We’ve seen a lot of search term sets over the years. Long ones, short ones, carefully crafted ones, and ones that clearly started as a good idea and then sort of spiraled. But every once in a while, a client comes to us with something that makes you stop, blink, and reach for your coffee.
This is one of those stories.
Two PMs, Two Visions, One Giant Tangle
The client arrived with a dataset and a search term list that had been built — as best we could tell — by their project managers, but the best we could tell there were two of them working from two very different perspectives. There was no apparent coordination between them, no shared framework, and no overarching strategy. What they had instead was the eDiscovery equivalent of two people trying to give directions to the same destination from different cities, in different languages, at the same time.
The result was a list that read less like a legal search strategy and more like a stream of consciousness. Terms were added as they came to mind. Then more terms. Then more again. By the time it landed on our desk, they had gone from 300 search terms to 700 — and they weren’t done yet.
We will admit: when we first pulled it up, we laughed. Not unkindly — but genuinely, because the sheer ambition of the thing was impressive.
Then we got to work.
The First Question We Always Ask
Before we could begin untangling anything, we had to go back to basics. We sat down with the client and asked the question that should have anchored the whole effort from the start:
What do you actually want to find?
It sounds almost too simple. But when you’re deep in a matter, juggling custodians and data volumes and deadlines, it’s surprisingly easy to lose sight of that fundamental question. Search terms become a way of hedging — if we include enough of them, surely we’ll catch everything, right?
Not quite.
A Dataset Is Not the Internet
Here’s a misconception we run into more than we’d like: people approach eDiscovery search the way they approach Google. If one term doesn’t work, try five more. Cast the net as wide as possible. Don’t leave anything out.
But a closed dataset doesn’t work like the internet. It has structure, context, and boundaries and it requires accuracy. In that environment, a scattershot approach — particularly one driven primarily by attorneys doing ad hoc searches rather than search specialists building a coordinated strategy — doesn’t produce better results. It produces noise. Enormous, expensive, time-consuming noise.
In this case, roughly 95% of the search terms had been entered by attorneys doing their own searches. Each one, individually, had logic behind it. But there was no big-picture thinking. No sense of how the terms related to each other or to the data. The result swung wildly between terms that were so broad they were nearly meaningless and terms so narrow they were almost certainly useless.
Overly broad. Entirely unhelpful. Often both at the same time.
This Isn’t as Unusual as You Might Think
We want to be clear about something: this situation was particularly bad. But the underlying pattern — search terms built without a unified strategy, without proper scoping, without someone who truly understands both the legal objectives and the technical realities of the dataset — is far more common than most people admit.
We’ve seen it in large firms and small ones. In first-time litigants and seasoned legal departments. It tends to happen when eDiscovery is treated as a task to be delegated rather than a process to be managed — when no one person owns the strategy from start to finish, and when the people constructing the search terms aren’t the ones who best understand how searches are done.
That’s not a criticism. It’s a description of how things go sideways when expertise isn’t in the room.
The Bigger the Mess, the More Expertise It Takes to Untangle It
This is the part that matters most.
When everything is going smoothly, expertise can feel optional. You might not notice it at all. But the moment things go wrong — when the search terms are a tangle, when the strategy has fractured, when the data and the objectives no longer seem to be speaking the same language — that’s exactly when real expertise becomes essential.
Untangling a mess like this one requires more than just cleaning up a list. It requires understanding what the client actually needs, mapping that to what the data actually contains, and rebuilding a search strategy that is targeted, defensible, and proportionate. It means knowing what to cut, what to keep, and what questions to ask along the way.
That’s the work we do. And it’s the kind of work that makes a genuine difference — not just in the quality of results, but in the cost, the efficiency, and the confidence with which a client can move forward.
When You Need It Most
We think of eDiscovery a bit like any skilled trade. When things are running smoothly, you might not need us. But when you’ve got a pipe burst at 2am — or 700 search terms and no clear strategy — you want someone who’s seen it before, knows exactly what to do, and can get things back on track without making them worse.
That’s what we’re here for. The messy ones are actually our specialty.