In 2018 there were 2,746 software mergers and acquisitions. There were 656 in the first quarter of 2019 alone. While the number of IPOs is on the rise, M&A represents the most likely exit scenario for most companies. Integrating two companies after a merger is tough. The stories of acquisition integration failures are legendary. Here’s a link to the 37 largest M&A failures in recent history. Years ago I was part of an acquisition integration project. I worked for a $40 million division of a conglomerate that acquired a $165 million business unit of a Fortune 500 company. This is the story of that project and what I learned from it.
In the late 1990s, I was a director in the marketing department of Sterling Software’s $40 million Application Development Division. Sterling had acquired KnowledgeWare in 1994. They split the business among two Sterling business units. In 1997 Sterling acquired Texas Instruments’ Software division for $165 million and integrated it with the Application Development Division. This story is about my participation in the integration process and the lessons I learned.
Sterling Software was a serial acquirer. Sterling executed the first hostile takeover in the software industry. In 1985, with the help of Michael Milken and junk bonds, Sterling completed the hostile takeover of Informatics General. Informatics General was the fourth largest independent software company at the time.
Before I begin I want to emphasize that there is more than one way to integrate two companies during a merger. Sterling Software was a company that grew primarily by acquisition. By 1997 they had completed more than 35 acquisitions. They had a process and a set of best practices for acquisitions. I have participated in and led several acquisitions and divestitures since then. There are many ways to integrate two companies during an acquisition. One way is not better than another. What is important is that the process reflects the values, culture, and lessons learned by your company.
Welcome to Elysian Fields
In June 1997 I worked at my division’s headquarters in Atlanta. Sterling Software headquarters was in Dallas. I got a call to get on a plane and come join the integration planning project. An assistant booked me a first class ticket. I arrived at The Crescent, one of the most prestigious office complexes in Dallas, complete with a 5 star hotel Crescent Court. The office were class AAA and the walls were decorated with famous art work (Thomas Hart Benton, N.C. Wyeth, a few Picasso’s for good measure). Everyone was dressed to the nines and there were many assistants to cater to whatever you needed (you need to send a present for your niece’s birthday? I’ll handle that). Meals were catered in from 3 star restaurants. You were in corporate heaven. Little did I know that this was all part of an elaborate Kabuki dance that I would understand once it was over.
Little did I know at the time the entire integration process had been carefully designed. After 35 acquisitions the Sterling executive team had learned what worked for them and what didn’t. As I look back I now understand that there was a lot they didn’t tell us.
After the initial shock of walking into the Elysian Fields, a logical question arose. When will we be going home? When I got the call to come to Dallas my boss simply said “pack your bags and get here”. Some colleagues had been invited into the process earlier than I had and they had been gone for ten days. When I asked the question when I would be going home my boss told me “when we finish all of the work. The hotel will dry clean your clothes and if you need something else Cindy will run out and get it for you.” I had arrived on a Tuesday and packed enough to last until Friday. I didn’t go home for three weeks.
At the Crescent, Sterling had gathered all of the potential candidates to fill executive positions in the new organization. It was a mixture of people from the existing Application Development Division and people from Texas Instruments. At its peak there were about 50 people from the two companies. There were also the senior executives from Sterling — the CEO, COO, CFO, VP Business Development, General Counsel, and a few special advisors.
One of the primary objectives of the integration planning process was to indoctrinate the management of the new business with Sterling’s management system. Like most successful companies Sterling had its ways of doing things. Their management system consisted of a few things:
- Standard organizational structures across all eleven divisions of the company. Each division had a common set of executive positions (Division President, Vice President Finance, VP Marketing, VP Labs, VP Sales, etc.)
- Standard structure for international direct selling operations and distributor operations
- Standard annual planning process
- Standard monthly and quarterly reporting and review processes
- Sterling Software Core Principles. A set of ten corporate and cultural values that actions and decisions were judged by. One of the classics is a saying “The Plan is The Plan” (more on this later)
For example, there was a standard for organizational charts — it included how to display names, titles, office location, and size of headcount managed. There was even a Sterling Software Style Guide — a 25 page brand guide that covered everything from official logos that had to be used, product naming conventions, dress codes, to even when and how you were supposed to wear the official Sterling Software pin.
The Model was a detailed financial model that managed the business and calculated what guidance to give Wall Street about revenues and earnings. It was the consolidation of the P&L, Balance Sheet, and Cash Flow Statement of each business unit in the company into one massive model. The Model was also used to determine at specific purchase prices whether an acquisition would be accretive or dilutive to earnings per share. The model had three cases: base case, best case, and worst case. The cases had different assumptions about revenues, expenses, depreciation, amortization, and cash flow. It as a massive Lotus 1–2–3 spreadsheet and it took one highly paid financial analyst full time to maintain.
The Model was worked out before the acquisition was even announced. A key objective of Integration Planning was to test out the model assumptions for each case and settle on the version that would eventually be taken to Wall Street. The contents of the model were never shared with the participants in the integration planning process. Yet its contents drove almost all of the final decisions made about headcount and operating expenses. The integration planning process helped to validate the model and evaluate the sensitivities of the edge cases. Since a large part of the wealth of Sterling executives was based on stock performance getting the model right was critical.
As mentioned earlier, the agenda for the entire integration planning process was never explicitly laid out for the participants. Sterling employees who had been through integration planning processes for other acquisitions generally quiet about their past experiences. Every evening you would be given an assignment or agenda for the next day’s activities. In retrospect the integration planning process could be divided into four major activities:
- Organization Building. The first step was to select executives to fill each slot in the top level of the organization. The process was then repeated for the next level of management. The process continued until each spot in the new organization was filled. What people didn’t realize at the time was that if a name was not on the chart, that person would be terminated from the organization when the acquisition closed. Needless to say there was a lot of horse trading at the end of the process to ‘save’ certain people once they figured out what was going on.
- Strategic Planning. The next step was to develop a high level strategic plan. It covered things like goals, core strategy, product portfolio and roadmaps, positioning, go to market plan, departmental tactics, special projects, and a financial plan. The strategic plan set the context for tactical action plans.
- Tactical Planning. 30/60/90 day tactical plans for each department. This covered everything from changing building signage, updating websites, transitioning employee benefit programs, customer contract novation, etc.
- Financial Planning. The end goal was to produce a complete financial plan for the new organization expressed as a P&L, Balance Sheet, and Cash Flow Statement. This became the ultimate expression of the plan and the basis for everyone’s incentive compensation. It was somewhere between the worst case and base case version of The Model used to drive the acquisition.
The central feature of the integration planning process was the daily executive review meeting. Each day your team had an assignment to produce something — an org chart, a version of the strategic plan, proposed names for product rebranding, etc. Once you had your deliverable, you had a review meeting in the corporate board room. The attendees were the senior Sterling executives (CEO, COO, CFO, General Counsel, VP Business Development, special advisors) and a good chunk of your business’ executive team. The board room had a massive marble conference table that seated about 25 people. Each executive had a dedicated spot — they would sit in the same place for weeks at a time. There were two massive display screens that also served as videoconferencing links.
Each day you would go in and give your pitch and get ‘feedback’. Sterling operated like a Marine boot camp. The strategy of ‘Tear them down and then build them back up’ was practiced every day. In the beginning the criticism was merciless. Part of the challenge is that the Sterling senior executive team had been through this process so many times they could spot bullshit from a mile away. Sloppy work or inaccurate numbers were not tolerated. Many times they would stop a presentation mid-stream and tell you to go back and do it again. You went to your work room, redid the deliverable, and came back the next day to present it again. The questions they asked were direct, piercing, and insightful.
These sessions were like tryouts — the team was trying to see how you thought, adapted, and responded to pressure. Some folks did not make the grade. The next morning they simply weren’t there. When you asked where Sue was the answer was ‘she had to go home to take care of some stuff’. Sue typically did not make the ‘keep’ list at the end of the process. Knowing that you could be the next one to be sent home put a lot of pressure on you.
As the days and weeks progressed the criticisms decreased and the compliments increased. Part of it was people learning how not to put their foot in their mouths. Another part was an intentional effort by the Sterling executive team to rebuild your confidence.
In retrospect one of the most stressful exercises was building the org chart. First, the Sterling executive team met and picked who the division president was going to be. This was usually done well before the integration planning process began. Next the senior team and the division president would meet to determine who was going to fill each VP slot. A key part of this was the ‘where you were born story’. During operational due diligence, the senior Sterling team would hold meetings with the acquired company’s management. One thing they did in these meetings was to ask each person that presented to tell their ‘where they were born story’. The story is a 3 to 5 minute summary of everything about that person from the time they were born until the time they were standing in that room. What you choose to say in that story had a lot to do with how the Sterling team evaluated you. The Sterling executive team always took copious notes at this stage in paper no the days and weeks progressedtebooks that they always carried with them. They would reference these notes throughout the entire integration planning process.
In addition to the story, each executive would rate you on a scale of 1 to 10. If you were an 8 they would hire you for the job you were in at the time. An 8.5 meant that you were better than average in that job. A 9 mean that you were one of the best people for the job in the entire company. A 9.5 meant that you were one of the best people in the entire industry at that job. Anything below a 7 was a relative degree of insult in terms of how bad they thought you were in the job.
When it came to filling the vice president positions the executives would go around the room. Each person would say who they thought should fill the position, what their rating of that person was, and why they felt that way. After everyone had spoken a consensus was determined and the position was awarded.
Once the VP slots had been filled the same process was repeated for each department. This time the voting group consisted of the Sterling senior executives, the division president, and the newly appointed division vice presidents. This was for director and manager level positions. This is where things got interesting because different VPs had different opinions on various candidates. It was also the first time that people from Sterling and the acquired company had a serious discussion about candidates from both companies. It was contentious at times but the consensus driven model produced good results. Once the management positions were filled, the rest of the org chart was built using the same approach. The entire process took 3 to 5 days to complete. If your name was not on the org chart you would be terminated (with generous severance) when the acquisition closed.
Deciding the fate of people you have worked with for years is tough. They are your friends, you know their spouses and even their kids. For financially driven acquisitions like Sterling did, large job cuts were inevitable. Payroll is 70% of most software company’s expense. The only way to significantly increase profits is through headcount reduction. In the case of Sterling’s acquisition of TI Software over 400 positions were eliminated. Some people could not bring themselves to participate in the process. They ended up leaving the integration planning process and returning home. Some were kept on in the new organization. Some were terminated when the acquisition closed.
Nothing builds a team better than a common enemy and a shared painful experience. In integration planning, the Sterling executive team became the common enemy and the process was the shared painful experience. A typical day would start around 7:30am with the first board room pitches scheduled for 8:00am. The day would end around 8:00pm or 9:00pm. This would go on for weeks. Sundays were a little different, the sessions would start at Noon but still run into the evening. Having your work publicly criticized in front of your peers, bosses, and boss’ bosses for days on end is disheartening. It definitely reinforced the ‘we versus them’ attitude.
After the day’s work was done the division’s management team would repair to the bar at the Crescent Court hotel. There they would commiserate or celebrate what had happened that day. People from the different companies and even countries could sit down and get to know each other. The war stories were numerous and memorable. These gab sessions would continue to 12:00 or 1:00am. Then it was time to ‘sleep fast’ and start again at 7:30. This cycle continued for six weeks. After week 3 we were allowed to go home for a weekend — primarily to pick up fresh clothes and say hello to our family.
During the planning process we were isolated. We were not allowed to discuss anything that was going on with the people back at our office. Both Sterling and TI were public companies. Disclosing material non-public information is illegal. One of the difficult things is dealing with people that you know are going to be terminated when the acquisition closes. You avoided all contact to minimize the chance that they would ask you if they were safe or not. Absolute confidentiality of what was going on in the planning process. More than one person was sent home and put on the ‘cut’ list for breaking confidentiality.
The combination of overwork, isolation, and stress built strong relationships among the people locked up in the planning process. Like a battle you never forget who was next to you in the trenches when the bullets begin to fly.
The rollout is the penultimate event of the integration process. It is the combination of the acquisition legally closing, the public announcement of the deal, the notification and termination of impacted employees and the briefing of all go forward employees. Press and industry analysts are briefed as well. Building and office signage is updated as well as websites.
For the participants in the planning process it is a day of intense conflicting emotions. On the one hand there is intense pride in having rebuilt the business from the ground up in less than 40 days. On the other hand there is regret and sadness when long time colleagues are terminated. There is a sense of survivors’ guilt.
The bond among the new management team is now almost unbreakable. They have been through a shared painful experience that few people will ever experience in their career. The bonds built in the Elysian Fields of Crescent Court will last a lifetime.
My participation in the Sterling/TI integration planning process was a defining moment in my career. When I arrived at the Crescent my division president told me that I was going to be promoted from director to Vice President, Business Development. There were only 40 other VPs at Sterling. It would have been my first executive job. Unfortunately that changed three days into organizational planning. Sterling was having a hard time finding a director of product management for Cool:Gen. Cool:Gen was the industry leading Computer Aided Software Engineering (CASE) tool. It was TI Software’s largest product line that accounted for over 70% of their global revenue. Due to its size, it would be more than 15% of Sterling Software’s revenues.
Sterling was having a hard time finding someone they trusted for this position. After three days of trying the CEO convened a meeting in his personal conference room. The attendees were the CEO CFO, Corporate VP Business Development, General Counsel, Group President, Division President, Division VP Finance, and me. After exchanging pleasantries, the CEO cut to the chase. He wanted me to take the director job instead of the VP job I had initially been offered. He went around the room and asked everyone to rate me in the director job. I was given ratings of 8.5 and a few 9.0s. After the ratings he looked me in the eye and asked me if I wanted the job. I smiled and said ‘Yes Sir!’. I was bitterly disappointed, but soldiered on.
The job was tough. People resented that a guy from Sterling was responsible for their core product. There were serious quality problems and major customers were pissed. I went on a worldwide apology tour to meet with the key customers and assure them the defects would be fixed. Six months after starting the job I was in Zurich Switzerland for a third meeting with an unhappy customer. I got a call from the Group President telling me I had been promoted to Group Vice President, Business Development. In one call I went from being a director product management to the #3 executive in a $200 million business.
In my new role I was responsible for mergers, acquisitions, and divestitures. I led two major acquisitions (Synon Corporation and Cayenne Software). I also led two divestitures, including one for over $10 million.
By the year 2000 our business had grown to over $300 million in size, dwarfing our nearest competitor by over $200 million a year. In February 2000, right before the DotCom meltdown Sterling was acquired by Computer Associates for over $2 billion — almost a 3x revenue multiple and a 7x operating profit multiple.
Originally published on Medium
Also published on Medium.