Every small decision waits for you
TL;DR
Your team is not slow. It is waiting on you.
Every question only you can answer is an answer with no home.
Researchers at UC Irvine measured it: 23 minutes 15 seconds to fully refocus after one interruption.
You do not need to delegate harder. You need to move the answers out of your head.
Count the questions for seven days. Then give each repeated answer a home your team can reach without pinging you.
Sunday night. You open WhatsApp and there are fourteen unread messages from your own team. Can we give this discount? What do we tell the client about delivery? What should the new hire do first? These are the same questions as last month. You answer them in bed, one by one, and the company moves again.
That is the job now. Not building the business. Being its help desk.
The team is not lazy and they are not stupid. They ask because the answers live in exactly one place: your head. When you are busy, or asleep, or on a flight, the work waits. Your company has a single point of failure, and it is you.
Why does every small decision land on the founder?
Because you are the only one who holds both the information and the authority. You know the client history. You know the margin. You know which supplier can be pushed. So every question routes to the one brain that has the full picture.
There is a line a coach used about his client: your org chart says Head of Sales, but your decision structure says check with the founder. The titles are fiction. The real chart is a star, and you are the center.
People do not ask because they cannot think. They ask because guessing wrong is expensive for them. If a person guesses the discount wrong twice, they stop guessing out loud. So they bring you the question. It is the rational move for them, and it is a disaster for you.
And you trained this. Early on, you were the fastest brain in the room, and the company was you. The habit that built the business is now the habit capping it. If every operational decision still needs you years in, the business has not scaled. It has just grown around a very busy person.
What does answering cost you?
More than the five minutes per question. Researchers at UC Irvine spent years logging interruptions across office workers and found it takes an average of 23 minutes and 15 seconds to fully return to a task after being pulled away. Not to glance at the screen. To reach the same depth of thought you had before. Their CHI 2008 paper, The Cost of Interrupted Work, is the primary source.
Then there is the residue. A University of Minnesota study gave it a name: attention residue. When you switch from an unfinished task to a new one, part of your mind stays behind, still processing the old one. You answer a two minute question and your brain keeps working on the last thing, not the current one, long after you have replied.
The sneaky part is that the day still looks productive. Researchers found that interrupted people finish their work anyway. They compensate by running hotter: more stress, more frustration, more time pressure. The bill does not show up as missed deadlines. It shows up as you feeling wrung out at 11 pm after a day that, on paper, went fine.
Zoom out and the pattern is everywhere. A recent UENI study of micro business owners found 85 percent struggle with time management, with distractions and poor organisation among the top obstacles. You are not uniquely bad at focus. You are running a structure that makes focus impossible.
Why does answering yourself feel faster?
Because it is. Once.
You read the question, you know the answer, you type it in forty seconds. Writing the rule down for the team would take ten minutes. So you answer. This is the trap: fast once, forever. Every forty second answer is a promise that you will answer the same question again next week.
Delegating feels slow because the first time costs more. The tenth question is where the maths flips. One ten minute note beats twenty interruptions at 23 minutes of refocus each, and that is before the attention residue.
I made a version of this mistake in Your AI tools sit next to the work, not in it. A fix that lives beside the work gets skipped on a busy day. The same is true of your answers. A rule in your head is a rule beside the work. It only works when you are there to say it.
What is the difference between giving permission and giving answers?
This is the part most founders get wrong. You keep telling the team, "Just decide, you do not need my permission." They nod. They keep asking.
Permission was never the problem. They are not asking whether they are allowed. They are asking what you would have said. Handing someone the right to decide without handing them the rule you would have applied just moves the fear. They can decide now, but they cannot defend the decision, so they still come to you.
The team needs your answers, not your authority. They need the discount rule, the delivery-date script, the supplier priority. The org chart fiction dissolves the moment someone can say, "I did what the founder would have done," and have that be checkable.
This is also why the knowledge-exit problem from If your best person leaves, what leaves with them? starts with you. You are the best person. Everything you know and have not written down is already a hostage situation. The team just has not called it that yet.
How do you find your questions?
Count them for seven days. That is the whole audit.
Star every message where someone asks you something. At the end of the week, sort them into three piles. Pile one: only you could have answered. Big money, hiring and firing, strategy bets. That pile is your job. Pile two: you answered, but someone else could have if they knew the rule. Pile three: this question has been asked before, and the answer existed nowhere.
Piles two and three are the system you never built. Most founders find that pile one is tiny. A coach who ran this audit with a founder found 26 decisions in two weeks, and 19 of them never needed the founder. The people just were not sure where their line was.
Do not fix anything in week one. Just count. You cannot hand over answers you have not listed.
What do you hand over first?
The repeated ones. If three people asked the same thing this month, that is your first rule to write down. Frequency is the priority list.
Then build a ladder with one rung per risk level. Something like this works for a small firm:
Decide and inform. Small spends, say under ten thousand rupees. Decide it, drop a line in the group, move on. Nobody asks first.
Decide and note it. Reversible choices within agreed bounds. They act, they log it where you can see it, you can overrule within a day.
Escalate with options. Client risk, money that matters, anything hard to reverse. They bring what happened, why it matters, the options, and their recommendation. Never just "what should we do."
Yours alone. Big money, people decisions, strategic bets. These stay on your desk, and saying so out loud is a relief to everyone.
One ladder, written down, in the place the team already works. Inc has a useful framing of this as decision rights: the fix is not better delegation, it is a system that says who can decide what, with what information, so nobody has to ask you at 9 pm.
Where should the answers live?
Not in a Google Doc graveyard. Nobody opens the playbook doc on a busy Tuesday. The answers have to live where the questions are.
The delivery-date answer belongs next to the order record. The discount rule belongs next to the quote. The supplier priority belongs next to the purchase list. If your team works in WhatsApp and Excel, the answer sitting in a folder they never open is no answer at all. That is the same lesson from You pay for software. Your team uses WhatsApp.: a system beside the work gets abandoned.
So the real test for wherever you put this is one line: can the team reach the answer without pinging you? If yes, you have started building the company. If no, you have just moved the filing cabinet. The same was true for chasing money: payment follow-up only stopped living in the founder's head when it moved into a system the team could see.
This is the thinking behind BaseEngine. Not features. Not seats. One system your team actually opens, where the answers sit next to the work they answer. It is the same placement logic we use in our AI automation work: inside the work, not beside it. The guarantee that matters is not on the software. It is on the day your team stops needing to ask.
That is the conversation I want this piece to start. This week, count the questions. Every repeat question is a system you have not built yet. And every system you build is one more evening you get back.
Want to work through this for your business? Start with the BaseEngine overview and ask one question: how many of the questions your team asked you this week could have been answered by a system instead of your phone.
Frequently Asked Questions
How do I know if I am the bottleneck?
Check three signs. Nothing moves when you are on leave. The same questions come back every month. Your calendar belongs to everyone but you. If all three are true, you are not leading the company. You are its help desk, and the fix is the answers, not the people.
Should I just tell my team to stop asking me questions?
No. That keeps the fear and removes the only channel. People will guess in silence instead, which is worse. The move is not fewer questions. It is fewer repeat questions: answer once, write the rule down, put it where the team works, and point to it next time.
What counts as a reversible decision?
Anything you can undo within a day or two without losing money or trust. A discount under your threshold. Reordering stock. Rescheduling a non-critical delivery. If the worst case is a small apology and a correction, it is reversible. Hand those over first, with a note-it rule so you can overrule.
What is a decision rights ladder?
A short written list that says who can decide what without asking. Decide and inform for small spends. Decide and log for reversible choices. Escalate with options for anything risky. Founder-only for big money, people, and strategy. The point is that the line is written down, so nobody has to guess where it is.
Where should the answers live so people actually find them?
Next to the work they answer, in the system your team already opens daily. The discount rule sits next to quotes. The delivery script sits next to orders. A playbook doc nobody opens is not a home. If the team still has to ping you to reach the answer, the answer has no home yet.
How long does this take to set up?
The audit takes one week of starring messages. The ladder takes an afternoon. The first ten answers take a week of writing rules when questions arrive. After that it compounds: every repeated question you answer in writing is one you never answer again. Start with the three most repeated questions and go from there.
By Aravind Bhati, Founder, BaseWorks