Hacker News new | past | comments | ask | show | jobs | submit login
Reflections on startup life: Week 76 (Y Combinator Interview) (timbull.com)
46 points by zaveri on Apr 27, 2011 | hide | past | web | favorite | 19 comments

I'm really disappointed in YC on this one. I'm sure there were some awesome companies selected, but http://trunk.ly flat out rocks.

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 :-)

You mentioned the service on Twitter and I signed up immediately and added my Twitter account. Within 2 hours of signing up I had used the service no less than 3 times to dig up links I wanted to share with other people. So damn useful.

Actually, I'd be willing to bet the people at YC agree with what you said about how useful and popular it could be... but how do you make money with it?

That seems to be the one point that everything hinges on here.

When I talked with Tim, I had three ideas for monetization. His/Alex's may indeed be even better.

#1 - Sponsorships/promotions on topically-specific "top" pages. Advertisers (like my company) pay top dollar for highly focused branding to specific audiences. If Trunkly created subject-specific pages like "Top Startup Links," "Top Travel Links," "Top Gadgets Links," etc. and earned decent traffic to them (which I think is very possible given how useful/interesting/powerful and SEO+social-sharing-likely those pages could be - think HN/Techmeme/Alltop/etc.) , that's a big advertiser revenue stream.

#2 - Tools for management/analysis/tracking/enhanced data for specific users. I'd personally pay $10-$50 a month and my company would pay $100-$500 a month to have all our links not just tracked and indexed, but show data, signals and analyses of what predicts more clicks/shares/etc. I've always wanted something that could tell me which features (length, use of words, time of day, when I shared vs. others, etc.) predicted higher CTR/engagement/RTs/FB shares/etc. and Trunk.ly's well poised to build that.

#3 - Social graph data ala Twitter + Facebook. Dozens of companies, in particular Bing + Google, are really interested in and paying top dollar for access to social sharing data about URLs. Trunk.ly knows not only what I share on Twitter/FB, but via my personal blog, LinkedIn, Quora, and anything with a feed that I add.

On the surface it seems an easier proposition to make money here than other companies, plenty of data to target ads well through the links being shared and like Twitter you can build in ads that are like regular content so people would be buying sponsored links that appear like the ones being saved.

I guess the issue is, the uptake of the product, and retention rate with the ratio of people actively using frequently to those using it as a link archive service which they don't revisit frequently as their links are being automatically archived.

The delicious aspect would be similar to say someone starting something in video and getting the question of how they are going to be able to make money when it's taken youtube crazy scale and availability of cheap internet bandwidth with that scale to turn a profit. Either you taking the same angle and are somehow going to do it better than Google or your clearly differentiating yourself in a way that will allow you to turn a profit without the same scale.

It seems contradictory to get a rejection email that bemoans the idea - we keep hearing that caliber of the founders is all that matters.

We care a lot more about the founders than the idea, but the idea still matters somewhat. And as we get further along, the differences between groups get smaller, so variation between ideas tends to account for more of the variation between groups. By the time we get to interviews we've narrowed the pool more than 10x, so we're basically choosing between good people with good ideas and good people with bad ideas.

Although I'll (perhaps understandably as this is my post / reflection that's been linked) not concede that our idea is bad, I do accept what Paul is saying here. I don't feel there was any duplicity or otherwise at all in the process - it's very open and straight forward.

From my perspective, of course Alex and I would have liked a different outcome, but I don't envy the decision that Paul and the team have to make either.

There's more than one post which has pointed out they don't always get it right. Only time will tell - YC was never a make or break for us.

Our immediate response (after initial disappointment) was "crap, maybe they're right". We've gone back and crunched the numbers and we still think there is a substantial opportunity here, but there's no doubt our forward planning is already being influenced by this outcome. We'll make sure we test these assumptions a lot sooner than we were initially thinking about.

Who wouldn't want the experience that was in that room to listen to you pitch for 10 minutes, and who wouldn't take some of that feedback on board and at least pause for thought?

Yes, it's an exaggeration to say bad ideas, particularly in your case. I wish the choice were so clear cut. In fact it's more like choosing between fairly good ideas that might grow into something great, and fairly good ideas that seem, on reflection, like they will run into trouble of some type.

It seems like companies that apply with no idea, then, would be at a noticeable disadvantage during interviews. Is this true?

No one applies with no idea now.

Given that the reddits initially had a bad idea, wouldn't this mean that if they applied today with a similarly bad idea, you'd have rejected them and not called them back?

Today, they would have applied with a better idea, in a later stage of implementation.

They can also be wrong about either or, they are people too. :) Like you said - only time will tell, and maybe we could be friends then, and life will go on.

Here is my writeup about this experience. It has two parts in it. The first part was written right after the interview. We were sitting in Mountain Views station and wanted to brain dump what we feel like at that moment. The second part is written after we got an email from PG, knowing that we were rejected.

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.

so today, Delicious is bought by youtube founders.

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.

"... On reflection I think the questioning about Delicious that we failed to articulate was not "how are we different" in a feature sense, but 'what will we do differently that makes money where they didn't.' ..."

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!

ok - monetizing ideas ... it might be a little out there ... but is there any way to corporitize trunk.ly? I'm thinking of user-accounts tied to a company, with analysis tools aimed at aiding knowledge management

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]

Blog on Trunk.ly rejection from Y Combinator - http://adamis.me/

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact