I was responsible for a portfolio of roughly $200 million spread across several studies.
Like many Clinical Operations leaders, I was expected to have a good handle on where spending was headed. Not just this month, but six months from now. Sometimes a year from now.
And if you’ve ever been in that position, you’ve probably heard questions like:
● Are we still on budget?
● Will we finish within forecast?
● How much confidence do you have in that number?
Fair questions.
The uncomfortable part is that sometimes the most honest answer is:
“I don’t know.”
Simply because there are things you don’t know yet.
The Forecast Is Only as Good as the Assumptions
Early in a study, a surprising amount of the budget is built on assumptions.
Enrollment assumptions. Site assumptions. Grant assumptions. Resource assumptions. Vendor assumptions.
Some turn out to be accurate. Some don’t.
A CRO may estimate grants based on the information available at the time. Then the study starts, sites are engaged, and actual grant expectations look very different.
Enrollment may be slower than expected. Or faster.
A protocol amendment may arrive earlier than anticipated.
Additional countries may be needed.
Anyone who has worked in clinical development long enough has seen some version of this.
None of it is unusual. It’s simply the nature of the work.
Forecasts Are Not Promises
Over the years, I’ve become less interested in whether an original forecast was exactly right and more interested in understanding what changed.
Sometimes we treat the first budget as though it should have predicted the future.
I’m not sure that’s a reasonable expectation.
The initial forecast is usually someone’s best estimate based on the information available at the time. As new information appears, the estimate should change.
In fact, I’d be more concerned about a forecast that never changes than one that does.
A forecast that never changes may simply mean nobody is challenging the assumptions behind it.
Where Vendor Management Fits In
Sponsors often expect vendors to provide accurate forecasts very early in a study. Vendors do their best, but many of those forecasts are based on assumptions that have not yet been tested.
That doesn’t mean vendors shouldn’t be accountable. They should.
But accountability and predictability are not the same thing.
I’ve occasionally seen sponsors become frustrated when spending diverged from the original plan, only to discover that the assumptions behind the forecast had changed months earlier.
The more productive conversation is often:
● What assumptions were made?
● Which assumptions changed?
● When did we know they had changed?
● How quickly did we communicate the impact?
● What actions did we take once we knew?
Those discussions tend to reveal much more than debating whether a forecast missed by a few percentage points.
What Good Teams Do Differently
The strongest teams I worked with weren’t necessarily the teams that predicted everything perfectly.
They made their best estimate; documented their assumptions; monitored leading indicators; compared forecasts against actuals; discussed risks early; learned from new information; and adjusted accordingly.
Nothing magical. Just discipline.
Over time, forecasts became more accurate because the teams understood the study better.
Not because uncertainty disappeared.
A Thought on Forecasting
A book that recently caught my attention is Superforecasting by Philip Tetlock.
One of its central ideas is that strong forecasters don’t necessarily succeed because they predict the future perfectly. They succeed because they’re willing to update their views as new information becomes available.
That feels familiar.
The strongest Clinical Operations teams I’ve worked with didn’t treat forecasts as fixed commitments. They treated them as working estimates that improved over time. As assumptions were validated or disproven, forecasts were adjusted accordingly.
Perhaps that’s how we should think about forecasting in clinical trials.
Not as an exercise in certainty, but as an exercise in learning.
Final Thoughts
When you’re managing large studies, there is pressure to have answers.
Sometimes that pressure can make us sound more certain than we really are.
But certainty and accuracy are not the same thing.
Sometimes the most honest answer is:
“I don’t know.”
Followed by:
“Here’s what we know today. Here’s what we’re watching. And here’s how we’ll know more next month.”
I’ve found that approach tends to build more trust than pretending to know something that nobody could reasonably know yet.
About the Author Kalyan Obalampalli (KO) is the Founder and President of Clin.AI, a disruptive start-up transforming how biotech and pharma sponsors select and manage vendors. With more than 20 years of leadership experience in clinical operations and outsourcing at major pharma and biotech companies, KO brings a unique perspective on how technology can close long-standing gaps in clinical development.
