The Hidden Cost of "Ask Sarah": Single Points of Knowledge Failure

Every company has a Sarah

Every business past a certain size has one. Ask about the weird vendor contract clause, ask Sarah. Ask why the client onboarding checklist has that one unusual step, ask Sarah. Ask what actually happens when a shipment gets delayed past the promised date, ask Sarah, because she has handled it a dozen times and everyone else has handled it zero.

Sarah is not a problem. Sarah is what happens naturally when someone capable spends years solving the same category of problem repeatedly and gets very good at it. The problem is what the organization does, or more often does not do, in response to Sarah's existence. Most companies treat "ask Sarah" as a solved problem rather than what it actually is: an unpriced, unmanaged risk sitting quietly on the business, growing larger every year Sarah stays and every year nobody writes down what she knows.

Why this risk is so easy to ignore

Single points of knowledge failure are structurally easy to ignore because the cost is invisible until the exact moment it is not. Every single day Sarah is at her desk, the arrangement looks completely fine, arguably efficient. Questions get answered fast, by the most qualified person, with no process overhead. It looks like the system is working.

The cost only becomes visible in a moment nobody scheduled: Sarah takes two weeks of leave and a decision that used to take five minutes takes three days and involves three people improvising. Sarah gets a better offer and leaves with six weeks notice, and the business discovers, in real time, how much of a specific function existed only in her head. Sarah gets sick unexpectedly. None of these are exotic scenarios. They are close to certain to happen eventually to some key person in any organization that runs long enough, and yet very few SMEs treat this as the "when," not "if," problem it actually is.

This is exactly why the risk tends to stay unmanaged. It does not show up on a P&L as a line item. It shows up, occasionally and unpredictably, as a bad week, a scramble, a client complaint that gets attributed to "we were short-staffed" rather than to its actual root cause, which is that critical knowledge had exactly one point of failure and that point failed.

What it actually costs when it happens

It is worth making this concrete rather than leaving it abstract, because abstract risk is easy to deprioritize against a concrete, near-term budget line.

Direct productivity loss. Every question that used to take five minutes with Sarah now takes hours of trial and error, escalation, or waiting, multiplied across everyone who would have asked her. For a genuinely central knowledge-holder, this can meaningfully slow an entire function for as long as the gap persists.

Decision quality degradation. The judgment calls Sarah made instinctively, the exceptions she knew to make, the red flags she recognized before they became real problems, do not get made as well by whoever is filling in, because that judgment was never transferred. This shows up as worse outcomes, not just slower ones: a client relationship handled clumsily, a risk that would have been caught early instead becoming a real incident.

Client-facing damage. If Sarah was client-facing, or her knowledge underpinned how client-facing staff operated, her absence often becomes visible to clients directly, in slower responses, inconsistent answers, or an obvious drop in the quality of service the client had come to expect. This is the most expensive version of the cost, because it risks the relationship itself, not just internal efficiency.

Onboarding drag on her eventual replacement. If Sarah leaves permanently, whoever replaces her (internally promoted or externally hired) faces a much longer ramp-up than necessary, because there is no structured record of what she actually knew, only a job title and whatever fragments colleagues remember secondhand. The replacement essentially has to rebuild years of judgment from scratch, in real time, under the pressure of live client and operational demands.

Why this is a business continuity issue, not just an HR issue

Framing this purely as an HR or training gap understates it. This is a business continuity risk in the same category as a data backup failure or a single-supplier dependency, and it deserves to be evaluated with the same seriousness. A business that would never tolerate having all its financial records exist in exactly one uncopied file will often tolerate, without a second thought, having a critical operational function exist in exactly one person's head with no backup at all.

The comparison holds up under scrutiny. A backup failure and a knowledge-holder departure produce the same category of damage: sudden, hard-to-recover loss of something the business depended on daily. The fact that one is a technical failure and the other is a human one does not make the second any less worth planning for; if anything it is more predictable, because people leave companies at a far higher rate than well-maintained backups fail.

How to actually quantify the exposure

Most SMEs have never tried to size this risk, which is part of why it stays deprioritized. A rough but genuinely useful exercise: list your five or six most senior or longest-tenured staff. For each one, ask honestly, what would break, and for how long, if this person were unreachable for a month starting tomorrow. Not "would we manage," everyone manages eventually, but specifically what would slow down, what quality would drop, and what would a client or a critical process actually experience during that month.

This exercise almost always produces a shorter, more specific list than people expect, usually two or three genuinely high-exposure people per functional area rather than the whole team. That specificity is useful, because it turns a vague, uncomfortable feeling ("we're too dependent on a few people") into a concrete, prioritized list you can actually act on, starting with whoever scores highest on both likelihood of departure and depth of exposure.

The version of this risk that is easy to miss

Most conversations about key-person risk focus on the obvious candidates: the most senior person, the longest-tenured person, the founder. Those are important, but they are also the easiest to spot, and precisely because they are obvious, many businesses have at least some informal contingency thinking around them already.

The riskier gap is often one level down: a mid-tenure staff member who has quietly become the only person who understands a specific, unglamorous but critical process. The one person who knows how to reconcile a particular report when the numbers do not match. The one person who understands why a specific client's account is configured differently from every other client's, and what breaks if that difference is not respected. These dependencies accumulate gradually, nobody decides to create them, and because the person holding them is not necessarily senior or visible in the org chart, nobody flags the risk until it surfaces the hard way.

Part of doing the exposure exercise honestly is looking past job titles and asking, function by function, who is actually the only person who fully understands how a given process works end to end, regardless of where they sit on the org chart. That question surfaces a different, often more concerning, list than simply asking who has been here longest.

A note on what this is not

It is worth being explicit about what capturing this knowledge does not mean, because the framing matters for how staff experience the process. This is not about making any individual dispensable in a way designed to justify letting them go, and staff will sense immediately if that is the subtext, which will make them guarded rather than genuinely forthcoming in a capture session. It is closer to insurance: something you do specifically so that the business does not lose critical capability the day someone is unexpectedly unavailable, whether for a happy reason (promotion, a well-earned holiday) or an unhappy one.

Framed and communicated honestly, this kind of knowledge capture is something most experienced staff actually welcome, because it validates their expertise as something worth preserving deliberately, rather than something the business simply takes for granted until the day it is gone.

The fix does not require Sarah to leave first

The instinct many businesses have is to treat knowledge capture as something you do reactively, once someone has already given notice, in a rushed handover period that is never long enough to transfer years of judgment properly. That is the worst possible time to attempt it: the departing person is disengaged, distracted by their transition, and the pressure to finish fast works against the depth the capture actually needs.

The better approach is proactive and unglamorous: capture the judgment of your highest-exposure people while they are still fully engaged and have no reason to be anything other than thorough, structured around the specific recurring questions and decisions that make them the bottleneck in the first place. This does not require a company-wide initiative. It can be scoped narrowly to the two or three people identified in the exercise above, run as a focused set of structured sessions rather than an open-ended project.

Done this way, the goal is not to make Sarah replaceable in some diminishing sense, and it is not a signal of distrust in her. It is closer to backing up a system that is currently running on a single, uncopied drive. Sarah remains exactly as valuable. The business simply stops being one bad week away from losing access to what she knows.

This is precisely the risk Decisionlore's boss and key-staff interview workshop is built to address before it becomes an emergency, capturing the judgment of your most exposed people while they are still there to get it right. If this exercise surfaced a name or two for your business, our pricing page shows what a capture engagement looks like, or you can get started directly.