The question I’d like to pose to my colleagues in the audit space, particularly employee benefit plan audits, is does it still make sense to offshore this work? Let me explain.
Employee benefit plan audits run on data; mountains of it. And to deal with it all, every year, audit teams would spend weeks doing work that felt more like data processing than professional judgment.
The traditional workload involved:
- Reconciling recordkeeper files.
- Cross-referencing numbers across systems.
- Populating workpapers that followed the same structure every single time.
So, when firms decided to offshore this work, it seemed they made a reasonable call. After all, if the work is repetitive and rule-based, and there are people overseas who can do it accurately at a fraction of the cost, why would we pay domestic staff to sit there and do it? The math was hard to argue with.
Offshoring spread through the profession quietly, but quickly, and most firms that went down that road will tell you it worked. At least for a while. The question worth asking now is whether something has shifted underneath the assumption that it would only be “for a while.”
What Has Changed?
Here’s what’s changed: The work that firms sent offshore for the above reasons is now the same work that current technology itself does faster, with fewer handoff points, no time-zone friction, and without institutional knowledge walking out the door every time a team turns over. When you dig deeper into “the work” it’s hard not to argue for automation over offshoring.
These tasks no longer require a person halfway around the world:
- Reconciling participant data from recordkeepers.
- Running contribution testing.
- Cross-referencing distributions and rollovers.
- Formatting and populating workpapers.
In fact, if done right, they barely require a person at all other than review. To be clear, I’m not criticizing the firms that offshore this work or the offshore entities themselves. It’s just an observation about timing.
Automation vs. Offshoring
As I mentioned, the decision to offshore made sense when the alternative was paying a domestic senior associate to spend two weeks on something mechanical. Moreover, it makes less sense when the alternative now is a system that:
- Processes the same data in hours.
- Flags anomalies automatically.
- Feeds directly into workpapers that your team actually reviews rather than just assembles.
The Knowledge Problem
The deeper issue is the offshoring costs that don’t show up on a spreadsheet. When work leaves the building, so does the learning that comes from doing it. Junior staff who never touch the data processing side of an audit develop differently than those who do. They miss the pattern recognition that comes from handling thousands of participant records, seeing where the outliers tend to cluster, developing an instinct for what doesn’t look right.
In the end, offshoring solved a cost problem while quietly creating a knowledge problem, and that trade-off was easy to ignore when the output looked clean. Technology doesn’t have that trade-off. When you automate the mechanical work internally, the people who used to do it don’t disappear, they move up. In fact, they spend most of their time on:
- The judgment calls.
- The exceptions.
- The conversations with plan sponsors about what the numbers truly mean.
The firm also gets faster and the team gets sharper. That’s a different proposition than what offshoring ever offered.
Looking Ahead
None of this means every firm that offshored made the wrong decision for its moment. Businesses work with the options available to them, and for a long time, offshoring was the most viable option available.
But the options have changed, and the firms that recognize this earliest are going to look very different in five years from the ones still running the same model.
The interesting question now isn’t whether to offshore or not, it’s whether the assumption that made offshoring logical in the first place still holds. Moreover, what does it mean for your practice if it doesn’t.
That’s worth sitting with.
