eBay/ StumbleUpon – a (long – sorry!) analysis

May 9, 2007

Marketplaces ChannelAdvisor By ChannelAdvisor

Note: This post is pure speculation on my part based on years of watching and working with eBay.  I have no inside information about their rumored talks/plans around StumbleUpon.

Today the WSJ (sub required) has more news about the much rumored eBay+StumbleUpon acquisition.  The big news is that a) it must be closer because the WSJ is looped in and b) the value is in the $75m range.

There are three school of thoughts on this one:

  1. Om has a theory about Skype being the driver for the acquisition.
  2. Valleywag has a humorous, but I think off-base theory: “Ebay has more money, from its outrageous monopoly on online collectibles and other auctions, than sense.” and is essentially buying SU somewhat randomly.
  3. Ok, here’s my theory (I know you were dying to hear this – right?)
    • eBay is run by a bunch of ex-consulting (Bain/McKinsey/Accenture, etc.) people so you have to start with a framework.  The latest eBay framework (e.g. this is CFO Bob Swan’s preso from a Merrill show I  attended – slide 4 snippeted below titled “where we play”) shows the latest eBay thinking on the internet and where the opportunities are.

Ebay_frameworkThis is pure speculation on my part, but I think in the halls of eBay
there’s a bigger version of this framework that looks at where eBay is
strong (this is the strong version) and weak (I’d imagine the weak
talks a lot more about search/social/content/online advertising and
maybe even SaaS) because those are areas that are growing faster than
eBay so it makes sense to try and add them to the ‘portfolio’.

  • Now let’s look at what SU does.  It’s a downloaded toolbar and website that a social network uses to make recommendations about various sites.  Based on the recos of the community SU can make intelligent recommendations.  For example, if I find 5 star wars related sites I recommend and someone else in the community recommends 5 sites and we have 2 that overlap – each of us is probably going to be interested in the 3 that don’t overlap.
  • If you’ve spent much time on TiVo, Amazon or NetFlix, you’ll recognize that this can be a very powerful (and often creepy) way to discover new products, movies and with SU – websites.

So here’s how SU fits into the eBay “where we play’ framework:

  1. Buy – You have the foundation for a nice little eBay buyer recommendation engine.  eBay has never been able to crack this one, so this could be great.  eBay has the ingredients ready made:  11m+ listings, millions of people in a community, etc.  One of the challenges is going to be mapping a general web reco engine to a database of listings that on average have a 7 day lifespan.  By the time you get enough recos from the community and build out the map, the landscape changes.  You could get around this by putting it on Store listings (longer life) or maybe EE (heck both) or shopping.com even.  This could be huge for Kijiji (just kidding, I can’t resist typing Kijiji – Kijiji, Kijiji, Kijiji, Kijiji – phew that’s out of my system now – try it at home folks, this is the strangest thing you’ll ever type).  Buy – great fit – Check.
  2. Pay – Once you’ve built out Buy, you can default to have the recos only serve up items that take PayPal.  Check.
  3. Communicate – I’m not a big Skype fan, so let’s assume that Om is onto something here. Check.

So right there within the “where we play”” I think you can justify $75m.  With $50b in GMV at a 10% take rate, you only need to increase GMV via a reco-engine $750m (1.5%)to get $75m in high-margin (you didn’t pay marketing) top-line dollars.  In my experience a good recommendation engine can push things much higher – 5-10-20%?  How much of Amazon’s revenues do you think come from recommendations? It’s not scientific, but I end up buying probably 1 in 5-10 of the recos, and that increases my Amazon spending substantially.  Also this could be a great way to increase activity of active users and re-activate registered, but inactive users (a growing problem for eBay).

Outflanking Google?
All that is great and in fact it’s a nice little smoke screen.  Here’s what I think the REAL value and thinking is with SU.  If I’m eBay, I’ve been kicking myself for, oh, 5+ years that I missed the search market.  Missed is an understatement really (whiffed?).  Here you have around $200b in value created that eBay could have easily participated in and totally missed the boat.  Not only have they missed the boat, but Google is in a position now to throw many flaming bridges over the previously unassailable eBay moat.

Sidebar:  There was an infamous WSJ article last year where WSJ discovered that McKinsey was commissioned to ascertain for eBay if Google was a threat.  Here’s an excerpt:

Few at eBay initially saw reason to fear Google, say people at the
company, in part because of a 2003 study it commissioned from McKinsey
& Co. McKinsey concluded that Google wouldn’t use its search
capabilities to break into e-commerce. That made Google a manageable
threat, say people familiar with the study. EBay’s dependence on Google
increased as it shifted ad dollars to online ads from traditional media
throughout 2004.

At it’s heart, Google is a search algorithm built on popularity to deliver unprecedented relevancy. The popularity is determined by inbound links (PageRank).  Then google has layered tons and tons of things on top of this core.

As history has proven, PR is a great system and works extremely well.  However if you wanted to build a different and potentially game changing system, what you could do is keep the popularity aspect of relevancy, but instead of links (easy to spoof, bait, programmatically create, etc. in theory if you had a big enough community of users recommending websites, you could bring a human/social element (e.g. maybe I want to read comments people have made, or chat with them, make them my friends so I’m alerted to their recos down the line, etc.) to the relevancy that might just be in a position to have a better mousetrap than google:  Social+human relevance+skype+eBay+paypal.

So for $75 and another $20-50m in investment, I think you could have something with a pretty decent search share. Success hinges on getting big chunks of the skype+ebay community into the system (piggyback downloads?) and then getting that community to start searching first in SU vs. a google.  If you can carve off 5-10% of share of search from that, now you’re in a position to build or buy a link/graphical ad network, and you go from there.  That’s easily a $1b EXTREMELY high-margin revenue stream and as Goog has proven, has the potential to be much larger than eBay’s marketplace core.

There are a lot of ifs in here, but I think it’s plausible and do-able enough that if I were Google or Yahoo!, there’s no way in heck I’d let eBay pick this up for $75m.  You could argue that the technology isn’t that unique and what not, but it gives eBay a pretty good start down this path that I wouldn’t want them to have from a purely defensive position.

eBay Strategies Readers, what do you think?  Am I way off base here and giving eBay too much credit or could this be the ticket back into the Search (finding) market?

P.S. I forgot to mention that SU fits Entertain – they have a youtube-esque feature now for finding and rating videos.