Build an agency that runs without you

Karim Marucchi is direct that the founder is usually the biggest obstacle to growth, because the instinct to hire people who think like you simply reproduces your own blind spots. The hustle that got you here, improvising, personally solving every client problem, holding the entire context in your head, becomes a liability at scale, turning you into the central node every decision flows through.

Most owners do not notice the transition happening; they keep operating as they did at five people while the team grows around them, and the bottleneck compounds. The fix is to hire for your weaknesses and then genuinely get out of the way, which is far harder in practice than in theory. Accepting that you are the constraint is the precondition for everything else in this playbook.

Marucchi offers a blunt diagnostic: if you cannot step away from the agency for two weeks without it breaking, you have not yet built a functioning management layer. It is a test you can run honestly against your own calendar, and the answer tells you exactly how much of the business still lives in your head rather than in your systems and people.

The goal is not heroism or being indispensable, it is a business that keeps delivering, invoicing, and communicating without your daily presence. Everything that follows is about passing this test on purpose rather than discovering the answer in an emergency. Treat the two-week test as the outcome the rest of these steps are engineering toward.

Steph Hudson's talk starts from the fact that having a team is not the same as having a plan: if people do not know what to do in your absence, staff alone does not protect the business. Move your two-factor authentication codes out of your phone and into a shared password manager immediately, because codes locked to your device are a critical single point of failure that can lock everyone out.

Instruct any stand-in to review the last four to six weeks of your sent box, not your inbox, because sent mail reveals your active commitments and open threads. Borrowing the Red Cross model of 72-hour emergency preparedness, the aim is not a full disaster-recovery plan but simply keeping things running through the first few days. Busproofing is the fastest, cheapest resilience you can build, and most agencies have not done it.

Before anything else, Hudson says, identify two specific people: a PAL, a Personal Alert Lookout from your personal life who will notice if something goes wrong, and a DEX, a Designated Executor who will act on your behalf in a business context. The two should be introduced to each other, so the handoff in a crisis is clear and trusted rather than chaotic and improvised.

This is a small, concrete step most owners never take, and it is the difference between an orderly continuation and a scramble when it matters most. Continuity depends on named humans with defined roles, not on good intentions or the assumption that someone will step up. Make the introductions now, while nothing is wrong.

Thomas Amos warns that SOPs built in a vacuum produce documentation nobody uses, so you have to map the end-to-end customer journey first and let it tell you which SOPs are actually needed. This is what gets the critical workflows out of your head and onto paper in a form the team can follow without you narrating it.

Done in the right order, the journey defines the process, and the process defines what can be delegated, standardised, or eventually automated. Documentation is how knowledge stops being founder-shaped and becomes team-shaped, which is the whole project of running without you. Start with the journey your clients actually travel, not the org chart you wish you had.

Natasha Golinsky insists you know exactly what you want before you post, defining required technical skills, non-technical skills, and personal non-negotiables, because searching for a specific person is entirely different from posting and hoping. Use platforms with verified reviews, Upwork and similar, where earnings, job-success scores, and client reviews cannot be fabricated, unlike a resume which is guesswork by comparison.

Make the job posting itself a filter, be specific about rate, skills, time zone, and engagement type, because a vague posting repels good candidates and attracts spray-and-pray applicants. Bury a magic question in the posting to catch those who did not read, and use names as a signal, a candidate who opens with Dear Hiring Manager after you signed your name is not paying attention. Good hiring is the raw material of a team that can run without you.

Golinsky's sharpest screening tool is the DM exchange, which she treats as a live communication audit that reveals more than any credential. A candidate who replies ready to start, no questions is a red flag, because good contractors have thoughtful questions about the work, the context, and the expectations.

The back-and-forth before hiring is a preview of every future interaction, so pay attention to responsiveness, clarity, and curiosity, not just skill claims. Attention to detail and genuine engagement in that first exchange predict whether someone can be trusted to operate without close supervision. Hire the person who asks good questions now, because that is the person who will make good decisions when you are not there to make them.

Marucchi argues onboarding must be deliberate and, yes, expensive: new hires should have a dedicated buddy, near-constant contact in the first 90 days including video presence when remote, and they should not be billed to client projects during that period. Cutting corners here is what produces hires who never become self-sufficient and quietly keep routing decisions back to you, defeating the whole purpose.

The upfront investment is what buys you a teammate who genuinely takes work off your plate rather than adding oversight to it. You cannot delegate to people you never properly brought in, and rushed onboarding is a false economy that costs far more in prolonged dependence. Treat the first 90 days as an investment in your own eventual absence.

Marucchi's core structural fix is three distinct leadership roles: an account lead who owns the client relationship and long-term revenue, a delivery lead who owns the current project's scope of work, and an operations or practice lead who owns internal team health and process. These roles should sit in productive tension with each other rather than in a reporting hierarchy, so no single person, especially you, holds the whole thing up alone.

This is precisely the management layer the two-week test is checking for, the spine that lets the business stand without the founder. Each role has a clear owner, a clear remit, and a clear counterweight, which is what keeps decisions moving when you are away. With it in place, the agency has a spine that is not your spine.

An agency only runs without you if your best people stay, so Marucchi emphasises real feedback loops, defined career paths, and genuine team development, not just hiring. People who see a future and receive honest, regular feedback are the ones who take ownership rather than waiting for direction.

Investing in their growth is what turns a team of task-doers into a group of decision-makers you can actually step back from. Retention of your key people is the quiet prerequisite for delegation, because every departure drags the founder back into the gap. Build the paths and the feedback that make staying and growing the obvious choice.

Marucchi calls headcount a vanity metric and points to revenue per head as the real measure of health, targeting three times average cost per head, with two times being the point where an agency starts building meaningful reserves. Keep no single client above 12% of revenue, because concentration turns one departure into an existential event that pulls the founder straight back into firefighting, and when a client exceeds it, partner with another agency rather than hiring frantically.

A financially sound, diversified agency is one you can actually leave, because no single loss threatens its survival. The numbers are what give you the freedom to step away without watching the dashboard from your holiday. Manage the business by these metrics, not by how busy everyone looks.

Marucchi models the culture that sustains a business the founder can leave: Crowd Favorite has operated as a distributed company since 2014 and reserves Fridays entirely for internal work, learning, and exploration, no client meetings and no billable work. Intentional culture-building is not a perk, it is what keeps a distributed team aligned without the founder personally holding it together day to day.

He also frames AI correctly in this context, a tool, not a strategy, that accelerates production but cannot generate original strategy, so the agencies that reinvest AI savings into upfront discovery and QA will outperform those that just do more faster. A deliberate culture and a clear-eyed view of AI are what let the systems and people you have built keep running long after you have stepped out of the room. Build the culture on purpose, because it is the last thing holding everything together when you are not there.