WTRS Executive Interview

Interview with Adrian Tuck, CEO of Tendril Networks.

April 24, 2007

George:
Can you give us a bit of background on yourself?

Adrian:
I'm British, in case you hadn't noticed. I was educated at Sandhurst, which is the British equivalent of West Point. I was a British Army Officer for 7 years, and served in Desert Storm and a number of other locations, so I had an exciting early twenties, and then I left. Suddenly I woke up one day and decided that with the way the trend was going, the excitement was going to go down and the admin was going to go up and I should find something else to do.

So I left, and I sort of made myself this list of the things I liked about the army and things I didn't like, and ended up feeling that I wanted to find something that was a meritocracy and creative and a bit of an adventure.

Both my parents had been school teachers, I'd never been in business in my life, and I knew nothing about it at all. I interviewed in a number of places in London, and did reasonably well during the interview process and got offered a job at a big bank, and a big advertising agency. Then I met these two guys who had founded a new company and it all sounded so exciting. I was so ignorant about how hard it was going to be in a startup. To the disgust of my parents I took the startup job.

The name of the company was Connect, it was a managed services business, managing email and voice messaging for enterprise clients. It didn't take any venture money and grew through an incredibly rocky road, but managed to grow into a viable, profitable business. I learned so much during that period. These guys were ex management consultants, excellent with PowerPoint and very visionary; I was sort of helping where the rubber hit the road. It was a fantastic learning experience for me and I was there for 5 or 6 years and saw the company grow dramatically during that period. So that was my grounding in startups and I realized that the ability to effect change was the thing I should look for in my career and what I now look for in a role.

During the Connect process I met up with a software company in Dublin who made software to manage telephony billing. I felt their business model wasn't quite right and I was just starting to get my confidence in this kind of stuff and the founder and CEO said - "well why don't you come in and help us change it?"

That company was Softech, based in Dublin, and I came in and ran the US operation. That was my introduction to the US. We were in California for a bit, then moved to Boston. That was an exciting turnaround opportunity for a software company in a fairly tough time when the telecom market was crashing.

At one point whilst at Softech, we were trying to raise money. As part of that process, I went to Polaris Venture Partners in Boston. One of the partners at Polaris called me after I did the pitch and said "we're going to pass on the company." "But we have this opportunity and think it's kind of interesting and we thought you had the enthusiasm and the skills - why don't you come and have a look at it?"

"Basically its two guys, engineers out of MIT and we think they have something but we just don't know what it is. We feel it's going to be big.". So they said to me, come in, help write the business plan, raise the funds, recruit a CEO, and basically figure out what the company is going to be." As part of the process, I met Bob Metcalfe, I found out he was involved in Ember, and you know, you can't turn down opportunities like that too often.

I found that Rob Poor and Andy Wheeler were a couple of really bright guys. I read Rob's PhD thesis and I listened to him talk and my mind was full of ideas about where the business could go (most of them wrong). I joined Ember as the first business guy, with the job to write the business plan, the go-to-market strategy and all those sorts of things. The idea was always envisioned that it would be sort of an 18 month gig during which we would hire in the right team and then I would go onto something else. But I'm still in this fight, so I'm clearly intrigued by it. Or no one else would take me.

We looked at what we had, which was an algorithm that allowed mesh networks to self organize and work efficiently. There were other algorithms around, stuff from the Tiny OS community and a bunch of fairly heavy weight versions of this kind of networking that were used in higher bandwidth communications. We had in front of us all these markets that other people had looked at; there was a company called Mesh Networks that got acquired by Motorola (I think). They looked at things like First Responders, I think their original plan was alternative cell phone infrastructure networks and so on. We decided what we had was a very, very lightweight algorithm; very small so it could fit in tiny processors, and that meant that we could go into the networking of things that had never been networked before. So we decided on the whole "Internet of Things".
My recollection of sequencing here may be a bit hazy but ZigBee started to emerge at around the same time. For a while we had a sort of ambivalent attitude towards the standard. When you have someone like Metcalfe on the Board the whole gospel is "Standards always win". We knew that we were going to need to embrace ZigBee, and he certainly encouraged that at every turn. So we made the decision to embrace the standard and some people will argue that it cost us two years, but I still think it was the right thing to have done. It's playing out well now, the standard is becoming robust and Ember has a leadership position within that. There were times when you look at things like Zensys, and it looked like they were getting some traction. We occasionally wondered could we have done the same if we'd stayed outside of the quagmire that is the standards world. By the way, did you see the Zensys whitepaper on ZigBee? The last sting of a dying wasp.

George:
My comment there was that they are about right at the time where they have historically gotten another round of funding. They are kind of in this 24 to 30 month cycle. And it doesn't look like they are getting funding right now. And then they threw that whitepaper out.

Adrian:
I read it for the first time this morning. It basically implies that anyone who would ever use this technology, let alone ship millions of things with it in, is an idiot. It's very 'courageous' of a company that ships a proprietary product with only one channel to criticize a spread spectrum technology like ZigBee that has 16 channels.

George:
Who is the guy that is going to call up Eric Smith (of Control4) and say, "Eric, your product doesn't work."

Adrian:
There have been times when I've woken up in the night and asked, did we do the right thing? Ember contributed a reasonable amount of IP into this process and we had conscious discussion about, do we sit back and see how this thing goes or should we step in and get involved. We made the decision to get in and for the most part don't regret it, I definitely think it's a bet that's paying off.

So that's how I got into the Ember world, and in the middle there we recruited Jeff Grammer as the CEO and made the transition to the chip business, and again it was the intent, for me to do that hand off and go do something else. But I couldn't tear myself away, I'm always intrigued by all the opportunities here.

The reason for switching to Tendril was that I recognized that the sort of first tier of this battle was becoming won. ZigBee was establishing itself, 802.15.4 was definitely establishing itself with multiple vendors, and the underlying technology was basically reaching the price points that made it a viable solution. I had watched Ember's first customers struggle dramatically with something that has nothing to do with what Ember did, which was to build an end-to-end system out this technology and I just felt, wow- there's a yawning gap between where everybody normally finishes and where ZigBee devkits start and there is nothing in the middle there and the customers are sort of meandering and having to build things themselves. I felt there's a play in here for a platform to manage this from the edge of the wired-world down to the wireless nodes, and back again. I looked around and Tendril had the makings of that platform.

The Tendril Generation One architecture was a little heavy, it forced traffic flows through a central point. It made sense for people who didn't worry too much about true ZigBee compliance, but was a little bit overbearing. So we made some decisions early on, after I arrived, to modify that, and to basically get out of the networking business. Let the networking people worry about networking on the wireless devices, we would help people build end-to-end applications.

George:
So that's where Tendril starts. At some point where there is a network of some kind. And you go up from there rather than trying to reach down?

Adrian:
Yes, our customers fall into two categories. First people who had built networks already but who require things like monitoring and deployment tools so they can coexist with their existing networks nicely. But a growing second set of our customers are starting from the ground up and so we're able to architect and dictate quite a bit more about how the flows work in the network. I think we'll see increasing numbers of those people. We have as an example this customer, in parking garages, and he knows little about ZigBee, all he cares about is "I need this sensor value at this frequency at this place", so we started with our API and we're able to say, OK here's the value we care about and here's what we care about it, and by the way here's the networking information we care about as well, battery life, network connectivity and so on, and just present it all back up on the application level.

George:
So he was able to come to you with a set of business problems and have you use the Tendril services and products to translate that into an implementation of the network?

Adrian:
Yes, and that's exactly what we're about. The bet we're making is that the second wave of people into this market will come with that set of requirements. "Oh, I've seen this technology, how cool it is, I'd like to apply it to my cattle farming business, my whatever", and that's the thing, the breadth of ZigBee. Then we could use our rules engine that allows you to build that application. It will be compelling for a lot of people.

The thing that we've done uniquely is build relationships with the stack vendors - we've now got partnerships with Ember, TI and ST - we're hoping to announce some more soon. This is real tight collaboration. The applications are compiled with this agent that we use, it's not mandatory, but really it helps the structure of the end-to-end application by adding intelligence to the wireless devices. We think that's a compelling differentiator, the way we're approaching building ZigBee applications. Its something, when you talk to other people about the space, they say its too tough to pull off in that its unrealistic to expect people to compile this little agent with their application at the time when they are building it. As a consequence, others have built ways to build, deploy or manage ZigBee applications by sniffing, or using polling methods. We fundamentally believe that the market is at an early enough stage that we can win that battle. Especially if we work well with the chip guys, give to the agent out for free, and let customers put it in with their application and see the advantage of it.

George:
If you use the analog of SNMP, the IT guys have been doing this for fifteen or so years.
Adrian:
Yes, we're trying to use that analogy to get people comfortable with what we do, and we're trying to say - look the agent is open, it complies with the standard and it's free. Luckily, its very small, less than 4K.

George:
So, in your view, from where Tendril is sitting in the industry, what kind of big trends do you see? Or is there anything that surprises you?

Adrian:
I've been surprised at the pace of meter reading or energy demand management activity that has taken off. I mean if you talked to anyone six months ago, I don't think any of us would have had any visibility of that, but now there are solid ecosystems of meter manufacturers, the IBM's of this world with their platforms that manage this kind of application, the utilities, the government bodies who are all coalescing around this technology as, if not the, then certainly a solution to the demand-side component of energy management. That's heartening to see, because it's a piece of infrastructure that's going to get put in place, it's a gateway in every home, effectively, through the meter. It will allow other people to build all sorts of applications. I've been pleased by that.

I am occasionally frustrated at the pace of the Alliance, but actually it's been moving much faster than it once did. It has a more streamlined, a more rational approach to making decisions than it did a year or so ago, but its never fast enough for me. I feel encumbered sometimes by the processes of collaboration. That's why I will never be a politician, I suspect.

George:
It's one matter if you're TI and you've got this gazillion dollar revenue stream coming through. It's another matter it you're a startup. You've got a limited window in which you can be successful. They can sit and debate this for a decade at a large company and not affect their bottom line.

Adrian:
You know, from what I have seen emerge over the last twelve months, it is my belief is that ZigBee is now the incumbent wireless technology in the commercial buildings, & residential space, and it is making good ground in some other spaces as well. I think that anyone else entering these markets now, anyone with a different wireless technology needs to define themselves in relation to ZigBee. And that is a strong position in for ZigBee to be in . I think that there are people who will do that successfully, Dust is one of them - they made a play as to why their solution is different, and I believe that it is, and I think it will suit the market requirements for their markets. But you can't just get into the building automation market now with some proprietary thing and say it's great. You have to really have a good story as to why it would be better than ZigBee.

George:
And apropos Dust in particular, it's hard to claim that what they're doing is proprietary, isn't it. In some cases they're a little bit ahead of some standards that are emerging but they're really trying to follow a standard.

Adrian:
I agree. As far as I can make out, Dust is a standards play.

I wouldn't like to be Zensys right now. That would be tough spot to be in, even if their stuff worked, it would be tough. I never see them in the market any more, I know they have products, but the momentum is firmly behind ZigBee now.

George:
Can you talk about how it came about that Tendril joined the board of the ZigBee Alliance? What do you bring that wasn't there? What, I don't want to say benefit, in our opinion there were some missing things that were not on the board and Tendril in particular brought two of those.

Adrian:
We made a number of decisions when I arrived at Tendril. When I was at Ember I watched companies hedge and fail. And the example I use is Millennial Net, who basically said "We're not a hardware company, nor a software company, we're not standards based, we're not proprietary, we're whatever you want us to be when we talk to you." And Tendril had some of that. We were sort of "Well we're ZigBee, but we're talking to Dust, and we're doing some Crossbow 802.15.4 too". We decided we need to bet on the thing that we're going to do first, not forever, but the thing we're going to do first, and we needed to do it well. One of the bets that we made was ZigBee, and we basically said we're going to put all our effort into ZigBee and so with that comes the knowledge that if you align yourself with a standard, like ZigBee, you owe it to yourself to do everything you can to make the standard a success. And in some ways, Bob Heile, who's not on my Board, or on my payroll, or has any ownership in Tendril, will have a significant impact the success or otherwise of Tendril. If ZigBee takes off, we'll float with that rising tide, and we'll do very well. If ZigBee collapses, we'll feel it too.

I was Ember's representative board member on the alliance for a long period of time, and had seen how the Board worked. I felt two things were important to bring to the table. One was a software perspective on ZigBee. It was sorely missing from the ZigBee Board. And the second was a sort of neutrality or a cross platform perspective. Because we needed to necessarily make our software work across a lot of different platforms, we have that perspective. Most of the people on the Board now are focused on one platform, because they build it, or because they're an OEM that has made a decision to use that platform. We offer those two things to the board. The other thing I believe is that a Board like ZigBee needs a number of different types of companies on it. It needs big names, people like Honeywell and Samsung to confer credibility, but it also needs little companies who do much of the work. We recognized, and we very open in our application that we weren't going to confer any credibility onto the alliance in terms of reach or brand recognition, but that we would work really hard to make the standard a success and that we would live or die by this alliance and therefore we would put the effort in commensurate with that. I think it was reasonably well received, I had a reasonable track record on the Board of doing things, I had been involved in a number of programs, and I am now again. We were able to get elected to the board, I think, in a window when people remembered who I was. It's a tough place to get elected to. The rules for getting in are pretty stringent, so we had to get a number of things aligned to allow us to get in.

If I'm honest, that's only part of the reason for applying - we wanted to shore up the standard and make it a success. The second two reasons for joining are less altruistic. The first of those is that a lot of business gets done around the board and we wanted a piece of it. There is also the credibility that we get from being associated with the Alliance that is important to us. Which leads us to the third reason for applying - two guys in a garage with a PowerPoint presentation claim that they do what we do. That's always a challenge with a software business; the barriers to entry are theoretically low. So as this market emerges as a big play, all sorts of people are going to enter it, if history is any guide. And so what getting onto the ZigBee Alliance Board and becoming a Promoter allows us to do is to keep our heads above that and differentiate ourselves a bit, and say we're serious, we're plugged in at all sorts of levels and that will be a compelling thing. I mean it's an expensive thing for a little startup to do but I imagine that at the point where somebody comes to buy us or at our IPO, board membership of The ZigBee Alliance will have tangible value.

George:
So, you could go to an Accenture or an IBM or anybody who's in the business of integrating those kind of systems and say, you know, rather than choose some no-name software company, look at us, we're actually involved in developing a standard.

Adrian:
Yes, and I think that's a valuable message to bring out - often we get into see people just because we're part of the Alliance, and they're interested in learning about the Alliance and then talk about Tendril is almost a secondary thing that happens, but its quite beneficial.

So, I have no regrets about joining as a promoter. And we debated it hard, as to whether we wanted to spend the money, and more importantly, to spend the time it takes. I probably put two days a week into ZigBee board activities or thereabouts, so it's a reasonable amount of effort. Yes, we're pleased we got in. It's getting tougher and tougher to get elected.

George:
Well, I think that's always been our observation about the Alliance board. That it's very heavily dominated by technology companies, basically silicon vendors, and it needed a leavening of companies who are actually out trying to either sell products based around this technology or integrate products.

Adrian:
Yes, and for a long time it was very dominated by scientists, not even engineers, sort of research people, for whom the idea of getting a product out was almost sort an anathema to their process. So whilst I was at Ember, one of the things that we did was to try and encourage ST, Siemens, Schneider, and so on to join the Board with people who had product plans, to try and redress that balance a little bit. It wasn't just me who did that, there were a lot of people who saw that, but that's part of the reason why the Alliance is more streamlined now. It used to be an incredibly democratic place where anyone with any idea got equal billing regardless of whether you were a PhD student or you were somebody who was a product manager. And so, all sorts of fanciful ideas got debated in a way that wasn't helpful.

George:
So, is there any point that you'd like to make that we didn't ask about?

Adrian:
I think our key message is that there is a new wave of ZigBee developers emerging and they care less about the zeroes and ones of this thing and they care more about the business values that this kind of technology can add to their systems. We feel we have a platform that allows people to quickly get to a scalable ZigBee implementation, and that we won't be alone in offering this kind of solution forever, but we are right now.

Adrian Tuck, Chief Executive Officer of Tendril Networks As CEO, Adrian focuses on expanding Tendril's customer base through multiple sales and marketing channels, including alliances with leading companies in the ZigBee network industry. Prior to Tendril, Adrian served as both Interim CEO and Executive Vice President of Ember Corporation, where he guided the company's market strategy through its critical early-growth stage and helped position the company for significant growth. Under Adrian's management, Ember became the leader in the embedded wireless field, where it currently holds a dominant market share. Earlier, Adrian was President of U.S. Operations and VP of Marketing for Soft-ex, a call accounting software manufacturer that he successfully helped reposition as a high-end enterprise software and services company. Adrian also served as VP of Business Development & Marketing at Connect Communications Group. He was educated at the Royal Military Academy Sandhurst, the British Army's prestigious officer training academy where Adrian graduated near the top of his class. With a singular focus on solving the unique challenges of ZigBee application deployment issues, Tendril is helping manufacturers and integrators propel the ZigBee market forward. Founded in 2004, Tendril's software product offerings are the very first to enable organizations to quickly connect their everyday "things" to the networked computing environment and generate new revenue streams.

More information about Tendril Networks here...

This interview ran in our April 24, 2007 newsletter issue.