What founder-led support really costs (a worked example)

Founder-led support costs more than the minutes you spend typing replies: every task you do by hand for a customer pulls in a call, an account lookup and the time it takes to get back to what you were building. In the illustrative example below, 40 conversations a week add up to about 31 hours a month of founder time, and three quarters of it goes to the 15 conversations that were tasks, not questions. The table is built so you can replace every number with yours in about fifteen minutes.

The numbers in this post are made up to show the method. They are not data from any company, and they are not aside’s data. Their only job is to show the arithmetic so you can run it on your own week.

Key numbers (results of the illustrative example below):

• 31 hours a month of founder time, from 40 conversations a week.
• €2,325 a month, at €75 an hour.
• 77% of that time goes to tasks, which are 15 of the 40 conversations.

What do you need to work out the cost of founder-led support?

You need four inputs, and each one comes from somewhere you can look.

1. How many conversations you handle a week, split into questions and tasks. A question is “how does X work?” or “do you support Y?”. A task is “change X”, “set up Y”, “add Z”. If you have never split them, sort your last 30 conversations into questions and tasks first; it takes about twenty minutes.
2. How long each one takes by hand. Not the reply, the whole thing: finding the account, the call or screen-share, the follow-up email, the Loom you record because the call ran out of time. Time three or four of each kind this week with a stopwatch rather than guessing.
3. How long it takes you to get back to work afterwards. This is the line most founders leave out, and the one that hurts most.
4. What one hour of your time is worth. Not your salary. The useful number is what that hour would produce if it went to building or selling, or what you would have to pay someone else to take the work off you.

Why the refocus line is real

The third input has research behind it. In a field study of information workers, Gloria Mark, Victor González and Justin Harris found that 57% of working spheres were interrupted, and that when interrupted work was resumed the same day, it took an average of 25 minutes 26 seconds to get back to it, with an average of 2.26 other activities in between.

That doesn’t mean you lose 25 minutes after every call. You were doing other things in that gap, and some of them were useful. So the worked example below counts a conservative 10 minutes of refocus per task that pulled you into a call or a screen-share, and none for questions answered in a batch. Adjust it to what you see in your own day.

The worked example: say you have 40 conversations a week

Say you run a B2B product with three people and a few hundred paying customers. You did the sort, and a normal week looks like this: 40 conversations, of which 25 are questions and 15 are tasks. Questions are answered by email in a batch, about 4 minutes each. Tasks are things like changing a plan, inviting a teammate with the right role or setting up an integration; about half of them turn into a call, and on average each one takes 12 minutes of your time. Say an hour of your time is worth €75.

Line	Formula in words	Example	Your number
A. Questions per week	Count from your sorted inbox	25
B. Minutes per question	Average of the ones you timed	4
C. Tasks per week	Count from your sorted inbox	15
D. Minutes per task	Average of the ones you timed, call included	12
E. Refocus minutes per task	What it takes to get back to work after one	10
F. Hours on questions a month	A × B ÷ 60 × 4.33	7.2 h
G. Hours on tasks a month	C × (D + E) ÷ 60 × 4.33	23.8 h
H. Total hours a month	F + G	31 h
I. Value of one hour	What the hour would produce, or what you would pay to delegate it	€75
J. Monthly cost	H × I	€2,325
K. Share that is tasks	G ÷ H	77%

The 4.33 is the average number of weeks in a month (52 ÷ 12). Copy the table into a spreadsheet, fill the last column and you have your own version.

Where does most of the cost go?

Most of it goes to tasks, so look at line K before line J. In the example, tasks are 15 of 40 conversations, less than 40% of the volume, and they take 77% of the time. That shape is what you should look for in your own numbers, because it tells you where the cost really is.

The reason is simple. A question can wait for a batch. A task usually can’t: the customer is stuck halfway through something and wants it done now, so it interrupts you, it often needs the customer’s account open, and it is the kind of thing that turns into “can we jump on a quick call?”. The minutes per task are higher and the refocus applies to almost every one of them.

This changes what “reducing support” means. If your answer is more documentation, you are working on line F, which in the example is the smaller one. Customers already try the docs first: Harvard Business Review reports that 81% of customers try to solve things themselves before contacting a person. The tasks in line C are, for the most part, what is left after they tried.

What this table leaves out on purpose

Three costs don’t fit in a formula, so they stay out. That keeps the number honest, and it also means the real cost is higher than whatever you get.

• The customers who never wrote in. Someone got stuck setting up the integration, didn’t ask, and quietly stopped using the feature. Your inbox can’t count them.
• Which hours the support takes. Calls land when customers are free, which is usually the middle of your working day, not the end of it. An hour lost at 11:00 isn’t worth the same as an hour at 19:00.
• What didn’t ship. Thirty hours a month is most of a working week. The feature that slipped a sprint because of it doesn’t show up as a support cost anywhere.

If you want to be strict, mention these next to the number when you share it, without adding them to it.

How do you decide whether a fix is worth paying for?

Compare the monthly cost with what the fix costs. Two quick formulas do most of the work.

Break-even hours. Divide the monthly price of any tool by the value of your hour. In the example, a tool at €99 a month has to give you back 99 ÷ 75 = 1.3 hours a month to pay for itself. At 22 minutes per task (line D + line E), that’s about four tasks a month that you no longer do by hand.

Which line does it reduce? This matters more than the price. Different fixes work on different lines:

Fix	Line it reduces	Line it leaves
Better documentation	F (questions)	G (tasks): the customer still has to do it
A chatbot over your docs	F, faster and at any hour	G, for the same reason
A product tour	Part of G, for paths you predicted in advance	Tasks outside the tour
Hiring part-time support	F and G	Your own budget; compare with what that person costs in your market
An in-app assistant that runs actions	F and G	Tasks you haven’t turned into actions yet

The last row is what aside does, so read it with that in mind. It answers questions from your documentation and says when it doesn’t know. For tasks, your page registers actions, which are functions you already have, like change plan or invite teammate. When a customer asks, it picks the action and its arguments, your own code opens the screen and fills the form where the customer can see it, and the customer presses the button that saves. When it can’t solve something, it hands the conversation to your team, in aside’s inbox, Slack or a webhook. Plans start at €49 a month for 200 conversations, down from €249 before, which was too much for solo founders and small startups, and answering from your docs only needs one script tag. Actions need someone who can write a function.

Whatever you pick, use the break-even formula with your own line G and line I. If the tasks line in your table is small, a chatbot over your docs or better docs may be all you need.

Redo it with your numbers this week

1. Sort the last 30 conversations into questions and tasks (here is how), and scale the counts to a week.
2. Time three questions and three tasks, start to finish, call included.
3. Pick a refocus number you believe. Ten minutes per task is a conservative starting point.
4. Write down the value of one hour, and say why you chose it.
5. Fill the table and look at line K first.

If tasks come out as the bigger share, the next step is reducing them, not answering them faster. How to cut support calls when you’re a 3-person SaaS walks through that, starting with the three tasks you repeat most. And if you want the story behind the numbers, the same question, every week is about why customers still write in when the answer is already in your docs.