Operational Due Diligence is relatively easy to measure operationally. How many managers were reviewed? Were reviews completed on time? Is the annual plan on schedule? How many findings were raised? How quickly were reports delivered?
Those numbers are useful. They do not tell you whether the ODD process is good.
A programme can complete every scheduled review and still produce inconsistent judgement, spend too much senior time on routine work, miss important organizational change or generate reports that investment committees struggle to use. The better question is whether the ODD function is improving the institution's ability to make and maintain confident delegation decisions.
That requires looking beyond activity.
Completion is a very low bar
There is an understandable tendency to assess an ODD programme through its workflow. Reviews enter the pipeline. Documentation is collected. Interviews take place. Reports are issued. Findings are tracked.
“A low number of findings therefore tells us surprisingly little about whether the process is working. So does a high number.”
A well-run process should do all of those things reliably. But operational efficiency and ODD effectiveness are not the same thing.
The purpose of ODD is to help an investor determine whether an organisation remains sufficiently governable for capital delegation and whether that delegation can be supported within its institutional governance framework. That means a better measure of the function is not simply whether reviews happened. It is whether the reviews changed, strengthened or confirmed the investor's understanding in a way that made the eventual judgement more reliable.
Occasionally, the best ODD review will identify a serious concern. Quite often it will not. A low number of findings therefore tells us surprisingly little about whether the process is working. So does a high number.
The quality lies in whether the right things were recognised and given the right weight.
Look for consistency without expecting identical judgement
One of the first things we would examine is what happens when different people review broadly comparable situations.
Professional judgement should never be completely standardised. Different managers have different structures, histories and circumstances. But reviewer identity should not determine the risk outcome. If one reviewer treats a particular dependency as material while another routinely accepts something very similar without discussion, the institution should be able to explain why.
Sometimes there will be a perfectly good reason. Sometimes there will not.
As ODD programs grow, this becomes harder. More managers create more information, more reviewers and more individual judgement calls. Lestrade's underlying methodology was developed partly in response to this problem: without a stabilising framework, judgement can become increasingly influenced by information volume, recent observations, personal preferences or the areas in which individual reviewers have particular expertise.
A finding is not automatically a reason not to invest
Most asset managers are not operationally perfect. They depend on people. They use third parties. They operate legacy systems. They make governance compromises. Processes evolve faster than documentation. Some controls are more manual than an investor would ideally prefer.
The presence of an issue is therefore not particularly informative on its own. What matters is the mechanism through which that issue could affect the organisation.
A small team, for example, may create key-person dependency. Or it may be entirely appropriate for a focused organisation with clear responsibilities, strong succession arrangements and good oversight. A manual process may represent weak scalability. Or it may be a sensible control in a low-volume environment where automation would add complexity without meaningfully reducing risk.
Good ODD should distinguish between something that is unusual, something that deserves attention and something that materially changes the operational risk of delegating capital. That distinction is where judgement begins.
Materiality depends on context
This is one reason ODD findings do not translate neatly into investment rules. The same observation can mean different things in different organisations.
Imagine two managers with significant dependence on a senior operations executive. At the first, that individual controls several important processes personally, succession arrangements are informal and the wider team has limited decision-making authority. At the second, the executive remains influential but responsibilities are distributed, deputies are experienced and governance does not depend on the individual being present.
The headline finding could be described similarly. The structural reality is not.
This is why institutional investors should be cautious about treating ODD as a checklist in which certain observations automatically produce predetermined outcomes. The purpose of ODD is to understand whether the organisation remains sufficiently governable for capital delegation under uncertainty. That requires context around the finding, not simply classification of it.
The first question should be: what is the residual risk?
A finding becomes decision-relevant when the investor understands what remains after the manager's mitigants are taken into account. Consider an organisation that relies heavily on a critical third-party provider. The dependency itself may be obvious. The more important questions are what the organisation has done about it.
Is the provider monitored properly? Are responsibilities clear? Could the function be transferred if necessary? Does management understand the concentration risk? Have failures or service issues occurred before?
The ODD conclusion should therefore go beyond identifying the dependency. It should form a view on the residual risk created by that dependency after considering how the organisation manages it. This is a more useful starting point for an investment decision because investors delegate to real organisations, not theoretical ones without weaknesses.
The question is rarely whether risk exists. It is whether the remaining risk is acceptable.
But risk is only half the picture
There is another dimension that is easy to lose once a finding has been assigned a rating. How certain are we?
Two managers may receive broadly similar residual-risk conclusions while the evidence behind those conclusions is very different. In one case, documentation, operational artefacts and interviews all point in the same direction. In another, the conclusion may rely heavily on management's description because independent evidence is limited.
Lestrade's methodology deliberately separates residual risk from confidence in the evidence supporting that assessment. That distinction can have very practical consequences. A moderate risk understood with high confidence may be easier for an institution to accept than a superficially low-risk conclusion supported by weak evidence.
“The real constraint in ODD is often experienced judgement. An efficient ODD process should protect senior judgement capacity rather than consume it.”
A healthy ODD process therefore produces conclusions that are comparable enough to defend, while remaining sensitive to the reality of each manager. That balance is more important than perfect standardisation.
Ask where your senior people's time goes
This is one of the most practical tests, and one that institutions can usually answer quite quickly. What are the most experienced ODD professionals actually doing?
If a large proportion of their week is spent coordinating interviews, chasing documentation, formatting reports, checking routine disclosures or repeatedly explaining administrative requirements, the function may be well staffed and still be using its scarcest resource poorly.
The real constraint in ODD is often experienced judgement. As complexity and review volumes increase, teams can develop growing backlogs, greater dependence on senior reviewers, increasing documentation effort and difficulty maintaining calibration. Technology helps with some of this. So can outsourcing. But neither is valuable simply because it makes individual tasks faster.
The more interesting test is whether they release experienced people to spend more time where their judgement matters: difficult interviews, conflicting evidence, structural dependencies, escalation and the interpretation of ambiguous situations.
An efficient ODD process should protect senior judgement capacity rather than consume it.
Does the process distinguish information from evidence?
This is harder to observe from management information, but it tells you a great deal about the maturity of an ODD programme.
Asset managers can provide enormous amounts of information. The question is what the reviewer does with it. A DDQ tells you how the organisation describes itself. A policy tells you how a process is supposed to work. An operational artefact may show what actually happened. An interview reveals how people understand their own responsibilities.
These are not interchangeable forms of evidence. And when they disagree, the disagreement may be more informative than any individual source. Lestrade's evidence architecture deliberately looks for convergence, contradiction and incompleteness across different evidence sources rather than treating information volume itself as assurance.
You can often see the difference in the resulting work. A weaker process produces a thorough description of the manager. A stronger one explains why the reviewer believes that description represents operational reality — and where there are reasons to remain cautious.
That is a much more useful form of assurance.
Good ODD should occasionally change its mind
There is a slightly uncomfortable test of process quality. Does new evidence ever change the initial view?
“Otherwise diligence risks becoming an elaborate way of documenting the first impression.”
If every interview merely confirms the DDQ, every document supports the existing narrative and every review ends roughly where it began, the process may be more confirmatory than investigative.
Good ODD should contain the possibility of surprise. An apparently worrying issue may turn out to be well controlled once the operating evidence is understood. A reassuring policy may become less comforting when people describe how responsibilities work in practice. A small contradiction may lead to something important.
This does not mean reviewers should search for problems. It means the process should genuinely allow evidence to move the judgement in either direction. Otherwise diligence risks becoming an elaborate way of documenting the first impression.
Is uncertainty visible, or does it disappear into the report?
Another useful test is to look at conclusions that appear reassuring. How much evidence sits behind them?
Complete information is structurally unattainable in ODD. Documents can be incomplete, processes change and interviews provide partial perspectives. A mature process does not pretend that these limitations vanish simply because a final report has been issued. Lestrade therefore separates residual risk from confidence in the evidence supporting that assessment. A manager may be assessed as relatively low risk while confidence remains weaker because important evidence is unavailable or contradictory.
“The report is not the end product. The decision it supports is”
That distinction is particularly useful when evaluating an ODD programme. If every conclusion appears equally certain, regardless of the quality of the underlying evidence, the process is probably hiding information that decision-makers should see.
Good governance does not require artificial certainty. It requires knowing where certainty stops.
Can the investment committee actually use the output?
This may be the simplest test of all. Take a completed ODD report and give it to a senior decision-maker who was not involved in the review.
Can that person understand what matters? Can they see why it matters? Can they distinguish a benign difference from a genuine operational concern? Do they understand where evidence is strong and where uncertainty remains? And is it reasonably clear what the institution should do with the conclusion?
If answering those questions requires the ODD team to provide a second twenty-minute explanation, the report may contain excellent analysis without yet being fully usable.
Institutional usability is part of ODD quality. Your positioning document makes this explicit: rigorous ODD should remain practical to use, integrate into existing institutional processes and produce work that is usable by investment committees rather than creating unnecessary operational burden.
The report is not the end product. The decision it supports is.
Watch what happens after a finding is raised
A good ODD process also has a memory. If an issue is identified, what happens next? Was it accepted? Remediated? Monitored? Did the manager make the change it promised? Did a similar issue appear again later?
This is where the difference between individual reviews and an ODD programme becomes visible.
An institution may conduct excellent point-in-time diligence yet lose important information between reviews. A finding that looked manageable two years ago may become more significant when it forms part of a pattern. A remediation that seemed convincing when proposed may look different when implementation repeatedly slips.
The function should therefore become more informed about a manager over time, not repeatedly rediscover the organisation from the beginning. If institutional memory sits mainly in the heads of a few experienced reviewers, the process is more fragile than it appears.
Perhaps the best test is whether ODD makes judgement easier
There are many metrics an institution can use to monitor its ODD function. Review timeliness matters. Capacity matters. Coverage matters. Findings and remediation should be tracked.
But none of them captures the central question.
After years of doing ODD, is the institution becoming better at deciding what really matters? Are similar issues judged more consistently? Are senior professionals spending more time on difficult judgement and less on routine processing? Can uncertainty be expressed rather than hidden? Do reports help decision-makers understand risk without reconstructing the entire review? Does the organisation learn from previous findings? And when evidence changes, can the judgement change with it?
Those are signs of an ODD process that is becoming more mature. The goal is not to complete diligence efficiently for its own sake. It is to create a system in which reliable judgement can be produced repeatedly, across different managers and different reviewers, even as the environment becomes more complex.
That is a higher standard. It is also a much more useful one.
