The Love of Linux Chapter 7: The Python Swallowing the Elephant
I use the analogy of the python, because after a python has swallowed its meal, it maintains the shape of its victim until it has had time to digest it. Caldera had from its inception, being trying to develop a global reseller channel. We knew that we needed to provide business solutions to customers and the integrator/VAR was, and is, instrumental in providing the total solution to the business customer. These integrators also provided much needed business applications. To build a channel takes time and money. To build a channel from scratch is very expensive and time consuming. Caldera began this process early on. Long before the IPO, we had contracted with MindShare to recruit resellers and begin to build a reseller channel.
With the capital and currency from the IPO, we felt that the best thing to do was to buy the channel and then grow it. We also knew that Linux could make an ideal small business server platform because it was very stable. But, to get it accepted into small business, you needed the reseller and existing customer base. It is a known fact in the industry, that the local reseller does have tremendous influence on what the customer buys. At this time, approximately 75% of what a small business owner purchased was based on the recommendation of the value added reseller they work with.
Ralph Yarro, the president of Canopy and chairman of our board, suggested that I look at Santa Cruz Operations stock. Ralph owned just a few shares and had been tracking their price. Early on at Caldera, Bryan Sparks, Ralph and I had almost jokingly talked about buying SCO and UNIX if Caldera was ever successful in going public. Ralph and I looked at the stock and saw that they were trading a little less than half of our share price. SCO had one of the best reseller/integrator channels in the industry, not to mention an impressive list of business customers like, McDonalds, Kmart, Taco Bell, KFC, only to mention a few. I mentioned to Ralph that I had met Doug Michels a few months earlier when SCO launched Tarantella on Linux. SCO had invited Caldera to participate with several other Linux companies on a panel at their press briefing. After the panel, Doug Michels came up to me and introduced himself and thanked me for participating. I gave Doug a call and asked him if he had considered selling SCO. Doug mentioned that if I had not called him, he would have called us in a couple of weeks. We set up a meeting in Las Vegas and I flew down and met with Doug and a member of the SCO board.
Much of our initial discussions were how to structure a deal that would allow new growth business of UnixWare and Linux to be separate from the old declining revenues of OpenServer, the old UNIX product that was the back bone of SCO’s success. What resulted was a quite complex transaction that the SEC had a hard time digesting. The original agreement and frankly, nearly all negotiations with Tarantella were incredibly difficult. The Tarantella management truly subscribed to the win loose mentality on negotiations. If we won they lost and if they won we lost. The concept that they had a real stake in Caldera’s future, owning a substantial amount of stock in our company and would benefit from a win win proposition appeared totally alien to them. It appeared to us that all they could see was the short term and they wanted as much of our cash as they could get. There were several times when the negotiations were broken off because we had reached an impass. On one such occasion, I received a call from Maureen O’Gerra notorious for getting the scope on all the UNIX and Linux stories in the industry before they appeared any where else. I had always leveled with Maureen and I think she appreciated that. She asked me point blank if there was anything going on with the purchase of SCO and I honestly told her no. We had walked away from the deal. However, within days, SCO responded favorably and the negotiations continued. When we made the announcement a couple of weeks later that we intended on purchasing SCO, I do not think Maureen ever forgave me. SCO had broken up their company into three divisions, the server division, professional services, and Tarantella division. Caldera purchased two of the three divisions. After the purchase, the remaining Tarantella division changed the name of their company to Tarantella.
The transaction took almost nine months to get SEC approval and complete. In the original agreement, Caldera did not actually purchase the OpenServer business. Tarantella maintained ownership and Caldera became the exclusive distributor. We owned the distribution channel and they owned the OpenServer product. The deal was structured this way because Doug was convinced that if the Caldera was saddled with decreasing revenue stream of the old business, investment analysts would never give the company a good valuation. The logic seemed sound, but we also felt that Doug wanted to hang on to his baby or at least the revenue from it and the negotiation process was tedious at best. Regardless, the resulting deal was far too complex. After six months into the SEC review process and continued cash drain on Tarantella’s quickly depleting resources, Caldera renegotiated the deal with Tarantella and purchased the Open Server business as well. But it was not until the cash situation at Tarantella became almost desperate that we could negotiate a reasonable agreement.
To manage the company during the SEC review process, Caldera and Tarantella set up a way for Caldera to begin to influence the direction of the two SCO divisions we were acquiring. Caldera held management and strategy meetings with the existing SCO management of the two divisions and were able to choose what resources we were going to keep and what resources had to go. We would then meet with the Tarantella management to get their approvals and agreement in a meeting affectionately called Fortnight. However, since Caldera did not know the company or the employees well, we had to rely heavily on the advice and input of the SCO division management. Unfortunately, they were still being paid by Tarantella and the bonuses seemed to provide more direction than either Caldera or the remaining Tarantella management team. During the SEC approval time, the Tarantella team paid incredible sales bonuses to the existing management team to ensure that they would sell the old OpenServer products and minimize any and all marketing expenses on either OpenServer or UnixWare products. At the time, Tarantella was quickly running out of cash and was milking the SCO revenue stream to support investment in Tarantella. Consequently, to meet the expense targets and get the bonuses, nearly all marketing of the SCO products stopped in the Americas for 9 months. Tarantella also provided incentives to the sales force to sell the old UNIX product Open Server, because the original agreement allowed Tarantella to maintain ownership of this product. Under the original agreement, nearly all the proceeds of OpenServer would have stayed with Tarantella and Caldera would have been paid a royalty to market and sell it.
Originally, Caldera intended to keep as much of the old management team in place to ensure continuity. Honestly, I had never run a 600 person company before and did not have the confidence that I or the original Caldera management team knew enough about the SCO business. But it was very difficult to get a clear picture on the state of the business because the information was being filtered thorough the remaining SCO division management. The remaining SCO management of course, did not want to be the bearers of bad news to the new owners. Unfortunately, this filtering continued up until the last several weeks of the first quarter of doing business as Caldera International following the close of the transaction. For example, the old SCO management team represented that we were going to meet the sales projections. We soon found out that they were over 5 million dollars off. Upon investigation, the projections from the sales force were very accurate, but this same layer of management had made a management uplift to the numbers at the beginning of the quarter, probably anticipating renewed interest in the products after the close of the sale. But, they continued to represent that they would meet the unrealistic projections. Unfortunately, we had kept additional resources in place during that time anticipating better revenues. This misinformation created credibility problems with the board. Being the CEO, I had to represent and defend the numbers to the board given to me by the management team without having complete access to the information.
Realizing that we were paying a premium in many different ways to have an extra layer of management and we were not getting the information we needed to run the business, we decided to let nearly the entire middle management layer go and risk making a few mistakes ourselves. As a lesson learned, if I were to do a major accusation again, I would negotiate up front for the old management team to transition out immediately or in a staggered manner as each element of the business is effectively transferred. You can learn a business pretty quickly by taking to the customers and solid business principles apply to all size of businesses. At Caldera, the Executive team made a significant effort taking the time to develop a strategy with key managers from all over the company. We hired several consultants to assist in this process which was very helpful. The Executive team would purposely stop in and meet with employees all over the world when we would visit a customer or attend a trade show. We flew around the world meeting with customers, OEMs and ISVs so that within months, we had a very good handle on the business and where we needed to go. Each member of the team, Drew Spencer, Benoy Tamang, Ed Donakey, Bob Bench, Reg Broughton, Jeff Hunsaker and others played a great role in helping to ascertain the needed strategy and develop confidence with the employees and managers. So if I were to acquire a company again, my recommendation would be to let the existing managers know they will be transitioning and negotiate separation packages up front to provide them the incentive to assist in the transition. This method is far less costly than trying to keep them on and educate them on a completely new way of doing things or leaving them to guess about their future. In some cases, the managers can make the transition and may be a great asset, but only a little time will tell. A good transition package with give you that time and allow you the opportunity to see if the managers are willing to accept the new direction and play a vital role in achieving the new vision.
Caldera now had the infrastructure it needed to deliver high quality business solutions around the world and the management team had a common goal, build a world-class business in a new world economy and make a difference. It took us longer to digest the SCO acquisition than we have anticipated. After four quarters, we had finally gotten through the majority of the one time charges we needed to take to clean up many items like 20 year leases SCO had entered into. We had nearly twice as much office space as we needed worldwide. The number of employees was nearly 650 and we had to pair the back to 395. The revenues were declining at nearly 10% a quarter and we were able to slow that down and level it off. At the same time, the economy and the entire IT stock market dropped to record lows. The Dot Com bubble had burst and the economy suffered. During this time, Caldera cut out nearly 36% of its costs on a quarterly basis for five straight quarters. While some of those cuts were in laying off people, we went through great pain to minimize that option. Management took fairly significant cuts in pay as did some of the employees who had been with the company for some time. The management team and the employees did a fabulous job. They consistently met their budget targets every quarter without fail. The experience was like trying to pull the nose up of a 747 whose engines quit during flight.