Thurston they do so through the eyes of the client and ask hard questions: "what money can you make that Dave couldn't?". There's one other thing the agency see, 76 weeks. They ask themselves, "Thurstons been on the market for 76 weeks & he hasn't yet how to capitalise! This is now our clients major concern. Can the candidate make the big bucks?"

Access denied!

Sad but that is one possible way to view the outcome. There are plenty of hot dates out there so get out there and find a way!

gigaom has this nice article: Can Delicious Solve Our Information Discovery Problem? http://gigaom.com/2011/04/27/can-delicious-solve-our-informa...

I think trunk.ly is right there tackling this problem. Had this article came out earlier and pg reads about it, the outcome would be quite different.

Something like a hookin to hootsuite? If you don't know it - its worth looking at. It integrates FB, twitter etc for IT Service Management.

I'm trying to think why a corporate would be eager to give you guys money for the value you add - and I hear knowledge management is a pain, so could you be part of that solution?

[just tossing ill-formed ideas around in my head hoping to help]