Tim Oren's Due Diligence

More on Wi-Fi hotspot business

More on Wi-Fi hotspot business models

David Sifry himself takes notice of the posts on hotspot subsidies by Don Park and myself, and responds with an illuminating discussion of the mutations of Sputnik's business model. Bottom line: David is happy to let Don risk his own money. Now that's evangelism!

Update: Here's a local broadband ISP that's trying to grow a hotspot network by crediting DSL subs who put up hotspots with 50% of the revenue derived, leveraging the customer's capital investment. Viability still unproven, but it's being done at minimal cost. An experiment to watch.

Peek-a-boo! From Sharp minds come

Peek-a-boo!

From Sharp minds come tiny little products. At Kokoro, the world's smallest megapixel image capture module. Coming to a place or device near you, that you won't notice, real soon now.

Eric Raymond unloads on SCO

Eric Raymond unloads on SCO CEO Darl McBride

Who bloody well deserves it. SCO is in a FUD fight it can't win.

Wi-Fi AP subsidies by carriers?

Wi-Fi AP subsidies by carriers?

Don Park speculates that sales of Wi-Fi APs could be accelerated if there were subsidies from carriers to buyers who deploy the APs as public hotspots. Since the idea was kicked off by a discussion with David Sifry, his notion is stated in terms of Sputnik as the carrier.

There's partial precedent for such an arrangement. Back when I was at CompuServe, we and the other online services and major ISPs had subsidy deals with the larger modem and PC manufacturers. If a buyer of one of their devices signed up for our service and stayed on for a few months, the manufacturer would get a modest kick-back from us. While I was never privy to the financial models of the modem or PC guys, one assumes that some portion of that spiff ended up being used as a weapon in price competition, and therefore ended up reducing the ASP to the end user. What's more novel here is the idea of direct subsidy to the purchaser, though that shouldn't be any logistical difficulty, given the number of rebate-on-activation type offers already out there.

That much said, I don't think the Wi-Fi market is mature enough for such a model to work as yet, and there are also some economic problems that may be inherent. Remember that when I and others were doing the modem spiffs, they were a competitive weapon to gain share in a fast growth market - dial up - that had a reasonably understood consumer value and business model. None of these qualifiers pertain in Wi-Fi hotspotting, as yet:

Wi-Fi overall is past the knee-in-the-curve and in fast growth, but the same can't be said of hot-spotting. Anecdotally, the average usage per Starbucks site is three sessions per day. The stable business model is not known as yet - it may be straight subscription, add-on subscriber feature, pay-per-use, free or discounted location amenity or (likely) some mix of those. No one in any of those models is yet profitable, meaning that a subsidy is just a way of going further into the red.

The AP price is a factor in the initial setup of a public hotspot, but labor costs of the install likely equal or exceed it, unless the location owner is technically sophisticated. On an ongoing basis, the costs will be dominated by maintenance (truck rolls), but even more by backhaul costs. This is a hidden and problematic difference between the startup of the dial-up ISP and hotspot models. Dial-up access piggy-backs onto the PSTN to get the users to a concentration point, where modem racks handle conversion to an IP backhaul, initially a T-1. When starting up a dial node, that piggy-backing assured that you could get a fast start on amortizing both the hardware and backhaul costs, assuming some reasonable subscriber base within the served area of the CO.

In contrast, the hotspot can use a cheaper DSL backhaul, but has much less assurance that it will be able to amortize its costs among users as the business starts up: Desktop PCs with modems had the nice characteristic of staying largely fixed with a CO serving area; Wi-Fi'ed laptops and PDAs sprout feet and walk away. The served area of a hotpot is inherently less than that of a fixed line CO. As we learn more about user behaviors, it may be possible to build models for viable concentrations of hotspots within an area (sort of like the video store, gas station, and pizza parlor siting models), and discover what degree of end user switching costs can be created by hotspot owners and carriers, but we're far from there as yet. This is an inherently more complex business model, and assessing the actual value to the carrier of a subsidy would be very hard.

So, recapping, subsidies are a weapon for carriers to buy market share in a period of fast growth. Lacking this growth, and a confidence that new sites will be profitable, a subsidy model isn't likely.

Kevin Starr is not in

Kevin Starr is not in a funk

Nothing like a historian with a long view of things to counteract the blahs induced by a worms-eye view of a down cycle. Check out Scott Loftesness' post on the perspectives of California historian Kevin Starr, and feel better about the Valley, the business and the state.

Tim and Ole talk Microsoft

Tim and Ole talk Microsoft strategy

After my post on the 3 Laws of Microsoft, Ole Eichhorn replied with two counter-examples [scroll to bottom of the post] where Microsoft appeared to abandon a market. An e-mail discussion ensued, which we agreed might also be interesting to our vast readerships. So, edited for continuity, we bring you Tim and Ole's fireside Microsoft chat:

[TO: The 3 Laws of Microsoft reprised: (1) You can always see them coming. (2) They never get it right the first time. (3) They never go away (unless the market is proven not to exist).]

OE: There's actually at least two counter-examples to (3): tax preparation software (they tried with TaxSaver and couldn't dent Intuit's share, so theyægave up ), and bill payment processing (they tried with Transpoint and couldn't dent CheckFree's share, so they gave up).

TO: You're certainly correct on the facts. However, I've always looked upon those products as being not an end in themselves, but rather tactics in an overarching Microsoft strategy: to get themselves into the financial transactions/intermediary foodchain, one way or another. If you look at it that way, there was also the STT initiative in the mid-90s, which was parried by the bankcard associations and turned into the equally abortive SET standard, and the more recent Hailstorm consumer side directory portion of .NET, which seems to be still-born. The one that has achieved some nominal success is Passport, though I would doubt that they are yet able to generate much leverage toward a real food chain position out of it. But I'd still argue that 2) they certainly didn't get it right the first (few) times, and 3) they haven't gone away, if you buy my Gates/Ballmer level grand strategy interpretation.

OE: Yeah, I agree that MS has been trying to get into financial services by hook or by crook.æThey really haven't given up, have they? Buying Great Plains seems to be their best effort so far, although that's accounting software and not really financial services.æPerhaps they'll leverage the vendor relationships GP has with businesses to offer payroll, tax, insurance, etc. like Intuit does for QuickBooks customers.æThat's become a great business for Intuit.æI really think the only reason they still offer Money is to get consumers into MSN and related services, a kind of loss leader.

TO: I read Great Plains as being more strategic in the server wars vs. Linux and Sun. I've used Quickbooks for quite a few years (our fund's on it, for heaven's sake!), but it does run out of gas when you have to go to a multiple user situation. I think MS wanted to position to catch businesses at that scale up point, with a clean all-in-one-box accounting solution that would then direct the customer toward using more of their stuff for analytics, HR, portals, etc. If you tried to do that on Linux (say), you'd be piecing things together. So I see it as a move to consolidate the CFO/COO functionality area in SMBs, being one of the battlegrounds of the larger server platform war, where MS has its hands full.

OE: We see things much the same.æI bet you're right about the GreatPlains acquisition and Sun and Linux.æ(And IBM) BTW, Steve Bennett is trying to move QuickBooks upstream as fast as possible; note their new 'enterprise solutions'.æMS better watch out for a Christensenian 'attack from below'!æI can see it now, QuickBooks Venture Fund Edition :)

OE: [But] Passport has zero traction for financial services; people use it for single sign on, but they don't use it as a wallet.æ(Really no client-side wallet has ever made it, for whatever reason.)

TO: Agreed, and agreed. I still remember when I was running CompuServe's ISP and CyberCash came around and suggested that I should pay them for the privilege of distributing their wallet, since it was obviously going to attract a lot of users. Heh! That one didn't happen.

OE: I remember CyberCash, too.æI was at Digital Insight at the time, and we had a similar offer.æWe actually did distribute the CyberCash wallet for about four months ($0 both ways), but since only about seven people signed up we dropped it. æMy sense is that there is an opportunity in there somewhere for the right product.æMaybe Google will give it a shot (as an extension of their toolbar, which seems to get more functionality all the time).

A requiem for virtual reality

A requiem for virtual reality

In a short Wired interview of Neal Stephenson, he reflects on Snow Crash, his first novel that became the ur-myth for the virtual reality investment craze of 1996-8:

For the most part, Snow Crash turned out to be a failed prediction. People have shown limited interest in immersive 3-D technology, so I think it worked better as a novel than as a prognostication. But it provided a reasonable, coherent picture of a particular kind of entertainment technology.

Stick-your-head-in-it VR has turned out to be a niche market, valuable in some applications such as training and modeling. Those who bet on it as a horizontal market, for entertainment, shopping or what have you, all died on the road. Remember VRML, the next big standard? Paragraph, Intervista, and Newfire? Mark Pesce and Tony Parisi? Even the avatar modeling tools have been reduced to gimmicks for selling clothing. A cautionary tale on taking a compelling, coherent yarn as an argument for a real market.

What's actually happening is in many ways more interesting. Rather than substituting a synthetic reality for the actual thing, information is becoming co-extensive with reality. Now it's mobile data and other gadgets, soon it becomes networks of sensors and information tags. Some have tried to gin this up into a ubiquitous 'augmented reality' that will happen Real Soon Now, often with visions of (again) data glasses that will overlay the virtual on the real. Take that with large grains of NaCl. Economic value is not evenly distributed in real space, and ROI will drive deployment. Check out the malls and warehouses first, long before you find sensors hanging on random trees. As for those HUD goggles, see the lesson above.

Myths about mobile phone capacity

Myths about mobile phone capacity

Every once in a while the Prof blows it, and this post about reserve mobile phone capacity is one such. Fortunately, before I started typing out a long rebuttal, I found that Steven Den Beste has done a much better job than I could ever manage, since he's actually a qualified CDMA designer. Bottom line is the same though: TANSTAAFL. You want reserve capacity in the system, then you pay capital costs up front, and they have to be cranked into the rate base, just like power utilities. With the difference that the requirement will have differing effects on competing carriers, depending on which technology they use: Den Beste follows up with a short primer on CDMA and other mobile voice/data systems. Bottom line: Qualcomm would end up a winner if reserve capacity were enforced, as CDMA's greater spectral efficiency would advantage carriers using it over others.

Non sequiter: Sandwiched between those two articles is Den Beste's glowing review of Miyazaki's animated movie Spirited Away. All I can do is whole-heartedly concur. If you haven't seen this yet, just go buy it. Trust us. And if Japan hasn't yet declared Miyazaki a national treasure, then someone's blown it.

Support the California economy -

Support the California economy - drink more red wine

It's good for you.

Richard Bennett is in a

Richard Bennett is in a funk, too

But unlike Jim Clark, he's bailing out to Oregon. Well, have fun, and remember to rotate your seating position periodically so the moss builds up evenly...