EBay acquires Hunch for $80m | Hacker News Reader

EBay acquires Hunch for $80m

uncrunched.com

158 pointsspeedracr93 comments

Could someone experienced breakdown what a transaction like this looks like for everyone involved? I'm super curious. I've never been acquired, so I really have no insight into the process other than what I've read.

Here's the facts as I see them: Crunchbase says Hunch started in September of 2007 and had 23 employees on LinkedIn when they exited. (TechCrunch calls it a 20-person team, so I'm presuming that's all the employees.) They've gotten $19.2m in funding, let's just say $20m. TechCrunch claims the sale was "around" $80m.

So what does the breakdown look like? Who gets what? What are the likely investor terms?

My totally naive guess would be that the investors got at least a 1x liquidation preference, maybe more. I mean, did Hunch have any revenue? So there was at most $40m to go around to the people at the company. Of course, most of that probably went to the founders. Would maybe 20% of that have gone to the 20-ish employees? So naively pretending that each of the 20-ish employees got 1% for four years, did they each end up with an extra $100k/year? What's the likely distribution of shares among employees?

What are the transaction costs (lawyers, taxes, etc.) for this sort of acquisition? How long will the employees have to be at eBay to get their earn-out, and will that earn-out be in addition to their common stock in Hunch? Will they end up being paid less to work for eBay during their earn-out than if they were on the open market?

Of course, their are many other reasons to do a deal like this (passion for improving eBay's recommendations, for example), but let's ignore that for now.

Here's a wild, random-ass guess:

- Let's keep it simple and say that since this is an experienced team, the investors invested $20m at a $60m post, so they have 1/3 of the company. Since it is an experienced team, I'm going to guess no participation or anything else hokey. 1x liquidation preference is likely but won't matter in this scenario b/c preferred will convert to common.

- With a 20 person team, you've probably given away 5-10% of the company to employees. The later employees probably did not get much, but then again, they weren't around for that long. Let's be very generous and call it 10%.

- That leaves the founders with 57% of the company.

- Let's just not count transaction costs, they're probably in the $75-$200k range, and it's not super material to this discussion.

In that scenario, the founders make $11.4 million each pre-tax, or about $8.55m post-tax (assuming all founders are equal, which is probably also unlikely here). The employees make $8m pre-tax total, or ~$350k each pretax, or ~$190k each post-tax, but the distribution is going to be skewed towards earlier employees. Investors make $26.4m, which is a pretty poor return for a 3-year investment -- investors are mostly looking at IRR, which means the longer an exit takes, the higher the outcome needs to be to make it worth it.

This is just a wild ass guess on what a transaction like this might look like, but lots and lots of things are likely way different than I guessed here, and I can guarantee that I've over-estimated the amount of money everybody made, because as a general rule of thumb that always happens.

EDIT: > How long will the employees have to be at eBay to get their earn-out, and will that earn-out be in addition to their common stock in Hunch? Will they end up being paid less to work for eBay during their earn-out than if they were on the open market?

Probably 2-4 years for the earn-out. The employees might get more retention bonuses to stick around, above what they would normally get from the transaction. They will not earn below-market due to the retention package.

you are calculating the tax here at around 25%. Wouldn't this count as long term capital gains tax (if you hold the stock for more than 1 year, which it seems likely since the company was incorporated in 2007) that is currently being taxed at 15%?

State/City long term capital gains in NY are around 11-12%. In California they are the same as the income tax rate of 9.3%. 25% is the blended Federal/State/City rate.

ah forgot about that - thanks for clarifying.

Chances are the options weren't restricted options that employes could exercise ahead of time and that there was a 4 year vesting cliff. Most people don't exercise options at vesting time so I'd bet that most of the options were taxed as short term capital gains (and that would mean that 25% is actually probably a bit low for most employees with where this would get them to).

This sounds about right. I would say that the investors probably own a bit more of the company than you guess because the cash didn't all go in at once but in 3 stages (I think? The crunchbase lists 2 different A rounds which doesn't make sense). Valuations for the earlier stages we almost certainly a good bit less.

I would be surprised if the investors agreed to this acquisition without having a 2x liquidity preference.

This is worthwhile for eBay, since only a 1 or 2% improvement in recommendations could be worth millions of dollars a year.

"It’s been a long journey for Hunch."

They were founded in 2008 and sold for $80m 3 years later. That is considered a "long journey" these days? Statements like this give me that "we're in a bubble" feeling.

I am somewhat disappointed since Chris Dixon has said Hunch would be in it for the long-term. (I've followed Hunch since it's early days* and always wanted to see it go big.) $80m is a nice pay-out though, the systems they've built have the potential to bring an immense amount of value to eBay's offerings, and a lot of that money will trickle down to other startups via angel investments.

* http://hassy.posterous.com/my-favorite-startups-hunch

    Hunch is not for sale. The founders have already sold companies before and have no desire to sell Hunch. We want to build it into a large company.
    
    hunch responded one year ago

http://www.formspring.me/hunch/q/1392572735

I've learned that in one year, one's goals can change dramatically.

It's also one thing to say something. But the situation changes when a ton of money is actually thrown right in front of you.

If the number is right, $80m is not a ton of money for people who have already had big exits and took $20m in VC to start this company.

It could also be that the company is doing far worse than the business plan they originally showed out to the VCs : If there's lower growth rates, it's far better for everyone to notch up a quick 3x, rather than see the business dwindle into obscurity (the alternative case, in the minds of the VCs, if they're not going to 'go big').

FlickR only sold for $35 million. They took some funding prior to that too, so it wouldn't have been a huge payout for Fake.

I'm suspicious when any startup makes this claim, not because I don't think they legitimately mean it (at the time), but because it rarely ends up being the case as markets, situations, products and lives change.

> Hunch is not for sale.

All companies are for sale.

As in that anecdote, all people are for sale, just that you have to find the right price.

I think Hunch was actually founded in 2009

Agreed. The "standard journey" for a big, meaningful startup is more like 7 years. A "long journey" in my book would be 10-15 years.

With some very notable exceptions (YouTube, Zynga, Groupon), built-to-flip and talent acquisitions dominate the 1-2 year exit companies.

Wow, this is like a knight in a white horse for Hunch.. I seriously doubt they were making any profit... This sounds like a talent acquisition. I mean this is ridiculous... but it goes to show you.. you gotta show up for the game (the startup game, I mean) to win it... $80 m for a site that doesn't even get 1 million uniques (per Compete) a month is a STEAL. I mean.. seriously, everyone who has a CS background should just start something on the side if a deal like this is possible.

> you gotta show up for the game (the startup game, I mean) to win it...

This type of thinking is why the valley is so full of zombie companies pumped up with capital, but with no real traction or any way of ever being profitable

PG and many speakers at startup school repeatedly said that its easy to spot the ones who are just doing it because they want to be "that guy" like Zuckerberg - the ones who want to play the game. Instead of starting a company to figure out what customers want.

What motivates you to start a company is really important. Especially during the downswings that kills most startups (google "trough of sorrows").

+1 on the talent acquisition, although much like Slide, About.me, this shows that once you're lucky, twice you know who to speak to if you've gotten to a sizeable exit before. Also helps that eBay is looking for its second act to avoid the fate of a crumbling Yahoo!

> everyone who has a CS background should just start something on the side if a deal like this is possible

Chris Dixon is a veteran investor and entrepreneur, with enormous visibility in the tech world. Contacts with potential acquirers are a non-issue for him.

A poor CS fellow could have created a version of Hunch that was 100x better and they still would have 1,000x less of an exit opportunity as compared to any startup Dixon founded.

Get real.

If this really is true then it's a sad indictment of the state of the startup scene.

Perhaps your hypothetical fellow should focus on building a "lifestyle business" instead of trying to get big and then get acquired...

Technically he should be a software developer with a $120-200k salary working at Google, saving money with compound interest and getting company stock. Thats the easiest way as a CS grad to become wealthy - if thats your goal.

That's when you don't factor in 'skills'. If you're 'any' CS grad, you're right. Be if your skills are above average (CS and business sense), this skews the return of investment of the startup.

Well, not any developer can get $200k at Google. So skill is important in both respects.

C'est la vie. It's not the startup scene - it's a basic principle of society, unfortunately. Who you know is just as important as what you know or can do. The best thing to do is find people to know by showing them what you can do.

Agreed, this is completely absurd. Say what you will about Chris Dixon (zomg he doesn't code, wtf!!!!!111), but the engineering team at Hunch is second-to-none.

Grandparent makes it sound like Dixon pulled this off in his spare time. He assembled a team of A-level engineers, then secured millions in funding so they could develop an awesome recommendation engine with almost no clear revenue stream. I challenge anyone to find a CS grad who can match that in his or her spare time.

EDIT: restructured on second read.

I think we need to move away from all the nonsense of "A-level engineers" or superstar developers and so and so and so.

Carl Henderson of Flickr fame looked like an A-level engineer back then (especially with his O'reilly book) yet some of the Flickr engineers baffled when they saw the Flickr codebase.

Sometime it's all smoke and marketing hype.

agreed. ooh.. a recommendation engine backed by semantic cloud-based peer to peer infrastructure... for all we know, it could just be aggregating data from twitter and seeing whether the word "good" or bad appeared more often to decide whether to aggregate it.

Have you used Hunch? Have you ever written a recommender system? Have you even read the source of one??

This isn't some naive Bayes spam filter from 1992. Hunch is really, really accurate.

Actually yes, I have written a recommender system.But if there's 1 thing I've learned over the years is that you don't go overestimating the complexity/intelligence of any system unless you take a peak underneath. For all you know, Hutch could be just hiring curators with good taste rather than write some complex system that would need to be constantly changed and updated. But people have all the incentive in the world to "over fluff" and overcomplicate their work to make it sound valuable. It's smart, but you gotta be skeptical.

Flickr was a bonafide cultural phenomenon before Yahoo let it die on the vine. That doesn't happen with crappy engineering. Less-than-optimal engineering, yes. That's the tradeoff between getting a product out the door and never delivering.

Netscape's code was pretty crappy too.

I never refer to a perfect utopia where the code is superbly indented, nicely commented, looks great, etc.

But hacks will always be hacks regardless the discussion of "getting product out of the door" vs "never delivering".

A few years ago I met a 'superstar' type developer who was considered hot shit amongst the usual crowd. I met him with a view of hiring him (or, trying to) mostly because of his reputation.

I think i must have been the first person who really drilled into his knowledge, because he was more interested in talking about who he knew, projects he had been associated with etc.

I got a funny feeling about him, and asked that he do a simple code project. A week overdue later, he asks for our FTP credentials to upload it. I say that we don't run ftp anywhere, and that he needs to scp to the box. He said he had no idea that ssh can 'copy files'. urgh. He later worked out that his GUI FTP client supported SSH and got so excited about it that he messaged me to tell me about it.

I have never judged the capabilities of a developer based on reputation since.

It's kind of hard to judge this superstar developer you mentioned based on his knowledge of some of those UNIX tools.

Josh Bloch admitted himself that he's kind of not really "up"-ing his knowledge/skill when it comes to tools (git, eclipse, intellij, etc) but he writes damn good code with just Emacs.

He knows his Math & Stats to debunk benchmark and stuff like that and he knows the pitfalls of complexity.

ye but the mistake I made was to take his reputation and have a pre-conception about his abilities to the point where I almost didn't test him on anything.

he also fell apart in the code test, which was pretty simple (using a library I had written, which he made a copy of and edited). I have no idea how this guy got his reputation, but it seemed to be more amongst non-tech tech people, if you know what I mean (who are easier to bullshit, I guess)

when I started my career I worked with a bunch of old IBM mainframe guys who had no idea about modern web stuff but were some of the best programmers I have ever met. I could take a problem to them in a language they didn't know (so for eg. VB.NET) and they would talk me through it.

In what universe is starting a company with no clear revenue stream a good thing?

100x "better" in what sense? Traffic? If so, I suspect suitors would line up.

While Dixon's reputation is a great asset, I wouldn't say it's a multiple-orders-of-magnitude difference.

I dunno, their recommendation engine was scary good -- the technology Ebay's acquiring isn't anything to sniff at. They reportedly have a defensive portfolio of pending patents: http://www.formspring.me/hunch/q/169651283274266480

Ebay's been on a spending spree this past year and I'd say this is one of the better deals they've made, depending on how long they've locked in Hunch's team.

I wonder if Google was in on the action for Hunch.

Apparently Google bid $60million for Hunch - http://www.businessinsider.com/not-for-sale-hunch-has-turned...

This sounds like a talent acquisition.

I would agree, and it's a damn good talent acquisition. Hunch's team is made almost exclusively of engineers who (a) went to some of the best CS programs in the world (CMU, MIT, Stanford), and (b) really really really know their shit (just have a 5-minute convo with one of them).

Here in NY, Hunch had a reputation for opportunistic hiring. They had lots of funding and an incredible core team, so they'd only hire engineers that astounded them. That's a great way to set yourself up for a talent acquisition.

Would anyone happen to know how much 'exit bonus' these MIT/Stanford CS Grads get once a company is sold?

40% of them will start their own companies once their non-competes expire.

There's really almost no such thing as talent acquisition , IMHO:

1. No dev is worth $1m. Their skills may generate as much and more, but only in a specific context. Being great at catching a football is worth $50m in context of the NFL, $0 without it.

2. You can't really buy talent. At best, you can keep it, working at 1/2 of its ability, until fully vested and moves on. Alternately, eBay could create an environment as challenging and attractive as Hunch - but (a) by definition, it's almost impossible and (b) if you could pull it off, you could hire the similar talent for much less than $80m.

You think Ebay can't just, you know, hire this kind of talent?

I'd think walking through the halls of those CS programs offering $250k a year in salary would get a a bunch of A-Level talent at a much lower price then they paid for Hunch.

I never used Hunch much, and I'm not sure if an 80 million exit with 20 million in VC funding is so great for the founders, but I suppose it gives eBay a footing in NYC - that's the part of the article I found most interesting.

15 million a piece. I don't know how long they were working on it but the very top billed actors tend to make, what, 20 million for movies that take ~3 years to make.

So you assume the VC's only scored a 1x on this deal and had no special treatment for their preferred shares such as accrued dividends, liquidity preference, participating preferred, etc?

I went by what the GP said.

Most top actors spend 3-4 months on average per film. It's a scheduling things where you need a lot of people to film a movie some of which have offers for 3-4 movies that year. So people are working 12 hour days in one short burst and a lot of time in editing / post production but there is a ticking clock because investors want their money back ASAP. Reproduction can be slow simply because there is not a lot of money involved in the project yet.

Top billed actors have an excellent chance of making their full rate on an average movie.

Founders have an excellent chance of making nothing on an average startup.

So what do they need to make exactly? For me, anything beyond FU money is pretty much irrelevant.

Chris Dixon talked a lot about "going big", growing big companies, etc. He talked a lot about a new "data tier" for the web and how Hunch would be an intelligent filter for all the data. I wonder what changed?

A chance to leverage synergies?

Something about Hunch's trajectory is utterly predictable and ever-so-slightly depressing. I suppose building to sell is a perfectly valid business model -- if you succeed -- but it makes me feel a little dirty. And if nobody buys you, you're well and truly f#%@$ed.

Although who am I to argue with $80 million.

And if nobody buys you, you're well and truly f#%@$ed.

Really? If you don't have a 'building to sell' roadmap in the first place, why be depressed? If you're building a long-term, stable company with real value and profits, you're not for sale in the first place.

"I have a hunch nobody ever meaningfully used Hunch."

http://twitter.com/#!/rafat/status/138633186300264448

Hunch is like my divorced neighbor who couldn't work it out, so now they're giving up and short selling the house and bringing down the values for everyone.

"It’s been a long journey for Hunch. The company was founded in 2008."

Knowing how long it can take for a small company with a good idea and product to launch, I wouldn't really call that a long journey.

Still, congratulations to them. I hope eBay is able to make something good out of what they have created so far.

I never felt the need to use Hunch on its own. Now that it can be more of a tool to supplement other objectives, it sure might be the right tool after-all.

Hunch was always more of an API company than a destination site.

Can you name some companies which were using its API? Their developer site doesn't take any names.

What I meant was the goal of the company was to build technology and license that via an API, not become a destination site.

did anyone actually use Hunch seriously?

Taking this question further, what did Hunch help you do? I tried it early and passed, really interested to hear what it has done for others.

I tried it a couple of times too. First time I figured it was basically a slightly better stumbleupon. Next time I tried it was still the same thing and never went back.

No freaking kidding - only thing I saw the founder do was interview people on TC. Never heard anyone mention hunch outside of those interviews. What is ebay thinking?

> What is ebay thinking?

Impressive back-end tech (the recommendation engine) that can now process millions and millions of items on eBay to help increase sales (something like a quarter of Amazon's sales come from recommendations).

Very hard to imagine ebay doesn't already do a good job of this with their years of experience. I mean it's possible, but I doubt it.

80 million dollars is a lot to pay for a recommendation engine.

Netflix had their own and paid a million dollars as prize payout for an absolutely state of the art system.

If it really was just a tech acquisition they massively overpaid.

Don't forget that Hunch's system is real-time (this is very hard), and is more than just scalable collaborative filtering for items of the same kind. They can ingest a Twitter or Facebook profile, infer your "taste profile", and start recommending things instantly; "things" meaning books, films, music, events, places etc etc.

I don't believe real time is very hard. Something that ebay can' already do. And if they can't already do it, how will hunch come in and solve it. It would probably require major changes to db's etc if that is why ebay can't already do it.

> I don't believe real time is very hard.

At scale it is. More here:

http://www.quora.com/Hunch/Is-Hunch-actually-doing-anything-... (this one got a vote from one of the founders)

http://hassy.posterous.com/my-favorite-startups-hunch

How does that vary much from what Etsy did with their gift recommendations application based on your/friend's facebook profile

See my above reply to suking--$80m is kind of a bargain.

As someone who builds recommender systems for a living, let me assure you that the output of the netflix prize was perhaps 10% of what you need to actually build a good recommendations product.

Sure, I wasn't implying eBay overpaid by 80x but 79 million dollars is a lot of product development once you've paid 1 million for a proven state of the art proof of concept.

Obviously eBay must see other reasons for the acquisition than just the tech itself.

Also, if it's also in large part a talent acquisition then they're going to paying more than 80 million since they're going to need to pay for some new pairs of golden handcuffs to keep the talent around.

Apple bought the tech behind Siri for $200m in 2010.

MS bought Powerset (and their NLP-type search engine) for $100m in 2008. They also bought Farecast (travel pricing forecasting) for $75-115m in 2008.

Hunch for $80m is a steal, as far as talent acquisitions go.

EDIT: I'm not sure if they were product or talent acquisitions, but does it really matter? All companies listed above obtained state-of-the-art algorithms and the teams that developed them.

Everything you just listed wasn't for talent - it was for products... try again.

In all cases, those companies got the algorithms + the team that developed them. Are we really debating the difference? Or were you sitting in MS/Apple/eBay's boardroom?

I frequent HN, Reddit, Twitter, etc. and have heard the name Hunch once or twice, but have no idea what they do and don't know anyone who uses it.

There's a different kind of bubble in the valley, it's a contained ecosystem that people get caught up in. Similar to the social bubble surrounding a college or church. People inside the bubble often cannot see the forest from the trees.

In the bubble it's easy to get to know all the trees, but from the outside it's just a forest and few trees are easily recognized or stand out from the others at all.

Anyone have any suggestions about how Hunch would scale to eBay levels?

Seems Hunch has gone the 'Mega Machine' route (1 T RAM, 48 Core [1]), from 700k users (80m Total TTHY / 113THAYs [1]), make 250k API users

How do you scale that out to eBay numbers?

[1]http://blog.hunch.com/wp-content/uploads/2011/05/110509-HUNC...

Here's an interview with Chris Dixon on the Ebay purchase.

eBay is an interesting challenge because so many of the listings don't have metadata like product IDs.

The work Hunch has been doing on the open web is a good fit for this kind of unstructured data.

http://www.betabeat.com/2011/11/21/chris-dixon-ebay-hunch/

Is Arrington rebuilding a proper tech blog as a counterpart to the growing news machine of Techeablecrunchbeat?

I think it's just in his blood at this point to blog about tech news. Good for him I say. If you love what you do then do it.

Whatever Hunch is really good at, we haven't seen it yet. I assume Ebay has, and that it has everything to do with crunching customer browsing & purchase behavior to offer them products they're most likely to purchase.

I think their predictions were actually really good. A while ago I tested several movie recommendation engines and hunch was the only one that delivered.

On the downside, they really asked a lot of questions, which made me uncomfortable with using it.

There is a great play around the collection of data outside of just social space. Hunch has done a great job aggregating this data via a social means and displaying the data to the masses in an appealing way to drive more people towards their "recommendation" engine. I work for a site called Ranker and we're focused on a similar data play with much greater focus on SEO.

* http://www.ranker.com/

I like their movie predictor: http://hunch.com/movies/

Reply on news.ycombinator.com