Your Latest Best Estimate (LBE) forecast is the perfect partner to your Annual Operating Plan (AOP). Well, they are perfect partners when they work together effectively, which isn’t always – or even usually – the case. To learn more about what teams get wrong and how they can get their LBEs and AOPs working together in harmony, I spoke to Ufuk Ezer, Director, Implementation at J+D Forecasting, who works with pharmaceutical and biotech companies on LBE process design. You can read the full interview within our LBE guide here.
How do teams keep one evolving base case running through the year, and where does that usually slip?
The most common slip is treating the base case as something you update once a month when the LBE cycle opens, rather than something that lives and breathes throughout the year. The idea is that you are always refining it as new actuals come in, as market signals shift, as assumptions are proved right or wrong – and then you lock it at a defined point to create that snapshot.
What typically breaks down is the locking discipline. Either teams forget to maintain version control, so the original AOP baseline gets overwritten and lost, or the working base case stops being updated between cycles because the team is stretched. When that happens, you arrive at the LBE cycle with an outdated starting point, and the process becomes a patching exercise rather than an honest re-forecast. The fix is simple in principle but requires real discipline: make the working base case a live document with a clear owner.
How does variance analysis feed into the next forecast, and how do you get teams to use it that way?
Variance analysis only feeds forward if teams treat it as a learning exercise rather than an accountability exercise. If variance review feels like it is primarily about explaining why you missed, people will manage the conversation rather than interrogate the assumptions. The right framing is: what does this variance tell us about the assumptions we are carrying forward? That is what changes the next forecast.
The question about what action should follow is the one that gets skipped – teams identify the variance, explain it, and move on. But if no action follows from a foreseeable miss, you tend to see the same miss being repeated.
Getting teams to actually use variance analysis that way requires it to be built into the process, not treated as an optional retrospective. It should be a standing agenda item, with outputs that feed directly into the assumption revisions for the next cycle.
How do you stop the LBE becoming a way to reset expectations when results are under pressure?
It is worth being direct about what is actually happening when the LBE starts moving in this direction. It is not always intentional. Teams face real pressure, results are coming in below plan, and the natural instinct is to update the forecast to reflect where things are heading. The problem is that once the LBE absorbs all the bad news, you lose the ability to measure performance against the original commitment. The baseline disappears.
Practically, this requires two things. The first is process discipline: the original AOP should be locked and version-controlled in a way that is structurally separated from the LBE working file. The second is governance: leadership needs to be explicitly comparing LBE versus AOP, not just reviewing the LBE in isolation. When the gap between the two is visible in every review, it creates accountability. When it is hidden, teams can manage expectations downward without anyone explicitly choosing to do so.
The full article covers the complete LBE process in depth, including cadence design, risks and opportunities registers, and how model design either supports or undermines the whole thing.
Read the full article here or get in touch to learn how we can support your forecasting processes.
