Key Takeaways
- The phrase "let me get back to you" appears harmless, but it signals a systemic failure in knowledge accessibility that costs organizations thousands of hours in lost productivity.
- By relying on human gatekeepers for routine answers, companies incur an "interruption tax," create consistency errors, and distract experts from strategic work.
- Modern AI-powered knowledge bases are specifically designed to solve these problems by delivering instant, verified answers from your existing policies and documents.
Someone asks you a question. You don't know the answer off the top of your head.
So you say the thing everyone says: "Let me get back to you."
It seems harmless. Professional, even. You're not making something up. You're going to find the right answer and follow up.
But think about what actually happens next.
You make a mental note to find the answer. Or maybe you write it down somewhere. You go back to whatever you were doing before you were interrupted. Later—maybe hours later, maybe the next day—you remember you owe someone an answer.
You dig through your files, ask someone else, or search the intranet. You find what you think is the right information. You double-check because you're not certain. You compose a response. You send it.
That's 15 minutes of your time, minimum. Often more. For one question.
Now multiply that by however many times it happens in a week. In your department. Across your organization.
The cost isn't just your time. It's the person who asked, who's either blocked, waiting for your answer, or has moved on and forgotten why they needed it. It's the cognitive load of carrying around open loops. It's the inconsistency when different people get different answers to the same question, depending on who they ask and when.
"Let me get back to you" feels like a reasonable response. It's actually a symptom of a knowledge problem that's costing you more than you realize.
The "Answer Person" Bottleneck
In most organizations, certain people become the answer people.
You know who I'm talking about. The HR person who's been there fifteen years and knows the history behind every policy exception. The operations manager who understands how things actually work versus how they're supposed to work. The engineer who remembers why the system was built that way.
These people are invaluable. They're also bottlenecks.
The Interruption Tax
Studies suggest it takes over 20 minutes to fully recover focus after an interruption. If your answer person gets interrupted ten times a day, that's most of their productive time gone.
Source: Gloria Mark, UC Irvine ResearchEvery question that comes to them is an interruption. If your answer person gets interrupted ten times a day with questions, that's not ten small disruptions—it's most of their productive time gone.
And they can't scale. There's only one of them. When they're in meetings, on vacation, or just overwhelmed, the questions pile up. People either wait, find workarounds, or make decisions without the information they need.
The answer person didn't sign up to be a human FAQ. They have actual work to do—strategic work, probably, given their experience. But they can't get to it because they're too busy answering the same questions they answered last week.
The Cost of Waiting
The person asking pays a cost, too.
Sometimes they need the answer to move forward. A decision is waiting. A customer is waiting. A process is stalled. "Let me get back to you" means they're stuck until you do.
So they either wait—which delays whatever they were trying to accomplish—or they guess. They make the decision with incomplete information. Sometimes that's fine. Sometimes it leads to mistakes that take much longer to fix than answering the question would have taken.
Even when the stakes are low, there's a tax on momentum. You're in the middle of something, you hit a question you can't answer, you have to stop and ask someone, and now you're waiting.
Context Switching
The context you had in your head evaporates. When the answer finally comes, you have to reload where you were and what you were doing.
For some questions, people don't even bother asking anymore. They've learned that getting an answer takes too long or is too much hassle. So they just don't do the thing, or they do it wrong, or they do what they think is right and hope for the best.
That's not a failure of the employee. That's a failure of the system.
Calculating the Hidden Expense
Let's try to put some rough numbers on this.
Say you have a 5-person HR team supporting a 500-person company. Each HR person probably gets 10–15 routine questions per day from employees. Policy questions. Benefits questions. Process questions. "Who do I talk to about X?" questions.
Each one is an interruption. Even if they answer immediately from memory, that's 2–3 minutes per question—finding the thread, parsing what's being asked, formulating a response. If they have to look something up, it's 5–10 minutes. If they need to ask someone else, it's longer, and now you've interrupted two people.
The Math of Inefficiency
At an average of 7 minutes per question and 60 questions per day, a single HR team spends around 1,800 hours per year on routine answers—roughly $90,000 in fully loaded costs.
(Calculated example)And that's just HR. Add finance, IT, operations, legal, and all the other departments that field routine questions, and the number gets significantly larger.
The Consistency Problem
The consistency cost might be even worse than the time cost.
When answers live in people's heads, different people give different answers. Not because anyone's wrong, necessarily, but because policies have nuance and memory has gaps and people interpret things differently.
Employee A asks about the work-from-home policy and gets one answer. Employee B asks a different person and gets a slightly different answer. Now you have an inconsistency, and someone's going to be upset when they find out their colleague got a different deal.
Version Control Chaos
Someone answers based on how the policy used to work, before it was updated last year. They're not trying to give bad information—they just didn't register the change. Now an employee is operating on outdated guidance.
The answer person might be consistent with themselves, but they're not the only one answering questions. And even they might give different answers on a busy day versus a calm one, or remember details differently over time.
When knowledge lives in a system instead of in heads, everyone gets the same answer. The answer that reflects current policy, not remembered policy. The answer that's been vetted, not improvised.
The Real Solution: Self-Service
The fix isn't to tell people to stop asking questions.
They have questions because they need information to do their jobs. Penalizing them for asking makes them stop asking, not stop needing answers. They'll just make more decisions without information, which is worse.
The fix is to make answers available without requiring a human in the loop.
This is what a good internal knowledge base is supposed to do. But most knowledge bases don't work. They're hard to search. They're out of date. They're organized in ways that make sense to the person who built them, but not to the person searching.
People try them once, don't find what they need, and go back to asking the answer person who knows the answer.
Instant Retrieval
The promise of AI is that it can actually deliver on what knowledge bases were supposed to be—a true employee self-service knowledge base. You ask a question in plain language. You get an answer drawn from your actual policies and documents. With a source you can verify. No more "let me get back to you." The answer is just there.
AI vs. Human Judgment
I'm not saying AI eliminates the need for human judgment.
Complex situations, edge cases, things that require interpretation—those still need a person. But the majority of questions aren't complex. They're routine.
"What's the policy on X?" "How do I do Y?" "Who handles Z?"
These are questions with answers that already exist somewhere. The problem is that "somewhere" is hard to find, so people ask humans instead.
Shifting routine questions from humans to AI doesn't just save time. It frees your experienced people to handle the questions that actually need them. The weird edge cases. The sensitive situations. The things where judgment and empathy matter.
That's a better use of expensive, scarce human expertise than answering "how many vacation days do I have left" for the third time today. This is where custom virtual experts shine—handling the routine so humans can focus on the complex.
The Aggregate Impact
The cost of "let me get back to you" is death by a thousand cuts.
No single question seems like a big deal. But in aggregate, across your organization, over time—it's thousands of hours, significant money, inconsistent answers, delayed decisions, and your best people stuck doing work that doesn't require their expertise.
You're already paying this cost. You're just not seeing it because it's distributed and invisible. Once you start looking for it, you'll see it everywhere.
JoySuite turns "let me get back to you" into "here's your answer." Employees ask questions, Joy answers from your policies and documents with citations they can verify. Your HR team gets its time back. Your employees get answers in seconds. Everyone wins.
