Facing The Absurd

Happy New Year’s everyone! Hope 2011 will be a great year for you!

Anyway, just wanted to share an idea I have had for some time now. Never really got a chance to do it myself initially, then I decided that the business opportunity isn’t that huge and I have to focus on bigger niches. It’s basically an aftermarket play exploiting the drop game, a domain flea market if you will.

The idea is simple – all of us domainers drop domains sometimes, we choose to let them expire. The idea is that somebody would build a service where you could push these domains, say 2 weeks before they expire, and offer them for $10 a piece + renewal fee. If two or more people would be interested in the same domain, a standard namejet style auction would follow. The service operator and the domainer would then split the proceeds 50/50.

The service would just need to create accounts at all the main registrars where domainers would push the domains and do some simple programming of the marketplace, not much work imho.

This idea is based on a simple premise: some people can find gems (in their perception) in your garbage.

I personally drop a couple hundred thousand domains a year. The thing is that my only criteria for renewal is if the domains make reg fee. If they don’t, I simply let the domain drop and there are definitely gems in these drops that other people can discover but I don’t have time for.

I hope I created a new buzzword “drop monetization”:). Hope somebody picks this idea up. Good luck with it!

I am fairly critical about the lack of evolution of the domain parking industry – it really hasn’t evolved much in recent years. Since PPC has fallen so sharply, it still is a puzzle to me why parking companies are not becoming more inventive when it comes to monetization.

One area I see completely unexploited is behavioural targeting on parked pages. When you look at ways how parking pages are optimized is is really all about contextual targeting. You basically have a universe of users hitting a particular domain, you optimize the ads based on the the domain name, what users are clicking on and maybe on location. However the point is that if you would take the universe of all US users hitting the domain for example, they would be all shown the same ads. Why not actually target the ads more to the particular user, for example based on his interests, gender, age etc and mix this behavioural targeting with the standard contextual targeting?

Parking companies themselves collect massive amounts of data by which particular users could be categorized. They could also partner with big ad networks that have collected even bigger swaths of data such as Doubleclick. Then use this data to improve targeting and monetization. If big e-commerce sites such as Amazon or Netflix can exploit this data, I don’t see a single reason why the parking companies couldn’t. This data is a potential goldmine that is just waiting to be collected by somebody.

Some things are irrational or absurd, without explanation. One of these things seems to be a complete lack of revenue data about domains on the various aftermarket platforms. Even though a domain’s parking revenue would strongly influence the selling price of a domain, nobody is even bothered with communicating this information to potential buyers who may be interested in bidding and revenue is a key factor for them. For some unknown reason potential buyers are forced to estimate and their accuracy determines their success. Buyers focused on traffic/revenue names on the various dropcatching platforms can strongly support this argument.

Domains that generate a constant stream of parking revenue are the most liquid part of the domain universe because most smart domain investors buy on yield, not on potential development potential, end-user resale value etc since these values are just hypothetical and speculative whereas yield=cashflow. There is always somebody who wants to buy yield. The people who have had the biggest success in this business focused on yield.

Cashflow domains generally sell in bulk portfolios for a given revenue multiple. Say you buy a portfolio o 1000 domains making $10k/month for a 5 year revenue multiple, so you pay $600k for it. You are buying/selling the revenue of the entire portfolio, you are not really looking at the individual domains.

This is where the opportunity in the aftermarket lies – If you would unbundle the portfolio and sell it by individual names, you might get a total of 7 years revenue, for example, for the entire portfolio. The reason being that certain people might see more value in certain names for which they are willing to pay more (because for example they have a better way to monetize the traffic). So somebody may be willing to pay a higher multiple for your travel traffic names, somebody for finance traffic names etc. Somebody may see development potential in the name. Somebody might assess the risk of the domain differently etc.

Somebody smart can quickly carve out this niche in the aftermarket and start focusing on revenue names and include detailed parking stats with every domain for sale. Buyers will then simply be bidding based on what revenue multiple they are willing to pay for the name, ideally in an auction format.

In a previous post about Bido, I said that I think Bido will have to fundamentally change its model to start making money. This is the direction I think it should go, because this is where potentially the money is and it is an unoccupied niche. Instead of focusing on names with predominantly speculative value that usually catch the eye of 1 or 2 bidders (as hard as you try), why not focus on revenue names where you are likely to get interest from tens of different bidders if you can create a liquid marketplace.

I was just looking through Bido at the recent sales page. Since Bido get’s so much PR and buzz, I was really surprised about the miniscule amount of volume in dollar terms. Looks like on a typical day maybe $1,000-$1,500 of sales go through. That’s $100-150 of margin for Bido a day. And a hell of a lot of effort is put into that from Sahar’s team with no doubt to even get that result. I don’t really want to show off or anything, but just to put that number in context, I make that kind of money in less than 10 minutes, 24 hours a day, just from parking.

What the example of Bido clearly shows us is how difficult it really is to create a new viable aftermarket platform and especially get the model right. I think Sahar&co will really have to fundamentally change Bido’s model and I sincerely wish them a lot of luck, because any efforts like this help increase liquidity, which is always positive for all of us.

Overall, if you look  at the various aftermarket platform models, I think only some work very well, some moderately and some don’t at all.

Somebody who I think got the aftermarket model working really well is Namemedia with BuyDomains etc. Why it is so nicely profitable is that to a large degree, Namemedia is what I call in the business of proprietary domain trading. They own the inventory (or most of it) that they sell, hence their margins are really thick. Whereas others just rely on their 10% cut, Namemedia takes almost 100%. That’s why they can market their names proactively. Dark Blue Sea has been trying to do something similar to that with it’s Domain Distribution Network, but they are clearly not even close to as good as NameMedia is on this.

Another aftermarket model that I think makes a lot of sense is the dropcatching-to-auction model of Namejet, Snapnames and Pool. If you create liquidity in the marketplace, you can snap up domains for $7 and sell them for $79 or even thousands of dollars. Obviously most of the inventory comes from preferred registrar partnerships so the margins are not that high (as they have to give a big chunk to the registrars), but these dropcatching services definitely take a bigger cut than 10% that for example Bido or Sedo rely on.

Rick Latona gets it right as well through his whole aftermarket package (newsletter, auctions, active brokering). He also engages in what I call a lot of proprietary trading, a lot of the inventory he sells is his.

To a lesser degree I don’t think the whole marketplace model of Sedo (on a standalone basis) is that awesome and profitable. On a typical month, Sedo sells something like $6 million in inventory, with a 10% margin of $600k roughly. However Sedo has a HUGE overhead to keep this operation running, spends significant amounts on marketing etc. There’s probably very little left of the $600k a month after all the costs. However why this model seems to work is the marketplace’s impact on Sedo’s parking business. Because of the marketplace, Sedo gets a lot of parking business, where it can make thicker margins. Pretty much all the small guys making $50 a month on parking park with Sedo now, but they probably have thousands or maybe even tens of thousands of them so it adds up. The impact of the marketplace on the parking side of the business is exactly why Namedrive went into this business with its NDX Market. Overall clearly, on a standalone basis, the marketplace model is nothing very profitable.

So bottom line is that if you want the marketplace model to work, you really need some kind of upsell to make it work – to parking, a registrar or something like that.

One thing that has been puzzling me for some time is the lack of institutional money in any structured way in the domain business. More institutional money is clearly a prerequisite for higher domain valuations.

When you look at it today there is only a little bit. Marchex/Fabulous/Tucows are publicly traded. Oversee, Demand Media, Skenzo, Name Media have all taken aboard funding, very decent amounts. Then we also had iReit, which sort of flopped. Various domaining companies managed to take on some debt such as Reinvent. Domain Capital at least brings a little leverage effect into the business (they have $30 million loaned out). But that’s pretty much it.

But why don’t we have more hedge fund-esque operations that would take on investor’s money, maybe even tie in a little leverage to increase ROE and start buying up portfolios? The only exceptions I sort of know of are DomainIvest.LU (they have raised their first 10 million Euro fund, which is now invested I hear), mad.biz runs some kind of private partnerships, where they bring in limited partners. I do a little bit of that as well. Maybe InternetRealEstate does some of that as well.

So what are the main reasons behind this lack of structured institutional capital?

One factor is that the first round of institutional capital that poured in sort of got burnt. This was before Google/Yahoo started heavily cutting payouts via various quality related claims, before the downturn hit etc. To really illustrate this: If you bought a portfolio in 2007, today it would be probably making 60-80% less on PPC than it did at the time of purchase.

Second is transparency. Michael Gilmour sums it up pretty well in his article here, so no need to elaborate further.

Another issue may be size. When you really think about it, the domain industry is pretty small. My estimate is that Google/Yahoo combined probably pay out about $40 million a month to the domain channel now. That’s already not much, again taking the more macro perspective (compare it to say the size of the bond market). Worse, the market is highly fragmented. There is not probably a domain portfolio owner that would own 10% of this market. Probably Oversee, Reinvent etc may be close to the 10%, but more likely in the 5-7% range, when it comes to their owned and operated portfolios. The domain biz may simply be too small to get on the radar of the big various funds.

And lastly, there is the issue of risk. There is the monetization risk (that ppc will further decline or a big upstream ad provider leaving the space and not syndicating its feeds to the domain channel), maybe a degree of type-in traffic fading away (more long term) and then there is the legal risk. I hope eventually somebody smart will find a way how to securitize the cashflow from domains and create domain derivatives that could for example separate the the yield of a portfolio and its risk. The same way that for example in the bond market you have credit default swaps (through which you can basically separate the yield of a bond from the risk of non-repayment). Doing this would be a huge boost for the business and would really help institutional money to flow in in masses.

So will be see an influx of institutional money coming into domains in the next 3 years?

I really think so. PPC is certainly not going to fall as much as it did in the last 2 years – I actually think it may be relatively stable and new monetization techniques (refer to previous post) may actually even bring a little bit of upside. I also think there is going to be a new breed of domainers-turned-domain fund managers that will start bringing in the institutional money – because the industry is so complex it’s rather difficult for an outsider to do that. And lastly, with us getting out the recession I think investors will have a higher appetite in risk again and start exploring more alternative investments again.

It’s likely we’re going to see some industry consolidation this year, probably at least one “blockbuster deal” for the industry, maybe even two. So what are my thoughts on this and what I think should happen?

I think this “blockbuster deal” will somehow involve Namedrive. Namedrive is the fastest growing parking company, they are very aggressive, smart and flexible. They have a strong foothold in Europe. They have a lot of expertize in monetizing international traffic (international traffic is the next battleground me thinks), especially using the AFD feed. My personal thought on this is that Oversee.net should take over Namedrive, it would be a pretty good fit. Oversee has just opened a Frankfurt office and they are clearly signalling they are interested in Europe. I think Skenzo could be a buyer as well. Sedo probably not, the antipathy between Namedrive and Sedo is simply too big.

Also, I see Fabulous really getting marginalized over the last 2 years. I think it’s an option for them to sell the parking side of their business. Could be a nice scoop up for Sedo, which seems to specialize on these smaller bolt on deals (e.g Parking Panel, acquiring the parking biz of Dotster etc).

On the Yahoo! side of the fence I can see potential of something happening. I can see Skenzo being acquisitive, they have the institutional capital and can make a move. Parked.com would be a decent target I think, it’s relatively small, could be picked up well and would be a good fit. HitFarm/Reinvent won’t probably take part in these consolidation games.

And then there are the deals that look unthinkable at the moment, but could happen. Oversee.net is probably looking for an exit eventually. It could be bought by somebody, likely coming from outside the direct domain business. Why not Google for example? Sounds crazy, you can never rule it out though. Could be an ad network as well, would be a pretty decent fit. Why not Verisign. Sounds even crazier. But they clearly have the fire power to do it. Something similar could happen with Sedo as well.

I’m just really thinking out loud here what can happen….