Books I have co-authored

Why did the vision for public W-Lan networks fail?

Today it is obvious that business models that are based on public Wlan networks will never live up to the hype from 2002-2004. The dramatic price pressure for mobile data via 3G/HSPA undermined the market for Wlan operators. The choice is easy between spotty Wlan coverage and comprehensive geographical service with a 3G dongle.

Other players within the industry can learn several important lessons from the relative failure of this business model. If we go back to 1999-2002 we see that there was enthusiastic interest in Wlan and the possibility to create a disruptive innovation that could marginalize the traditional mobile operators.

The core of this hype was a simple analysis where the capacity/cost ratio for a Wlan access point was compared to a large 3G base station. A Wlan base station cost over 1,000 Euros and had the ability to transfer 5.5 Mbit per second. A large 3G mast cost 100,000 Euros and had a capacity of 4.5 Mbit per second. This capacity had to be shared by all the users within the mast’s area of coverage which meant that two surfers who each streamed 2.25 Mbit per second consumed the base station’s entire capacity. The cost advantage was dramatic. 3G was 100 times (10,000%) more expensive.

However, this is an oversimplified calculation: 3G micro base stations do not cost 100,000 Euros. When a 3G network is purchased there is built-in support for functionality such as roaming, handover, radio planning, access control, billing, etc. Wlan operators have to pay to develop solutions for this when their networks expand and their customer base grows.

Another factor is that Wlan base stations only cover an area within a radius of 30 (indoors) to 140 meters (outdoors). Geographical Wlan coverage requires many access points. But even if one adds the costs for the core functionality of the network as well as the cost of deploying a large number of Wlan access points, the cost advantage in favor of Wlan remained (even though the cost advantage was significantly lower than a factor of 100). In addition, Wlan networks have a higher capacity per square meter than 3G.

That’s how things stood seven years ago. Wlan was popular among the IT community, academics and VC investors. Wlan was embraced in a similar way as Linux and Open Source. It was only telecom operators and Ericsson/Nokia/Siemens who were upset when everyone criticized their 3G plans and the sky high prices they paid at the 3G auctions. I have to admit that I was also a Wlan enthusiast at that time.


What happened thereafter was that the competitors (the telecom operators and Ericsson et al.) responded to the threat, development took an unexpected turn and the Wlan operators shot themselves in the foot.

If one read between the lines about how Ericsson and the others in 3GPP reasoned before the Wlan threat became serious, the vendors’ plan was to begin by selling basic 3G networks at full price. The next phase was to sell capacity upgrades over several years at full price and finally begin marketing 4G (“LTE”) well after 2010.

When the Wlan threat appeared the telecom vendors were forced to rush their product development as well as market their upgrades such as HSPA earlier than they had planned. When the indebted operators were to be convinced to order 3G networks during the difficult period from 2002-2004, price pressure resulted in even lower prices. All of this eroded parts of the Wlan operators’ cost and capacity advantage.

Another factor that unexpectedly contributed to the undermining of the Wlan business model was the unused capacity in the finished 3G networks after 2006. The customers slow transition to 3G and the problems for pure 3G operators (such as Hutchisons 3) to acquire customers lead to desperate operators who were stuck with newly built networks without enough traffic. When 3 began to market fixed price 3G dongles they redefined the market with an offer that many other mobile operators were forced to copy. The 3G dongle became stiff competition for Wlan. Lower bandwidth, but much better geographical coverage.

It was actually the Wlan operators themselves who sabotaged their own market. By continuing to have unreasonable prices and not offering attractive roaming packages they missed the window of opportunity that existed before the 3G networks were fully up and running.

What originally made the concept of Wlan so exciting was the vision of a small-scale network deployment where the initial costs were minimal. If all of these networks had worked together and used one clearing house that was in charge of roaming, the users could be offered a large aggregated “network” with good coverage. Unfortunately, this did not take place.

The prices were set far too high and the operators did not ensure that the roaming contracts were reasonably priced. Considering the spotty coverage and the fact that few Wlan networks provided overlapping coverage, the operators should of course have created a single roaming pool. But they didn’t. They didn’t even understand that they should take advantage of the market for pre-paid cash cards. The prices were set for business users who were forced to access the service and login was made unnecessarily difficult.

Despite the fact that the Wlan industry was unable to live up to its promises and it will probably prove to be a wasted investment, we should all be grateful for the attention it has received over the last nine years. Wlan (and its cousin WiMax) have put pressure on Ericsson and the traditional telecom operators. This has increased the pace of innovation. The threat from Wlan operators will prevent greedy telcos from inflating their prices. The total market for all types of Wlan products has grown thanks to Wlan operators, and this has pushed the prices down.

This article has previously been published on my Swedish blog.

1 comment to Why did the vision for public W-Lan networks fail?

  • kevin

    in the future everyone will end up using wi-fi because someone will create a system that has a duel-feature long range or high capacity. as far as costs go, wi-fi is cheapest in terms of $ per bit of bandwidth. that’s what everything will come down to.

    the business model is there. if you can cover 100 users and charge them $10 per month then that’s $1000 per month revenue. The dsl / cable line is only $50 per month.

    the problem left is technical. lots of multipath and fading decreases range and capacity. a proper system will use these features to it’s advantage.