AUDC...High Tech Features
AudioCodes: Not for Sale
By Hadass Geyfman
Shabtai Adlersberg (46) and Leon Bialik (40) lacked for nothing. Six years ago, they were cosily ensconced in senior slots at the very top of DSPG. They had a fair sized share package in their names, and, routine problems aside, the two men, approaching mid-life, could start enjoying the comforts of their executive armchairs.
There was just one problem. They didn't particularly like treading some pre-appointed path, especially when asked to do so. One fine day in 1993, fed up with being regarded as professional troublemakers responsible for giving DSPG managers a chronic migraine, Adlersberg and Bialik decided to take the high jump without a safety net. They departed for AudioCodes, then a chip development start-up. They had no definite direction, the market was unclear and the future shrouded in darkness.
Less than six months ago, at the end of May 1999, AudioCodes was floated on Wall Street having been catalogued in not less than two of the hottest categories in town. Since then, they have been opening the newspaper every morning to find their share executing a crazy one-way trajectory – upwards. Not for nothing did analysts dub the company the most successful launch by an Israeli company on Wall Street this year. At the time of writing, the share price stands at $67, up almost 500% since its IPO and 163% since its second issue on October 14.
"Globes": Were you surprised at the issue results?
Shabtai Adlersberg: "The biggest surprise was the IPO", he says, "we did not expect the share price to rise so high in so short a space of time. We discovered we had been listed both in the communications components category and in the enabling technology category, because our communication cards enable the manufacturer to build a system even if he understands nothing about VOIP access servers. And investors are courting both categories.
"In retrospect it looks simple, but when we approached the underwriters, we had to work very hard to persuade them there was a new market developing here. We had to prove to them that this product had economic advantages, and that it had buyers. Luckily for us, at that time, several other companies engaging in the field went public, and that boosted awareness".
Analysts reconsider
Despite their favourable Wall Street reception, the issue results were not the first indication that AudioCodes was in possession of hot number. The first time Adlersberg and Bialik realised they were onto a success story was in the last quarter of 1998, when the big orders from communications giants start pouring in. Not even the most optimistic of analysts could have anticipated what then happened to the company's revenue graph.
Last year, AudioCodes posted revenues of $9.2 million. At the beginning of this year, analysts assessed that it would wind up the year on revenue of $17-18 million and 2000 on $39 million. Constantly changing data convinced them to rethink that assessment, and they then published revised expectations for $27 million by the end of 1999. Orders meanwhile keep flooding in, and 1999 is now expected to wind up with revenue of more than $30 million, and 2000 with $45-48 million.
Is there something about the company's strategy that explains this?
"I think one of the things that enabled the breakthrough in sales was that the company's employees were capable of swerving onto a completely different tack while working, and we did that, as it happens, several times. Originally, AudioCodes' main focus was on communications components. But at a certain stage, it transpires that some of our customers were unable to utilise them. Worse, there were customers who had invested a great deal of energy for a nine-month or even a twelve-month period who were likewise unable to arrive at a working product. We realised that if we wanted our customers to rely on our components, we had to concern ourselves with their time to market.
We accordingly decided to develop a product at a higher level of maturity, a communications card with a large number of chips. With this card, the customer need not make any great contribution, and can get to market faster. This change gave sales a material push, and we expect that by the end of the year, communications card sales will overtake component sales. Luckily for us, none of our rivals identified this need.
"Something else we thought important was the fact that in this industry, there are manufacturers that only make chips and others that only make cards. There is usually not enough flexibility and not many solutions are offered that diverge from the manufacturers' principal sphere of occupation. Therefore, the broad range we offer, from component level to card level, including interim combinations, enables us to offer more solutions and thereby enlarge our target groups".
What about technological edge?
"By means of the one chip or communications card, we can handle a larger number of channels than the competitors. Also, the voice quality we provide and our modem and fax signal transmission quality, which neutralise phenomena such as hoarseness, long delay and so forth, is higher than that of the competitors' products".
Discord in relations with DSPG
After the departure of Adlersberg and Bialik, DSPG decided to invest in AudioCodes, and actually acquired its majority shareholding. Today it holds 17.7% of the shares, while Adlersberg and Bialik each hold 15.4%.
When the two men left, both parties declared at every possible forum that "we parted as friends". Adlersberg insists on sticking to this version of events, firmly denying rumours of any dispute with DSPG, but sources associated with the company report things differently.
"One mistake DSPG management evidently now regrets", says a capital market source, "was the decision to reject out of hand an innovative idea of Adlersberg's. It is this idea that currently underlies the AudioCodes product technology.
"Adlersberg and Bialik decided to leave and senior DSPG managers told Davidi Gilo: ‘let them go. We have the technology, we don't need them'. But Gilo decided that he wanted to go into the new company as a strategic investor.
"In 1995, a number of changes took place in DSPG, and it stopped functioning as a strategic partner. But also in that year, it sent emissaries to Adlersberg and Bialik offering to buy their share in the company, which was then 40%, for $20 million. This met with a firm refusal.
"In 1998, AudioCodes got another offer. This time, from a large US investor who offered to invest in the company and buy part of the founders' stake at a company value of $50 million. Adlersberg and Bialik refused again.
"Later that year, AudioCodes refused a third purchase offer, this time from a US communications firm, at a company value of $120 million.
"Today the company is valued at $1.1 billion".
Will you sell the company at some stage?
"No. Our trend is to develop the company and continue to be independent. We have no intention of seeking to be bought out. On the other hand, I know the game rules and I know this is a shark-eat-shark world. If an offer should some day turn up that the shareholders find appropriate, not just financially, but also in terms of how the company will look in the future, what it will engage in, what direction it will follow and whether it will also be identified with a market leader, we will accord it due consideration. At the moment, in any case, all resources are directed toward corporate development".
There is a rumour that Intel approached you a week ago with an offer to purchase the company at something in the range of $2 billion".
"There are lots of rumours. I know of no such offer".
Internet telephony will be poor quality
The object of AudioCodes' systems is to increase in calls volume capacity of a given line. In recent years, chips were adapted for IP telephony applications and the company's products constitute the building blocks in systems that bridge between traditional telephony networks based on circuit switching and networks based on packets switching.
Voice transmission on IP network currently suffers from two main problems – loss of information packets, and their failure to reach their destination on time. These are both detrimental to voice quality. In order to overcome the problem, the company develops a condenser called NetCoder that is capable of materially compensating for loss of information packets and improving sound quality.
Whereas in other technologies, the loss of 3%-5% of information packets considerably reduces speech quality, the NetCoder technology is capable of maintaining high speech quality, even if 8%-10% of information packets are lost.
Adlersberg is at pains to state, at the start of the interview, that AudioCodes' developments are meant to facilitate voice quality in telephony via IP based communications networks but definitely not Internet telephony.
"IP is a method of transmitting information over the network in information packets", he says. "In this way, the information passes more efficiently and rapidly than in the existing system. The Internet, on the other hand, is a random collection of networks which are, in fact, IP-based, but each of which has a different character and there is no way of controlling what happens on them.
"And when you can't control terminal-to-terminal delay, you can't promise there will be no collisions between information packets and that there will be no loss of such packets. You can't give service. The only places where we install telephony on Internet is in remote locations, where there is, in any case, no quality communications infrastructure".
So what do we actually need IP for?
"To enable the combination of picture, voice and data on the same infrastructure, using a cheap and efficient network technology. Anyone wanting to construct service baskets that combine all these types of services, will have to transmit them on a single network and in the same form. This capability is contingent first and foremost on the use of IP or ATM networks".
What is your current market share?
"I suppose that once the market develops, more players will come in with huge resources - such as Texas Instruments, Intel, Motorola and others. And then, if we can hold more than 20% of the market, that will be a certificate of honour for us. My assessment is that we need to be capable of being 15%-20% players in the open market".
Aren't you worried about competing against communications giants?
"What can we do? The market is developing and the giants are coming in with tremendous resources. But I am optimistic, I believe we have a technological edge and a timing advantage".
Even though they have several times your resources?
"Money isn't everything. The most vital and most important resource her is manpower. There aren't very many experts in this field in the world today, and it's very difficult to get your hands on those that do exist".
What technologies will dominate the market, in your opinion, and which do you intend to go into?
"I estimate that as far a voice transmission is concerned, the ATM networks jointly with IP will bite off most of the market. Frame relay will evidently serve for data transmission only. It should be borne in mind that pushing systems and protocols is affected by the strength of the players behind them. Therefore, when Cisco and Ascend and 3Com push IP, that causes VOIP to develop. The big telecom companies such as Nortel, AT&T, Newbridge and Sprint are behind ATM and therefore this technology, too, will develop. There are no dominant giant players behind frame relay though, and therefore this technology will not develop as much.
"This year, we already intend going into voice on ATM, mainly in combination with DSL systems. Apart from that, we are now developing a technology that will enable us to combine VOIP with rapid dial modems, for Internet access. This combination will create a new generation of services, by means of which the consumer can obtain from the ISP telephony services in addition to surfing services. This server will be capable of integrating between the possibility of telephone calls and Internet access using the same communications system".
Your customers include Lucent, Alcatel, Cisco, Motorola and other communications giants. How do you persuade such companies to integrate components of a young Israeli company into their products?
"Our first customers were start-ups. They included a company from Dallas, Texas, called Celsius, which Cisco acquired. And that was how we reached Cisco. We reached Lucent through a company be the name of SDX, which was Britain's third largest manufacturer of exchanges, until being bought up by Lucent. Alcatel encountered our chips three years ago and examined them".
Asked about making the transition from development functions to managerial functions, Adlersberg replies that "what doesn't kill you strengthens you. As a rule, we have cumulative experience of many years of measuring up to the market, and we have developed an eye for what is economic and right and what is not. We have also taken pains to study what happens in other companies, when they succeed and when they fail.
"Since we did not identify the market when we started the company, and we knew it would take us two years to build a technological infrastructure, we decided there was no point in setting up a large sales complex right at the beginning".
Has worker morale changed since issuance?
"The workers here have always made a huge investment, but since we went public, I sense that they have simply taken wing. Suddenly the whole world is looking at the company, and it is important for people to be identified with success".
How is it to get up in the morning with a company value of over a billion dollars?
"Frightening. At that value there are expectations from management, there is responsibility toward the shareholders, and you must not let anyone down. After the IPO we did not want to be a shooting star. We wanted to go on showing a continuous growth, and that is the only concern that keeps prodding us forward. Personally, whether I am worth fifty, ninety or a hundred and twenty million dollars, it makes no difference. That is not what motivates me".
Published by Israel's Business Arena November 18, 1999 |
