Chapter 12 - Networking: Market Order: LANs 1983-1986
12.19 Concord Data Systems
Notwithstanding the existence of INI, CDS management still felt they were in the driver’s seat to win the GM business. After all, they were the ones who helped GM pull off NCC. These views were held in spite of the fact that ever since NCC, GM had been dropping hints that: “we want our standard to be 10 megabits.” Soon GM began informing CDS of their requirements, not being educated by CDS, as had been past practice. Unfortunately, CDS had decided long ago to base their first products on 5 megabit per second RF modems, not the more difficult to develop 10 megabit per second modems. To now innovate a 10 megabit per second RF modem and then re-engineer their existing products would be difficult and costly, not to mention time consuming process.
CDS had little choice but to continue selling the advantages of 5 megabit per second products. Miller remembers their logic:
Look you’re better off with six channels. You can make an argument you’re better off with six 5 megabit per second channels that fit into standard TV channels than you are with three 10 megabit per second channels not using standard TV cabling.” But GM was insistent. “It was a little tougher sell. They were beating us up that the next revision of MAP was going to specify 10 megabits. I probably didn’t do enough selling.
As the spring approached, Miller spent less and less time calling on GM; partly because he didn’t enjoy the best of relationships with the GM buyer, and partly because he didn’t have the solution GM wanted. Nonetheless, CDS management remained optimistic. No one was willing to think the unspeakable. Miller recalls his attitude:
Hey, no way they can go with INI. They’re not going to make it.
The spring of 1985 brought ruthless competition in the dial-up modem business spurred by new modem chips from Rockwell. The GM contract took on even more significance. Miller remembers:
Oh, they had teams come and – oh, I mean, it was a whole big thing. Well, it was EDS.21 It got all complicated then because EDS was in the fray. So part of the problem was Ungermann-Bass and GE said they’d stand behind it. Yeah, we’ll solve your networking problems.
And then the decision came. GM had selected INI over CDS. A stunned Miller recalls:
I didn’t believe it. Couldn’t believe it! Just couldn’t believe it. It was terrible gloom.
Losing the GM business was a severe blow; one too painful to accept fully. And then, as if winning was not enough, INI aggressively pressed their advantage. Again Miller:
And also INI had so much money. They were throwing money around like crazy. They were – Schoendorf was in all these articles. You know: ‘This is my LSI chip.’ They were putting salesmen in the field like it was going out of style. And we were losing money. I had been beat up. We didn’t have enough marketing. It’s not right. We hired a marketing guy and he turned out to be a disaster. There was – so there was some management problems unraveling at the same time.
CDS did not become uncompetitive instantly. They would sign Honeywell as an OEM. What had happened, however, was CDS was no longer the perceived technology leader. Yet anyone reading the trade press before the GM decision easily could have concluded CDS was the market leader - as their Board members and sophisticated institutional investors had. Miller remembers the period following the loss of the GM account and the aggressiveness of INI:
I mean the stakes got higher. The ratchet had gone up a couple of notches because now we have this well funded guy out there throwing money at the problem – got two big companies, Ungermann-Bass was no small company and GE also behind them. All these bucks to spend. And hey, we’d spent net $10 million, at that time. But we still had a good reputation. We still – I mean, it wasn’t a total disaster. We’re not out of the market. So it was us and one other big player. And then we always – if anybody was sophisticated enough to measure performance, we always won. And INI was out making promises like crazy, while at that time you couldn’t see whether they could deliver or not. It was a mess. That’s when it was unraveling a bit. And at the same time, the modem business was less profitable. It was more competitive. At the time, we also had all our financial statements combined together, so it was hard to determine what the hell was going on. So I was under real pressure.
In the December 1985 issue of Data Communications, an article titled: MAP: A user revolt for standards, prominently mentioned CDS, albeit INI was clearly the market leader. In the article, CDS represents: “Board-level prototype implementations of the MAP chip set22 will be available in March 1986, and general deliveries are slated to begin in the second or third quarter of 1986.” The same article reported: “So far, INI has spent $25 million in organizing and developing its MAP board products, and it expects to spend another $25 million expanding its sales and service force and continuing development of MAP-compatible products.”
Despite continuing optimistic forecasts for the MAP-token bus market, Andy McLane, Board member and the company’s original venture capitalist, was concerned. Could CDS compete successfully against a strongly motivated, well-funded competitor; especially given CDS’s weak financials? He argued the company needed to find a corporate partner, a partner with money, a strategic partner like the one UB had found. Miller remembers the negotiations with Gould, the parent of AMI, the semiconductor foundry developing CDS’s token bus chips:
We decided we wanted a corporate partner. That was the thing to do. So Gould had been after us. We had some very strong discussions with Gould. In fact, we got an offer on the table for a minority investment. But the problem is, it was taking you over without taking you over. Betwixt and between is no good.
CDS had faced this choice years earlier with Analog Devices. Just as then, they declined. Miller remembers:
And then we had heavy discussions with Allen Bradley. Rockwell had just taken over Allen Bradley at the time. I mean, Allen Bradley was an OEM and it made a lot of sense. And at the last minute, they said: ‘Well, we’ve decided, no we’re not going to pursue this anymore.’ They got spooked by the deteriorating modem business and the price we were asking. They ended up buying Interactive Systems from 3M for a song.
The Board of Directors of CDS scrambled. Straining for ways to put a positive spin on a deteriorating situation, they considered splitting in two: a LAN company and a modem company. After all, wasn’t that what potential corporate partners had been saying? Again Miller recalls:
Both Gould and Allen Bradley said: ‘We’re not interested in this modem business!’ You know, the financials were going to shit, so they got scared off. And so we talked with a bunch of others too. We knew that MAP was being talked about as this fantastic market, it’s going crazy, so we had a lot of people courting us.
The idea quickly became a decision. Given the growth and excitement of the LAN business, it was concluded money would be easier to raise for their token bus company. So they formed Concord Communications Inc. with the mission to become the leader in token bus local area networking. Shortly thereafter, management finally had to acknowledge the simple fact that the project to develop VLSI token bus chips with AMI a disastrous failure. Not only did it leave CDS without the chips it had been counting on for success, it had cost the company $12 million with nothing to show for it.
So we waited – and actually it bled the modem side. It was awful. So we couldn’t hire people and, oh it was awful. And there was also a big rivalry in the company. Big jealousy in the modem people: ‘Oh, those fucking guys have been a bunch of prima donnas.’
The Board also decided to go after new management. “And we also made the decision, hey, I’d be best – I was pretty burnt out probably by then. My strength wasn’t operations anyway. We got a guy who was a friend of TA’s to come in and be here two to three days a week. We went out to raise money for the LAN side alone. And we got $7 million in June 1986.”
The LAN market, especially the MAP segment, had not lost its luster. Concord Communications raised $7 million with a company valuation of $22 million in June 1986.23