It’s nearly done - three days of Web 2.0 Expo excitement are over. Web 2.0 Expo was a conference which would have cost you a whopping EUR 1,300 if you want the full deal and didn’t book ahead, held at the fabulous Berlin Congress Center, was very well organized and turned out to be simply… boring.

I don’t even think it was the fault of the organizers - somehow most sessions (at least the ones I attended - which might be a problem of adverse selection) missed to inspire, missed to create a connect to the audience and too often defaulted to ’state the obvious’. Now this all might not even be the fault of the speakers - it might as well be the audience.

It definitely has been a while since I saw such a rather uninspired crowd. Gone are the days when you literally could feel the buzz, the energy to create. Now all you can see, hear and feel is the looming gloom of an economic downturn… or was it the simple fact that we are all sooo fed up by the damn Web 2.0 buzz?!

Anyway - I leave this place rather uninspired, yet knowing that these are great times: It’s a time where we will build real companies for real people with real business models. I - for one - am looking forward to this. It will be tough. It will be hard. But it will be good.

Now - back to London and get some real work done.

As the details about the recent layoffs were rather few and far between (the official version is still that eBay laid off 10% of their worldwide workforce - nothing else) the real meat behind the story is slowly trickling in: It looks like the cuts in Europe basically meant that eBay slashed the local teams in each and every European country to a very bare minimum (often literally only a handful of employees) - and combining marketing and marketplace functions on a pan-European level in Switzerland and the UK respectively.

If you ask me that’s a rather bold move - and one which might backfire heavily: If there is one thing eBay should have learned with all their failed expansion plans (Japan or China anyone?), it’s the simple fact that eBay’s business is a rather local one and that you need the local face towards the seller and buyer community. With these layoffs eBay basically gives up on this idea and acts like you can run eBay like a machine - no interaction needed. I heavily doubt that this will work even in the short run - and that most larger Powersellers (who contribute a significant part of eBay’s revenue) will now finally throw the towel; looking for better managed outlets like Amazon Marketplace or their own webshops in conjunction with search & price comparison engine marketing.

Somehow this makes me rather sad - I still think eBay has a very important role to play in the still growing world of ecommerce. Sad. Really sad.

Update: The German magazine Wirtschaftswoche just posted an article about the changes at eBay Germany and confirmed the information I had. [27/10/2008]

Ron Conway, one of the most prolific angel investors in Silicon Valley send an email to all the CEOs in his portfolio companies, giving them some war-proofen advice on how to weather the current market conditions. TechCrunch has the full story and also the complete text of his email - go read it, it’s very worth it.

In the last couple of weeks I had a lot of questions and discussions from and with startups both in Germany and the UK - they are obviously all concerned about the current market conditions and want to know what this means for their financing. Conway raises a couple of really good points which I can only underline - the two most important things to do at the moment are:

• Lower your burn rate to give you another 3-6 months runway. The more time you have, the more likely it will be that you will survive the current shake-out.
• If you are currently raising capital - raise as much as you possibly can (even if that means you need to give up more equity than you originally planned to do). Again - more money buys you more time and more time will give you a significantly better position to weather the storm.

I would add to Conway’s points that this is an exciting (though tremendously hard) time - personally I predict that companies who survice the current market conditions will be the big winners a couple of years down the line (very similar to what happened after the crash in 2001). So - if you can stand the heat, it’s a great time to start a company! Seriously.

Recently I held keynote speeches at two Etail events in Munich, Germany - one at Ciao (the price comparison and product review site which was recently acquired by Microsoft), the other at the Computer Reseller News Etail Summit (the leading B2B IT magazine in Germany). In both keynotes I spoke about Web 3.0 / Semantic Web and how it will shape the things to come both on the general web as well as in Ecommerce. Some people asked if I can share the slides - which I certainly can. So here we go (this is the English version, if you need a German language version, email me):

I just finished reading Van Lindberg’s “Intellectual Property and Open Source: A Practical Guide to Protecting Code” (published at O’Reilly) - an excellent book for everyone interested in the broader topic of IP (intellectual property), the US system of patents, trade marks, trade secrets and copyright and how developing code (not only in an open source environment) fits into this.

As an author Van Lindberg is an odd fish - he is a software engineer and praticing attorney. Which means that you will find a lot of legal principles explained through analogies in coding - which in turn makes them rather easy to understand (if you are a techie that is). The book starts off with some general information about the legal system in the US, gives great explanations on the various topics which are connected to IP (copyrights, patents, etc) and then goes on and explains the very real implications this all has on open source software development (both from an open source project as well as contributor point of view).

What makes this book so relevant for pretty much every startup which develops code or makes use of freely available open source code is the simple fact that it gives you guidelines on what you can and can’t do with your code, how you can protect your code and what to look out for when you deploy open source software. All this can bite you into your backside when you go through a thorough due diligence process with a VC firm in later stage financing rounds. And make no mistake - this book might be written out of a US-law perspective, the underlying fundamentals are valid in pretty much the whole western world.

OStatic recently reviewed the book as well (which was the reason why I went out and bought it) - you’ll find their review here.

I meant to write about The Receivables Exchange a while ago - I first stumbled upon them through an article in Inc magazine which struck me as a really interesting business model worth investigating.

Receivables Exchange in a nutshell is a marketplace for receivables (open invoices) - they basically took the age-old model of invoice factoring, shook the existing business model quite a bit and created an open exchange where companies who want to sell invoices and companies who are willing to buy them for a discount (their spread) come together. It’s an interesting approach - basically looking at an industry which is notoriously closed and intransparent and bringing a maximum of transparency into it. It might not be the sexiest industry - factoring is often referred to as the ugly pig of financing (which generally means there are not many competitors). It will be interesting to see how they perform under the current market conditions - theoretically they should be able to profit handsomely as a lot of small and medium sized companies have massive problems getting cash from their banks.

The company is an interesting case study as they went into a market which is not tremendously obvious (it’s not the next social network for dog lovers), it’s B2B (hey - B2B was sexy once… where is all the glitter gone?), it has a business model (and a robust one) and it’s a marketplace (thus bringing transparency into notoriously intransparent markets). Good to read and think about - here’s the Inc article.

I just read a post on Jochen Krisch’s excellent Exciting Commerce blog (it’s in German, here’s the english version of the blog — the article is not yet translated) about eBay’s latest foray into improving their crumbling business model. According to Caroline Malifaud from eBay’s Advertising Team they will introduce a new system for sellers to advertise their listings on eBay aptly labeled eBay AdCommerce in the next few months.

What AdCommerce will do, is to place keyword-based text ads on the search results page. These ads can be booked by eBay sellers on a CPC-basis and contain title, description, an optional image and link to the respective sellers listings. Ads are bought in an auction process (similar to Google AdWords) - the ad from the highest bidder will be displayed.

So far - so good. Plain, simple and horrendously stupid! First of all eBay will display the ads at the bottom of the search result page - pretty stupid if you want to optimize for clicks (the so-called below-the-fold area - everything below your initially visible area of a website - receives very little clicks). As sellers will only pay per click this might not be a huge problem - other than the fact that sellers will get little clicks from this promotion. A bigger problem will be click-fraud - you can bet on the fact that competing sellers in a category will simply click on the ad from their competitors to deplete their ad budget. I don’t see how eBay will effectively manage this - and given the assumption that most sellers will operate with relatively small budgets and in clearly defined categories this will become a significant challenge.

But the biggest failure lies in the very nature of this program: I can’t see many eBay sellers being too happy about the fact that eBay makes them pay triple - first they pay their listing fees, then they pay their feature fees to promote their item in the search results and then they should buy keywords? Well, well… let’s see how this will go down with the already annoyed eBay seller community.

Having said all this, the one fact which makes me wonder the most is the fact that eBay tried exactly the same thing a couple of years ago (with a partner: AdMarketplace) in the US and Germany and the program was a total failure. Only very few sellers were actually willing to cough up the money to triple pay, then they all fought with click fraud, they received very little total clicks and the conversion rate (the ratio between clicks and purchases) was despicable.

I don’t get it. Normally I don’t rant about eBay - I believe it’s a great company with a ton of fantastic and super-smart people and the whole concept still has so much room to grow. And what do they do? They take a complete failure and roll it out again - without making it better. Seriously eBay - you can do better!

Over dinner tonight with a good friend, I discussed the somewhat misguided approach of a lot of young startups to rely on advertising as their primary (and often single) revenue stream. Which made me think about the underlying economics and how blatantly obvious the flaws in this business model are:

Assume that your effective CPM (eCPM) is $1 (which is high - believe me, pretty much regardless how great you think your content and how good your targeting is, for a startup it’s nearly impossible to achieve a higher eCPM) — which means that for every 1,000 page impressions you get $1. Now assume that you have a tiny little office, a half-way decent server (you might be clever and run your stuff in the cloud, which unfortunately doesn’t change the argument), some infrastructure (phone, computers, electricity, a hot coffee from time to time) and pay yourself and your two buddies, who work with you on your startup, a little salary to pay for your personal expenses (again you need to sleep somewhere and eat something) and your monthly costs run rather easily in the range of $5,000 to (more likely) $10,000. Which means you need to generate and sell 5 to 10 million page impressions. Not an easy feast to start with…

Now imagine your eCPM is not $1 but more in the range of $.25 (which is the grim reality for most young companies) and you already need to generate and again sell between 20 and 40 million page impressions to pay for your little company. This assumes that you can sell all your PIs which again is highly unlikely - oh yeah, and your costs will go up because your servers suddenly need to serve not 5 but 40 million page impressions. It’s getting more complicated already…

And now the final stage - you want a VC to invest into your company. These guys don’t want to see revenues along the lines of $5,000. They want to see $50,000 and more. Much more. Let’s assume you need to generate a mere $100,000 per month. And your targeting and content is so great that you manage to sell your inventory for a whopping $1 eCPM; you also manage to sell 50% of your inventory (which again compared to industry standard is high). That all means that you need a mere 200 million page impressions (!) to generate this level of income. Not exactly a piece of cake - and exactly the reason why I don’t believe in ad supported business models.

So please, please, please - if you have a startup (or think about starting a company), please sit down with our good, old friend Excel and run some realistic numbers. Don’t cheat on yourself - it doesn’t help you and will only come back to you once you need to either pay the bills or justify your business model in a discussion with an angel/seed investor/VC.

Have a different point of view? Send me an email or comment here - I would love to hear your thoughts.

After mentoring 8 out of the 22 startups at Seedcamp on Tuesday and Wednesday here are a few observations - and some thoughts about their respective implications:

First of all let me repeat an earlier statement: All in all the teams were/are stronger this year than they were last year. Having said that I think the one observation which stands out for me is the fact that pretty much every single team I mentored was really (and I mean really) weak on the business model. The most common answer when asked about the business model (or to make it dead simple: The question about how to make money) was ads - most of the time combined with the fact that the team had some magic sauce to make ads more effective. None (!) of the teams actually thought about the fact how to sell these ads - they expected that by simply putting out their ad units, they would sell automatically. This is too bad - as it is the number one question a VC will ask these teams and it should be the number one thing on the forefront of their heads anyway. Also there is a ton of great resources out there which help you figure out how to systematically think about this problem (two of my favorites are Guy Kawasaki’s book ‘The Art of the Start‘ and Dave McClure’s deck ‘Startup Metrics for Pirates‘).

Secondly I found the teams rather unprepared for their mentors - there were a ton of amazing mentors, the teams knew beforehand whom they are going to meet and yet they pretty much never had any specific question to their respective mentors. This is too bad as I believe the teams could get even more out of the whole experience by preparing specific questions for their mentors.

Lastly (and here I clearly side with the teams) I have to admit that the amount of information which rained down on the teams is simply overwhelming. I had situations where six mentors met a single founder - and every single mentor had some valuable advice. Which is great - but can be a lot to digest.

Now we are looking forward hearing tomorrow who will win this thing… All the best to each and everyone and especially all the teams I mentored.

P.S.: The quote of the day is from one of the leading VCs who said the following in one of the mentoring sessions to a founder:

If you don’t have a dog - take the cat hunting.

John Lilly, Mozilla’s CEO, recently gave a talk at Stanford university about how management works (and happens) in a heavily decentralized organization such as Mozilla. John made the slides from his presentation publicly available on Slideshare - here’s the link and here’s the deck:

Check it out — these are really interesting and entertaining 57 slides. Don’t miss slide 11 and 12 - John proves that you can boast about your achievements without sounding the slightest bit arrogant.

What makes this deck so special though is not the nice way it is layed out, nor is it the fact that John can tell a truly amazing and astonishing story with the whole “rising as phoenix from the ashes” angle around the birth and rise of Firefox; there is actually a tremendous amount of information in the deck which can help any organization in tackling their challenges while growing:

• The innovation and decision making process at Mozilla often happens at the edges of their organization — often by volunteers, not employees. Incorporating this into an organizational model is a massive challenge for an organization. In corporations we see time and time again that employees in the ‘lower ranks’ make very, very valuable suggestions - which don’t make their way up to the decision making levels and thus will never be implemented. I would even go so far and say that Mozilla in this respect is not unlike any large organization - I bet that even an organization such as GE has a ton of innovation happening on their edges, but they don’t excel at the ability to make sure that this innovation actually gets incorporated into the company (not to speak about the fact that decision making certainly doesn’t happen on the edges of an organization).
• The need for communication - I’m sure you heard it times and times again and we all try to live it. But in the end most organizations pretty much fail and we only communicate (internally as well as externally) more or less the bare minimum. Mozilla in contrast is tremendously open with their communication — and pretty much all their communication channels are two-way (Wikis, Blogs, IRC, etc) instead of simple one-way channels (the weekly memo send from the department head to his team).
• On the topic of decision making there is a lot we can learn about the way Mozilla tackles the challenge to have open discourse and participation with the need to actually make decisions (we all know the problem - 10 people discussing from very different points of view: Nothing gets done).

If you don’t have time or the patience to flip through the deck (I guess you then also wouldn’t have read this far anyway), check out slide 54 which summarizes John’s key learnings in 7 simple bullets.

A little while ago (sometime in April to be half-way precise) I attended the Mini-Seedcamp event in London and met George Bevis, founder of a British startup called SpeedSell. I was intrigued about SpeedSell primarily out of two reasons: One - the company shares the same name as a company we acquired during my time at ChannelAdvisor (and where I believe that ChannelAdvisor still holds the trademark for) and two - SpeedSell sets out to do something we tried in a pilot while I was at eBay (we failed miserably by the way - which doesn’t mean anything anyway):

SpeedSell will buy your old console and computer games, picks it up at your place, gives you a guranteed price for your game and then tries to make its money back by selling the item on eBay (or later through other liquidation channels). Here’s an explanation in their own words.

So far so good. Here are a couple of the main challenges you have to overcome in this model to make it work:

• First of all you need to have a very good idea about the price you can actually sell the item for once you bought it from the consumer. If you get this wrong your whole margin can go south - or, in case you play it very conservatively, the purchase price you offer is simply too low for any sane person to sell his stuff to you.
• Second you have to deal with people who will sell you citrons - they claim that the disc works perfectly well, that all the manuals are included, etc. As this becomes your responsibility once you’ve sold the item to another consumer you have to test pretty much every item which comes in. That’s labor intensive and therefore expensive.
• Cost control in general is one of the major issues - you have so many moving parts in this game that it can become quite a nightmare to keep your margin in a healthy band.

The first issue is a rather interesting one - you can obviously use eBays past sales data to determine future prices. eBay even has a program where they license this data to you (which by the way is another drag on your margin: if you simply scrape the eBay site to extract this data, eBay will - sooner or later - come after you and shut you down; thus you need to buy their data through the Market Data Program). The problem here is that eBay’s data is notoriously ‘unclean’ - take for example a simple search for a mobile phone. eBay’s search (and thus their data) will return anything from the actual phone to batteries for this phone to bling-bling to pimp your phone. Cleaning this data to make robust price predictions is often not an easy task - it works quite well with media titles, which might be one of the main reasons SpeedSell started with PC and console games. It gets rather nasty once you turn to stuff like sporting goods, electronics, computer equipment or - god beware - fashion.

Having said all that - SpeedSell was one of the three top rated companies in April (the selection was done by a panel of well-known VCs) and is also one of the 22 finalist for next weeks Seedcamp week. SpeedSell is not alone in this market - the German Trade-A-Game is doing very well (with a slightly different business model) and the US-based ReCellular is doing phenomenally well with a similar business model in the market for used mobile phone.

In general I believe that this model has legs - if you are able to make selling for the average consumer easier than it is today (eBay was a huge leap - yet it still is quite a pain in the neck to sell something online) you have a massive market in front of you.

So - congratulations to George for his Seedcamp nomination, I’m looking forward catching up with you guys!

I just came across Virshi (labeling themself “The ultimate Shopping Experience” - quite a mouthful, eh?). Virshi is the latest project from Glaxstar, the company behind the Glubble kids browser (I wrote about this market a little while ago) and is build as a Firefox add-on. This in itself is quite exciting as I am a firm believer in creating products on top of browsers - using the browser as a platform (also something I wrote about a few weeks ago). But that’s not all - Virshi sets out to create a new form of shopping experience by putting some rather useful functions into a Firefox sidebar:

Virshi allows you to store products you found on the web in the sidebar, organize them by a multitude of criteria, share and discuss these results with your friends and find the best price on the web. If that all sounds a bit complicated I suggest you hop over to the Virshi website and watch the short demo video - it’s really nicely done and explains the concept in a few seconds.

What’s really interesting about Virshi is the fact that this is one of the next wave of applications which will greatly benefit from Web 3.0 / semantic web technologies (see also my blog post from a few days ago). What Virshi sets out to do (at least in my opinion) is to treat products as data, combine this data with lots of other, relevant information and create a new, website independent view on your shopping experience. It’s not hard to imagine that this can really take off once we have more and more data available in a portable format - price comparison data, consumer and professional reviews, stock information, geo-information about stores who carry the product, etc.

This might be early days for products like Virshi - but if you want to see (one possible) future for eCommerce I suggest you check it out!

Recently I have been thinking quite a bit about the implications Web 3.0 (the semantic web) will have on eCommerce. What strikes me as probably the most important implication is the sheer fact that by having data available in a format which machines can easily digest and understand, we’ll see more and more applications which will run on top of existing websites.

Let me give you an example: Imagine a world in which the data from price comparison sites, product review services as well as company rating services is easily available in semantic formats. Now you could build a super-robust plugin for your browser which allows you to right click on any webpage which shows a product (say the Amazon product page) and access an interactive layer which presents you with the best prices for this product on the web, cross-references these prices against company ratings (how happy were customers with the displayed etailers) and also shows you some great customer reviews for the product itself. You see that eBuyer.com sells the item cheaper and has an equally good rating from it’s customers as Amazon. With a single click the plugin will bring you to the item page from eBuyer.com where you can purchase the item.

Technically you could build something like this today - and some companies already did. Usually these services/plug-ins/extensions are rather poor as their data sources are not “clean”. Information needs to be scraped, the matching needs to be done by applying some level of guesswork, etc.

Now - how would this change eCommerce? How would it change buyer behaviour if the next etailer is not only a click away but you also get all the relevant information presented in an easy to digest format. It’s an interesting future we are heading to - and a great opportunity for entrepreneurs building new services around the infrastructure of Web 3.

I just read an article in the Wall Street Journal about artists shunning the Apple iTunes music store mainly due to their wish to earn more money. That’s all nice and I can understand the argument that Apple might be a bit unflexible in their pricing (every song costs $.99 - no discussion) and that artists fear that iTunes becomes the de-facto standard for digital downloads (well, in a way it is, I guess).

What really, really struck me was the following statement in the article:

Some artists see their albums as one piece of work, and don’t want them dismantled. Their handlers believe they can make more by selling complete albums for $10 to $15 than by selling individual songs.

This stays in stark contrast with the comment from Aram Sinnreich, a media professor at New York University:

This is a last gasp for the album format. [...] Most albums have only one or two good songs surrounded by little more than “filler material.”

Combine this with the following statement from the same article:

In addition, customers have demonstrated a clear preference for buying singles instead of entire albums.

WHOOOOHAAAA - we obviously have a strong customer preference, a great opportunity for every company to deliver a product which customers really want. How good is that? But what does the music industry do? Sell complete albums because they make more money.

This is so unreal… And they still wonder why kids download their stuff illegally?! Serious guys - you either learn to adapt, or you just die. And nobody will care.

In case you haven’t heard about it - Mozilla (the wonderful people behind our all most favored browser Firefox) released an alpha version of their latest and greatest invention today: Ubiquity.

What Ubiquity sets out to do is basically bring the power of mashups to everyone - Ubiquity brings together formerly disjoint parts of the web to create a new, intelligent environment right inside your browser. I spare you my probably pretty bad explanation and let Aza Raskin from Mozilla do the talking:

Ubiquity for Firefox from Aza Raskin on Vimeo.

Now - if you are not already thrilled by what you are seeing, consider for a moment what Ubiquity means in the upcomming world of the semantic web (or Web 3.0). The moment we have data properly tagged with meta data through XML, RDF and SWRL Ubiquity becomes a truly powerful concept which will change the way we interact with the web. Bring it on!