When Facebook bought Instagram, it paid a billion dollars for a photo-sharing website with almost zero revenue.
Putting a price on cyber assets is no mean feat.
Yahoo forked out $US35 million for Flickr in 2005. That deal had its critics at the time but Flickr was soon to become one of the top 50 websites in the world.
Facebook, itself the most over-hyped float in sharemarket history, swan dived from day one. Its billions of eyeballs, indeed the giddy growth in its audience, have proven no panacea for a flawed business model.
Listing at $US38 a share, it now fetches $US24. Facebook boasts 1.1 billion users, about 12 million Australians among them, yet its income is roughly $5 a person a year.
Meanwhile, Twitter goes from strength to strength. Who would have thought it? The most banal and shallow form of communication since smoke signal technology.
This windy preamble brings us to the float of the latest cyber stock, iSelect, which promises, according to some in spruiker-land, to "shoot the lights out".
It is certainly shooting the lights out on the fee front, with $11.9 million going to the float
brokers Credit Suisse and other advisers.
The market value at iSelect's $1.85-a-share listing price is $479.3 million. The company made $13 million in net profit on $113 million in revenues last year. On a price-to-earnings basis, the promoters point out the stock compares favourably to internet peers, Realestate.com.au. Carsales and Seek.
For a growth stock, with earnings, in the online space, this float looks like a no-brainer.
The rub is that, unlike Realestate.com.au, Seek, Carsales or Trade Me, the company does not control the inventory.
iSelect is more of a Webjet or a Wotif, a clipper of tickets. It runs comparison websites. The bulk of its earnings, which have grown impressively since inception a few years back, come from health insurance.
You click through, check out the best insurance deal, buy the product and iSelect takes a cut. Yet it has no dominant market position like a Seek or an eBay. If you are looking to buy or sell a house, you go to Domain or Realestate.com.au. You know where to go.
Comparison sites are more tricky. They have less pricing power, they have to advertise to maintain a market position. Barriers to market entry are low.
Then there is the small matter of earnings. iSelect's accounting is both creative and confusing. There are two types of commissions: upfront and trailing (on product sales as long as the client buys the product) and no clear information on cancellation rates and so forth.
If you are a bit old-fashioned, you may have a penchant for recognising revenue as revenue rather than revenue which may or may not arrive at some point in the future. But, what the heck, some people just like to book things upfront, in line perhaps with the 'why not?' accounting principle.
There can be no quibbling, however, with the timing of the iSelect initial public offering. Internet stocks have had a huge run - the likes of Carsales are up 100 per cent.
This must be close to the top, by the looks. And, thankfully for the equity markets faithful, it is unlikely iSelect will "do a Myer" and ruin the game for everyone.
Readers may recall that Bernie Brookes and the private equiteers from TPG flogged Myer at the most exquisite time. The online shopping demons were yet to haunt the sector, retail was buoyant, the government had just sallied forth with its "cash splash" and out came Myer, denuded of its hard assets, chockers with debt and marketing like there was no tomorrow.
Unfortunately, it spoilt the game for everybody else.
Regardless of the performance of iSelect - and they probably should have priced it more equitably for the aftermarket - its precursor is Virtus.
The world's first publicly listed fertility business is about the same size - a market capitalisation of almost $500 million. It debuted this week at a 9 per cent premium to its $5.68-a-share issue price and closed the week, premium intact, at $6.25. An impeccable start to public life.
For beleaguered capital marketeers this incipient rebound in offerings is like water in the desert. Only 44 companies listed on the ASX last year. That was well down on the 103 in 2011 and the bull market deluge of 250 listings during 2007.
Back to iSelect though. It might do all right, if management can push it into other comparison markets, the revenue-recognition thing works out all right and there are no price wars.
But rivals are nipping at its heels, namely British entrants compare themarket.com.au and helpmechoose.com.au.
And when you see the promoters, particularly Nine Entertainment, selling out, and $11.9 million of the $213 million raised going in float expenses - with no compelling reason to raise cash, except to cash out ... it makes you think twice.