Ask an agency owner when they last ran a proper campaign for their own agency and watch the pause. The website case studies stop at a client from two years ago. The blog has an April post and then nothing. Meanwhile three client accounts got a strategy refresh this month, because a client asking is a deadline and your own agency asking is just a thought you had on a Sunday.
The billable hour is structurally against you
Every hour spent on your own agency's marketing is an hour not billed to a client, on a P&L where billed hours are the entire point. New business work gets scheduled into whatever is left after delivery, and delivery always wins, because delivery is the thing a client is actively watching. So "grow the agency" sits permanently at the bottom of a list that starts fresh every Monday with client work at the top.
This is not a discipline problem. It is what the incentive is built to produce. An agency that is fully booked has no spare capacity to sell itself, and an agency with spare capacity has no cash to spend selling itself. There is no month where this feels easy, which is exactly why it needs to run on its own rather than wait for one.
Your own referral pipeline has one flaw your clients' don't
You tell clients not to rely on one channel. Your own new business, for most agencies, is almost entirely one channel: referrals from past and current clients, plus whatever inbound drifts in from a case study somebody happened to read. That is a fine channel and a bad only channel, because it runs entirely on relationships that are not yours to control. A champion changes jobs. A client goes quiet for six months and so does the referral that would have come from them. You did the work that earns the referral and then waited for someone else to decide to make it.
The agency has never had a client of its own
Turn the lens on your own agency the way you would turn it on a client's business. Who is the actual buyer? Not "growing companies," the way half of agency positioning reads, but something you could search for: Series B companies that just hired a first in-house marketer and will need an outside partner for the next eighteen months, or local businesses in three specific categories who are still running the ad account themselves and clearly losing money on it. You already know how to write that sentence. You have written it for a dozen clients. You have not written it for yourselves, because the client work is what pays this month's rent.
Run your own new business the way you run a client's
This is the actual fix, and it is not more hours, because there are none. It is running your own acquisition on the same engine you would otherwise be assembling for a client from five separate pieces: a data tool for the list, a sending tool, a CRM someone has to keep updated, a scheduling link, and a phone that increasingly goes to voicemail because nobody at the agency owns picking it up. Every one of those seams is invisible labor you would never let a client absorb, and you absorb it for yourselves constantly, because there is no account manager assigned to the account named after your own agency.
Rocketship is one engine instead of that pile: describe the buyer once, it finds them, writes to them from your own Gmail, reads and answers what comes back, books the meeting onto your calendar, and answers your phone. It also builds the page a prospect lands on, on your own domain, and takes payment through Stripe when a deposit is due. Point the same engine at your own agency's buyer that you would build for anyone else's, and running it costs almost nothing next to a single billable hour.
Searching, counting how many of that exact buyer exist, and previewing twenty-five of them by name are free, no card required. That matters here specifically, because it means you can build and sanity-check your own target list on a slow Friday afternoon without touching the budget, the same way you would insist a client do before committing to anything.
Run the numbers your own agency would ask a client to run
Launch is $24.99 a month with a hundred credits. Adding a company to the list costs one credit. A verified phone and email cost five, charged only when found. A hundred companies added and something to say to each of them is well inside a month's plan, and it is less than an hour of one senior person's billable rate. If three or four people are touching new business, Frontier at $79 adds a second worker and the scheduler, so a reply that turns into a meeting lands on the calendar without anyone manually chasing a time. Mission Control at $149 runs three workers and unlimited apps, which is closer to what running acquisition for your own agency alongside several client accounts actually needs.
Want the agency's own line answered the same way clients' lines are, instead of ringing into whoever happens to be near the phone? That is five dollars a month on top of the plan, not part of the base price, and worth adding once the agency's own outbound starts producing calls back.
The account you keep deprioritizing is the one you own
Nobody is going to fire the agency for neglecting its own growth. That is exactly the problem: there is no client on the other end of that account holding you accountable, no monthly report due, no deadline that moves the work up the list. The fix is not finding more hours. It is putting the agency's own acquisition on the same machine that runs everyone else's, so it keeps moving on weeks nobody remembers to think about it, which is most weeks, for every agency, including the ones with excellent client results to show for it.