Bluesocket has been a puzzle to us for five years. It had a highly regarded range of wireless LAN gateways, but in 2003/4 it lagged on the move to wireless switches. Since then, it's popped up occasionally, telling us it has a technology lead, and reporting positives moves in Gartnerspace. But real evidence of its performance has been scarce.

Our readers are clearly also puzzled - a recent blog post sparked a continuing burst of queries. We asked why the company sold Pingtel, the open-source PBX company which it only bought it last year?

Readers tell us the company has downsized drastically, and closed offices abroad - but these could just be people with some axe to grind.

So we are pleased to have a chance to hear some explanations from CEO Mads Lillelund, in a phone interview.

Still winning customers?

As a preliminary, he defends Bluseocket's customers and products. The company started with a security gateway that handled standalone access points, he says, but: "as the market changed, thin access points became the defacto standard. In 2005/6 Bluesocket developed into a full solution wireless LAN company, and built up 2,600 customers in the WLAN area" (a December 2007 release claims more than 3000).

The company's strength is in education (about 1000 customers), enterprise (another 1000 plus) and healthcare (around 450) he says, with the remainder of its customers in smaller vertical markets. The hospitality sector is growing, he says: "We're now having an opportunity to work with world class systems integrators, in the hotel and convention area."

"It's a strong portfolio of customers," he says. "We generate about 44 percent of our quarterly revenue from new customers and the rest from the existing customer base."

Leading in 802.11n?

The company started shipping an 802.11n fast Wi-Fi access points in May, he tells me.
That's later than other vendors, I point out.

Well, he counters, most vendors are using the same [Atheros] silicon, and the other vendors must have been selling APs based on early spins of the silicon, he says. "If they were selling earlier than us, they had to be on non-GA [general availability] chips."

I don't believe this - and after the interview, I check, and find Aruba's claim to have sold 10,000 Atheros-based 802.11n APs before May. (Cisco claims 50,000,but that is on earlier Marvell silicon).

According to his contacts in the supply chain, 802.11n has grown fast, to make up about 15 percent of the APs shipped by vendors, he estimates, and "that mirrors our sales, except there are very large deals in the pipeline that would swing this to a higher percentage of sales."

"I say that most customers would buy 802.11n APs, if it wasn't for the price," he says, claiming to beat the price of other enterprise 802.11n APs by around $100 to $200.

Bluesocket's AP is also one of the very few that can do full 3x3 MIMO on both Wi-Fi spectrum bands at the same time, within the limits of 802.3af Power over Ethernet he says: "We are not throttling down in any way, shape or form." The claim is made in detail in a position paper.

One other vendor claims a thin enterprise AP working at full stretch on standard PoE - but Siemens has had its claim verified.

The achievement comes form Bluesocket's experience, launching the first enterprise access point to use the MIMO technology which forms part of 802.11n. "Those APs already had 133Mbit/s throughput," says Lillelund. "We had to make the leap from 133Mbit/s to N. That helped the engineering team."

So why sell Pingtel?

With all that going on, we ask, why has the company laid off staff, and closed offices abroad? And why sell off Pingtel which Lillelund said was a strategic acquisition a year ago?

"We're focussing on the bottom line of company," Lillelund tells us. "We've been in the market since 1999, and we want to be there going forward. We want to ensure there is enough cash in the company to ensure a long term future for our employees."

He digresses on product strategy, explaining that wireless LAN is an overlay on the LAN, but that will change: "In the next twelve months, it will become more integrated. Fifteen percent of all the RFQs [Requests for Quotation] in the data networking space are already being consolidated to include LAN switches and wireless. Wireless is a feature function of the data network environment."

Given that, Bluesocket is moving towards software, rather than bigger switches: "Other vendors are building larger switches - a 20Gbit/s, 40Gbit/s or 80Gbit/s box - but that's not sustainable. We embarked on a software approach two years ago - going into a hardware race is not something I think is productive. We want to modularise the software, so it can start to reside in certain areas of the network."

The PingTel acquisition was part of that, he says: "We looked at some companies to go after in the early part of 2007; we settled on Pingtel. We're not shy about doing acquisitions. We saw some opportunities. With Pingtel inside the company, we were able to get an understanding of how voice operated. They had some very interesting customers. It was a good learning experience."

But if so, why sell it on, to Nortel?

At first he prefers to talk about the how, not the why. "All the people who came in, the whole company was absorbed into Nortel," He won't tell me how many either; he won't tell me whether PingTel grew or shrank.

Where did Blueocket's FMC plans go?

But we push him a bit. A year ago, Bluesocket was going to use PingTel to build converged voice services, and become a fixed-mobile convergence (FMC) player, wasn't it? Whatever happened to that?

"That was a side element to the acquisition," he responds, eventually. "That wasn't the reasoning behind it. We saw fixed-mobile convergence, but not as a major driver of wireless LANs. We see other applications as being at the forefront."

"Pingtel never stated they were an FMC company, but there were elements that you could use to develop something there," he said. "There was an element there that could be capitalised on at some point. if the FMC market should take off, it was something that Bluesocket could capitalise on - but we never really saw it move that quickly."

So FMC didn't boom the way it was supposed to, we ask, and you abandoned it?

No, he says. There never was an FMC plan. The way he tells it, Bluesocket only ever wanted to have PingTel in-house for a while, as a learning experience: "We did a lot of cross training, a lot of learning in terms of the whole voice architecture. It's been very good for the last twelve months."

I tell him that doesn't really square with his comments on last year's acquisition press release. What he said then was: "The opportunity lies in the fact that the wireless LAN and unified communications markets are converging. As a result of this acquisition, we are better positioned to deliver an integrated solution that meets the mobility needs of customers across a variety of industries."

I think "delivering an integrated solution" of the "wireless LAN and unified communications markets" sounds like an FMC plan, but he tells me I was reading too much into the announcement, and we let it go at that.

So if Bluesocket planned to sell PingTel off, how did the money work out? Did Bluesocket make or lose money on the two transactions?

"You are making an assumption there," he says, and zips his lip.

What does he mean by that? Are our readers right to call it a "roll-up" by venture capitalists who backed both companies? Was there any money at all involved in the PingTel purchase?

"I can't comment on what outside people may call it," he says. "We have several investors, and one common investor on both sides. The investor base of Pingtel is different except for one investor with a stake in both companies. That made it easier to put the two companies together."

So no answer then,

Staff cuts or a re-focus?

What about Bluesocket's shrinking staff? Readers have sent anecdotes about a reduction in Bluesocket's headcount.

"We have trimmed in certain areas," says Lillelund.

Can he give us any figures? Is it as low as the 40 people readers have claimed? "It would be higher than that."

How many employees has Bluesocket lost since this time last year? "We normally don't comment on employee levels."

What about office closures? While preparing this article, I have rung a few of the nine international numbers on Bluesocket's website, and got no response. By the time the article is finished there are three numbers in the US, the UK and China (the rest are there in Google's cache).

It's re-focussing, he says."In North America, the number of feet on the street is up - and we make 75 percent of our revenue there."

It sounds like the company has shifted almost completely dropped direct sales: "We generate 85-90 percent of our revenue through distributors and resellers," says Lillelund. "Most of our direct sales are legacy accounts. We went two-tier in 2005, and since then have signed up distributors."

Overall, the company prefers resellers to direct sales, as they get more coverage, he says: "Granted we are a small company, if we can get to the table we have a very strong technology proposition."

It's a long and interesting conversation. He doesn't answer a lot of my questions, but I don't expect him to.

I come away with a picture of a company keen to make the best of what it has. But anyone planning to buy Bluesocket kit - or the company itself - is going to have to probe a bit deeper.