It's been hard to write this essay on Skype and Google, because the world is changing faster than I can put the words together. And whilst I'm happy with the general thrust of the argument, I'm sure much of the detail will be muddled. I hope you've got the patience to stick with it, and look forward to your feedback.

Putting together Skype and Google, whilst no match made in heaven, does have a lot of synergy. The Skype client, or something very like it, has the potential to re-invent telephony. (That, after all, is the point of the Stupid Network, not disintermediating legacy voice toll charges.) It just requires you to stop thinking of telephony as an application, and instead just see it as a feature of a bigger communications framework.

In my unhumble opinion, one of the biggest unsolved problems in "IP communications" (or whatever we call it) is how people and corporations interact and transact. To go buzzwordish, Skype is the seed from which the ultimate B2C or C2B multi-modal interactive transaction engine can be built. Google has the cash and commercial relationships to do it, but lacks the user base to grow it from. Skype is one quick way to access that user base; Google's actions suggest they believe they can out-distribute and out-market Skype.

Google's business model

Here's Google's situation. Google isn't a search engine company; it brokers connections between people and corporations for profit. The media properties provide context for the relationships. Search is just one, albeit dominant. Orkut, despite its fizz-bang-pfizzzzzzaway, is sort of the social setworking equivalent of Skype's on-net calling. It's always going to be a free service, and just providers more contextual fodder. And so on through photos, books, maps, etc. — all just fodder for connection people and corporations for fun and profit.

I'm just doing a precis of what other observers have written, no new insights. Move along, please.

Note that broking of connections and relationships is more than just advertising. Google is more than adverts. Otherwise, why not just make all AdSense ads non-clickable? That's what adverts are — calls to action without an integrated fulfillment mechanism. As such, "on-line advertising" is an oxymoron.

The e-commerce value chain

Google is competing in a long transaction value chain that looks something like this:

• Demand stimulation/market formation. Self-awareness of user need, awareness of market solution. The domain of traditional marketing.
• Capture attention. The user is presented with ads, and eventually sees a proposed solution to a problem the user has. In the user's mind, the connection is ready to be made. In the olde world of directories, this is (i) finding the category of vendors who match your problem (often somehting that isn't intuitive if you're after something more complex than a taxi or flowers), and (ii) filtering on the locality/capability criteria you have. By the end of this stage the user feels "I am aware of a potentially relevant solution to my acknowledged problem".
• Connection. The user clicks on a link. The connection is only one-way; the advertiser doesn't know who the user is, or what they really want. An extended Yellow Pages advert is the analogue version of connection. The user is now engaged with a particular soltion provider and is paying attention to their message.
• Contact. The user and advertiser engage in bi-directional contact. The user presents some form of identity (e.g. gives a phone #, e-mail address, etc.). This is like calling the 800 number.
• Transact. The user's requirements are codified, and a non-repudiable contract is formed to deliver some good or service.
• Settle. Payment is remitted. A third party like a bank or Paypal is normally involved.
• Fulfillment. The goods are despatched.
• Delivery. The goods arrive.

Google's competitors aren't search engines per se. Google is competing for transaction value chain slices against eBay, Amazon and even vertical search like Craigslist. Of course, chop off the search engine leg today, and the Google animal as a one-trick pony falls over. But Skype could equally be another leg on the Google animal.

Structuring transactions

Amazon and eBay have their own twists on transaction enablement. The former is pretty full-spectrum, with the web site recommendation engine doing the marketing on behalf of the product vendors, all the way to actual despatch of the goods — and even delivery for some digital goods. eBay integrates payment and settlement (Paypal), but is weak at the capture stage because you need to go there first to initiate a search. There's no "did you know someone can solve this problem for you?" eBay sidebar as you browse around. On the other hand, eBay has memory of your identity and reputation; Google just delivers anonymous bodies to the shop door and disengages from the transaction entirely. That's a strange way to treat customers, not even attempting to make an introduction. Impolite, at the very least.

Both eBay and Amazon are able to deal with a relatively narrow range of possible commercial transactions, huge as the volumes may be within that range. Simple stuff like "I want a red one" in Amazon and eBay means adding another SKU. Every nail is hit by the "add another inventory item" hammer. Whilst there are platform APIs, the ability to define new forms of interaction with the user are very constrained by the existing business processes and transactions.

So Google = totally, unstructured transactions with no integation of user identity; eBay and Amazon = structured transactions, with limited flexibility, and some user identity (but isolated within their commerce island).

The future transactional user exterience

Fast-forward five or ten years. What sort of transaction and commerce experience do you expect to have? Picture yourself buying a vacation, talking to an agent. The agent is co-browsing with you; perhaps you would like this hotel? Or that one — it's a bit nearer the beach, for just a little more money? You would never have to dictate basic personal data. The agent would know you are indeed who you claim to be. The agent might not, however, get any of your personal data, or even phone number. They can contact you, but your identitiy is opaque, and if they spam you later, you can cut them off easily.

More succintly put, this screen is a big space that screams "fill me with interactivity":

(Yes, even test calls should be interactive…)

Goodbye IVR systems, hello rich media, up-sells, rapid navigation, and ease of use. The future of VoIP is more than cool icons and graphic design.

It isn't hard to imagine more and richer scenarios of "beyond the traditional phone call". And as Russel Beattie notes, we have the technology to do this right now by using extensible presence protocols. Why can't the agent push a web form to me to view in my VoIP client?

Skype and Google: the yin and the yang

So, why Skype and Google? What do they have to offer to each other?

Skype's weakness is that it is so PC-centric, and the PC is a lousy telephone. Check out the spoofish Engadget from 1985 page and scroll down to the old luggable mobile phones. We may laugh, but that's about the size and shape of my laptop, which today is my primary communications device.

The PC is the ultimate chameleon device. USB makes the form factor and sensory aspects extensible. A general-purpose CPU, screen and keyboard cater for a huge variety of possible interactions. It is the best place for the primodial swamp of next-generation communications apps to quickly iterate, mate, spawn and die.

The PC is also Google's home. Whilst both Skype and Google share an ambition to "go mobile", it's a long and fraught journey. Neither will make much impact in the short term.

Remember how Google's delivery of customers to vendors is almost completely unstructured. This contrasted with eBay and Amazon where the marketplace is highly structured in advance. A phone call is also highly unstructured today. Google is naturally placed to capture this high-end of the market. Leave eBay to worry about highly commoditised simple transactions; Amazon to take the mid-market where more complex transactions exist. Google is particularly suited to retail of complex, niche, high-value goods and services.

It would be interesting to compare the average eBay sale, average Amazon sale, and average Google-brokered transaction.

So the attraction of Skype to Google is time to market. Rather than waste time repeating basic communications stuff, just buy into an existing technology and user base where most of the bugs have been ironed out. Why take on the execution risk of something outside your core domain?

The smart, interactive VoIP client is also a green field that is easier for Google to shape in its own image. Trying to interpose Google-mediated transactions into existing web sites is probably a non-starter.

There are serious challenges with integrating existing automated call handling systems, and a Google-Skype alliance would suggest some deep conversations with the Avayas and Nortels of the world. But if they won't play along, jusy leapfrog them, and leave them marooned in the isolated archipelago of traditional telephony.

All that said, Skype needs Google a lot more than Google needs Skype.

Skype's strategic quandries

Chris Drury suggests:

The Skype business plan is to:

1. build a huge captive island of users by leveraging the arbitrage value of free calling around the world, and
2. charge users to connect to other islands — PSTN, SMS, IM, email — and for premium services to be determined.

Well, kinda. Skype is off doing a whole raft of stuff.

There are three directions in which the Skype business model can develop:

• Down. They're burning bizdev dollars on trying to sell connectivity with service, through things like SkypeZones. The upper layers of the stack get to support the lower ones — the ultimate "downsell". But why get tangled in the turmoil of connectivity, where players are either gasping for air in commodity markets, or entrenched by political fiat and don't need you?

• Across. Sell bridges from one application island to the next. But the biggest opportunity is passing them by — getting call termination revenue by using existing mobile numbers as SkypeIn numbers. Skype should be the landline displacement play of mobile operators, but isn't. Another missed opportunity. What they also aren't doing is working out ways of charging people other than the users for getting on and off the Skype island. That's what Google does, and extraordinarily well.

• Up. Bridge to the invisible "finance" layer that floats above the application layer in the protocol stack. That's what transaction integration does for you. Make the money wind blow harder across your island.

What you can't do is stay on your island. There's no money in telephony or IM within your island. Everyone can afford the ferry fare to leave.

Skype is both an extraordinary success and inspiration, as well as a frustration and disappointment. The Mongols may have swept across and subdued the civilised world on the back of a few horses, but that's the exception rather than the rule. Most invasions require big resources and manpower. Skype's limited resources are too diffused. Is a Skype toolbar really the biggest strategic imperative, something that cannot be done by a third party? An in-house video solution? Yet another web presence server?

I've seen this story before, and it doesn't have a pretty ending. Informix thought that a faster, sleeker database would win. Oracle crushed Informix by building a better sales and distribution network. Microsoft out-distributed the world by bundling with PC sales. Goodbye Borland, Netscape, WordPerfect et al as tier-one players. Yeah, limp on for years, but don't be a threat.

Telecom games are always won by the person with the stronger distribution strategy, not the biggest feature set. Skype is losing focus. When you've got really limited resources, you need exceptional alignment to strategy. Does Skype have a single product planning strategy professional on the staff? Anyone who can operate the tools of the trade? When you're presented with such mind-boggling choices of which way to turn, no one mind can hope to capture and process it all.

Google has the dream distribution network between consumers and businesses that Skype needs. Google has massive resources, enough to absorb some mistakes. Skype should have sold out when they could. Maybe Google are releasing VoIP clients to just depress the price they have to pay. But it's a dangerous game of chicken.

Looking forwards

Yahoo is a media company, and is unlikely to be the commerce bridge. eBay is a real threat to Google, and eBay buying Skype would be a setback for Google. It isn't hard to see eBay aligning with, say, Ask Jeeves and using all the Interactive Corp. properties as seeds for an integrated search and transaction experience. Amazon is a similar story. Microsoft has execution problems of its own, but knows what's at stake and has boundless cash and armies of developers to throw at it. Google's aura of invincibility is largely hubris. They need to diversify up the transaction chain.

Five years from now is quite a long time. Remember we're only celebrating 10 years of public e-commerce. I still believe that there will shortly be hundreds of millions of people making "phone calls" on the stupid network, doing things that were impossible on the old telecom network. But whether they'll be doing it on Skype is becoming less likely. Skype needs someone with strategic discipline to whip things into shape. Stop duplicating what your developer program partners are doing. Stop anything that isn't going to make you the best in the world at what you do, a market of one.

This is not a rehearsal

The prize is potentially very big. Think of the sort of commercial power Visa and the banks have today, and the huge profit they shake off as a result. They're vulnerable as intermediaries in the new order. The stock prices during the .com boom for Amazon and eBay were only crazy in terms of timing, not magnitude. Some of the players need to grow up fast to meet the challenge.

UPDATE: Just to be explicit, the "down" strategy leads to a small pot of gold because nobody has much to share, "across" to a medium-sized one, and "up" to a sodding large one. Down is limited by the size of the connectivity market; across by the communications application market; up by the size of the economy as a whole.