Every founder I talk to knows their hourly rate. Almost none of them know what their backend costs them per hour. That gap is where a surprising amount of money lives.
I have watched this for a while now. The number is almost never small. It usually lands between $150K and $350K a year, and it does not show up on any invoice.
The Admin Tax, Priced Out
Start with you. Say you bill $250 an hour, or your effective rate works out to that once you divide revenue by actual working hours. Now count the admin. Inbox triage. The invoice run. Scheduling back and forth. The proposal you rewrote from scratch because you never saved a template. The client update you sent at 11pm because it was the only quiet hour you had.
Fifteen hours a week is common. Some of the founders I work with are closer to twenty-five.
At fifteen hours and $250 an hour, that is $3,750 a week. $195,000 a year. Just you. Just the backend.
Now add the rest. A part-time admin at $28 an hour for twenty hours a week is roughly $29,000. A bookkeeper. A VA. A contractor you pay to fix the CRM nobody set up correctly. Then add the leads that went cold because nobody followed up within a week, which you will never see on a spreadsheet but you feel every January.
That is how the range gets to $150K to $350K. It is not one big expense. It is a tax paid in small increments, all year, mostly invisibly.
One thing I want to be clear about: this is not an argument for firing your admin. The person is usually the best value in the building. The tax is on the system, not on them.
Why the Real Number Is Higher Than the Math
The hours are the easy part to count. The expensive part is the lag.
A proposal that goes out four days late wins less often than one that goes out the same day. Not because the client is impatient, but because you were the third option they were considering and the first one to respond gets the momentum. Every founder I know has lost work to this and blamed it on price.
A lead that sits in an inbox for three days is usually gone. Not always. Often enough that the pattern is real.
Then there is the switching cost. If you do revenue work for forty minutes and stop to answer an invoice question, you do not lose forty minutes. You lose the forty plus whatever it takes to get back into the thinking. The deep work that makes your business good happens in uninterrupted blocks, and manual admin is a machine for interrupting blocks.
That is the part that does not fit in a founder bottleneck cost calculator. The calculator gives you the dollars. The quality of attention is what quietly drives the number up over time.
Run a Founder Time Audit Before You Change Anything
You cannot automate what you have not seen. So before you buy anything, do a five-day founder time audit. It takes about two hours total and it will annoy you.
Track in fifteen-minute blocks for five business days. Toggl Track works. A Google Sheet with three columns works just as well. The tool does not matter. The honesty does.
Tag every block with one of four labels: revenue work, delivery, admin, or nothing. "Nothing" is real and you should count it.
Then go back through and mark each admin block with two questions. Could someone else have done this? Could a system have done this? The first question is delegation. The second is automation. Most founders find that 60 to 70 percent of their admin blocks answer yes to the second one.
One more cheap trick. Search your sent email for "just checking in" and "circling back." Count them. That number is a map of where your process has holes.
The One Workflow to Automate First
Founders ask me what to automate first in their service business constantly, and I give the same answer every time. Not invoicing. Not content. The chain from new inquiry to booked call. It sits at the front of your revenue pipe, it runs constantly, and it fails quietly.
Here is what I would build, in order:
One form. Tally or Fillout, embedded on your site. Not a contact box that says "send us a message." Four or five real questions that qualify the person and give you something to prep with.
One place where the answers land. Airtable or Notion. Every inquiry in the same table with a status column. This single step kills most of the "I think someone emailed about this" chaos.
One automated reply, sent within a minute, with a booking link. Make or Zapier handles the trigger. Calendly or SavvyCal handles the booking. Write the reply in your own voice, warm and short, not something that reads like a car rental confirmation.
One nudge at 48 hours if they have not booked. Most of your recoverable revenue lives in this step and almost nobody builds it.
One prep doc, generated from the form answers, that lands in the calendar event. Now you walk into the call already knowing what they need.
That whole chain is a weekend of work. It is also measurable, which matters, because it lets you answer whether AI automation is worth it for consultants and studios with your own numbers instead of someone else's case study.
I start here and not with invoicing because this workflow touches revenue directly. When it gets faster, you see the result inside a month.
Why I Would Wait on the Next Tool
Most founders I talk to are paying for six to ten subscriptions and actively using two. Adding another tool to a process nobody has mapped just makes the mess move faster.
The order that works is boring. Map the process. Simplify it. Automate the simple version. Then add AI where judgment is genuinely required. Skip to the AI step and you get an agent that is confidently wrong inside a system nobody designed.
Most AI automation for service businesses fails at that same point, and it is never the model's fault.
I have been teaching this since 1998 and the pattern has not changed. The tools get better every year. The order does not. If you want to see how we run this inside a real business, the workshop walks through the build live, and this breakdown is where I lay out what different levels of support actually look like.
The admin tax is real and it compounds. The good news is that the first fix is usually small, cheap, and visible within a month.
If this resonates and you want to build this kind of infrastructure in your own business, the Mastermind is where we do the work live. You can learn more at mastermindshq.business.