Investment example 1 · Portfolio reallocation
A single investment model was never going to work equally well everywhere.
At OpenText, we were planning against a significant pipeline requirement across a large portfolio. Going into FY25, the business needed approximately $6.5 billion in open pipeline against roughly $5.2 billion in current pipeline and outlook. Underneath that company number were very different businesses, buyers, sales motions, and points of demand.
A single investment model was never going to work equally well across all of them.
I had become concerned that one part of the portfolio, Experience Cloud Solutions, was benefiting from advantages that made its performance difficult to compare directly with other businesses. In my view, corporate event investment and the SAP relationship were contributing to a halo around the business.
I could not prove every part of that hypothesis yet. But I had enough indicators to believe we should act.
So I shifted investment toward other business units where I believed additional support could make a greater difference.
Example 1 · Testing the decision
Act, then create the measures that can prove you wrong.
That was the first decision. The next step was just as important.
I did not want the organization changing direction every time a new data point appeared. Once we made the shift, I asked the team to keep investigating the assumptions behind it. We needed to understand whether the original judgment held up and whether the businesses receiving additional investment began to respond.
My recollection is that Experience continued to perform strongly while some of the businesses receiving greater attention began to improve. That gave me more confidence in the direction, but we continued testing rather than treating the decision as settled forever.
Investment example 2 · Channel mix
Let each business determine the mix.
A separate investment question was how much each business should place into events compared with digital, account development, and organic programs. The answer was not the same across the portfolio.
In FY25, events represented more than 60% of the last-touch mix in IT Operations Management and roughly half in Cybersecurity. The shares were lower in Content, Business Networks, and Experience. Some businesses received more event support. Others held or reduced events while adding capacity in other areas.
Then we changed the mix again.
By FY26, event share had come down across Content, Experience, Business Networks, IT Operations Management, Application Delivery Management, Cybersecurity, and Corporate Sales. In many of those areas, the balance moved toward account development and organic programs.
That was not an admission that the earlier decisions were wrong. It was the point of the process. Make the best decision you can from the commercial evidence available, watch what happens, learn, and adjust again.
Performance context · Across both examples
Read the response without claiming one cause.
The broader performance record provides context for these choices, but it is not proof that any one investment decision produced the result. OpenText Q4 FY24 pipeline reached approximately $679.6 million against a $605 million target, or 112% of plan. By June 2025, Content, Experience, and Business Networks were among the stronger performers, while several other businesses were described internally as turning a corner.
I would not attribute that performance to one budget decision. Sales execution, market conditions, programs, leadership, portfolio dynamics, and many other factors contributed. Investment evidence should help a leader make a better next decision, not create a causal claim the data cannot support.
What connects the examples
Waiting for perfect information is still a decision.
These decisions changed how I think about executive judgment.
Waiting for perfect information sounds disciplined, but sometimes waiting is itself a decision. If the business needs a change, a leader has to be willing to form a view from the signals available and act.
The discipline comes afterward.
Be clear about why you made the decision. Create the measures that can prove you wrong. Give the change enough time to teach you something. Then have the willingness to hold the course, adjust it, or reverse it as the evidence develops.
I became much more comfortable making decisions that way. Not guessing. Not pretending the data was more certain than it was. Making a reasoned choice, then deliberately learning whether we had made the right one.