Disclosure: This post contains affiliate links. If you sign up through them, I may earn a commission at no extra cost to you.
Getting your first client is a confidence problem. Getting your tenth is a systems problem. Most new agency owners solve neither, because they spend month one building a website nobody visits and month two waiting for it to work.
This is the unglamorous version: how ten paying clients actually arrive, in roughly the order they arrive in, and what you need in place so numbers four through ten don’t collapse under the weight of numbers one through three.
Clients 1–3: they already know you
Your first three clients are almost never strangers. They come from your existing network — former employers, people you worked alongside, the friend who runs a dental practice, the founder you met at a meetup two years ago.
This feels like cheating. It isn’t. Warm leads convert because trust is already priced in; you’re only negotiating scope.
What to do this week:
- Write down 40 names. Anyone who knows your name and has a business, or knows someone who does. LinkedIn connections count. Old Slack workspaces count.
- Message them individually. Not an announcement post. A specific, short note: what you’re doing now, who it’s for, and a direct question — “do you know anyone running paid ads who’s frustrated with their results?”
- Ask for referrals, not favours. People say no to “hire me” and yes to “who should I talk to.” The referral ask has a far higher hit rate and costs your relationship nothing.
- Charge from client one. Free work anchors your value at zero and free clients are the least respectful of your time. A small paid pilot beats a large free one every time.
Expect roughly one client per 30–40 warm messages. That’s a normal rate, not a failure.
Clients 4–7: pick one niche and go narrow
Around client four, referrals thin out and you have to generate demand instead of harvesting it. This is where most agencies stall — because “we do marketing for small businesses” is unsellable. Nobody wakes up looking for that.
Narrow until the pitch writes itself. Not “email marketing” but “abandoned-cart email flows for Shopify skincare brands.” Not “SEO” but “local SEO for multi-location dental practices.” A narrow niche gives you three compounding advantages: your outreach message can be specific enough to be believed, your case studies transfer directly from one client to the next, and referrals travel inside the industry without you.
Pick the niche where you already have one result — even a mediocre one. A single “we took X from A to B” is worth more than a portfolio of tasteful mock-ups.
Then run outreach with actual volume:
- 50–100 targeted prospects per week, researched by hand or with a list tool, not scraped indiscriminately.
- Lead with an observation, not a service. “Your checkout page loads in 6.2 seconds on mobile” opens more conversations than “we offer CRO services.”
- Follow up four to six times. The majority of replies land after the first message. Most people quit at one and conclude outreach doesn’t work.
- Track everything. Sent, opened, replied, booked. Without numbers you can’t tell a bad list from a bad message, and you’ll change the wrong one.
This is also the point where a spreadsheet stops being enough. When you’re juggling 200 open conversations across email, LinkedIn, and a phone, deals fall through the gaps silently — you never find out about the one you forgot to follow up with.
A CRM with built-in pipelines, automated follow-up sequences, and a booking calendar solves this. GoHighLevel is the common pick for agencies specifically because it bundles the CRM, funnels, calendar, SMS, and email automation into one $97/month Starter plan rather than four separate subscriptions — and because at the $297/month Unlimited tier you can white-label it and give each client their own login, which becomes a selling point rather than a cost. Note that email, SMS, and phone usage bill separately on top of every plan, so budget for that.
If you’re not ready for an all-in-one platform, a dedicated email tool like Moosend handles nurture sequences cheaply while you keep the pipeline in a spreadsheet a while longer. Just be honest about when the spreadsheet starts costing you deals.
Clients 8–10: stop selling projects
The trap at this stage is that you’re busy but not stable. Ten one-off projects means ten sales cycles a quarter and a revenue chart that looks like a heart monitor.
Convert to retainers as early as you can defend it. Retainers do three things: they smooth cash flow, they let you invest in a client’s results over months instead of weeks, and they cut the amount of selling you have to do to stand still.
A workable ladder:
- Paid audit or pilot — £500–£1,500, two to three weeks. Low risk for them, and it gets you inside their data.
- Fixed-scope project — the fix your audit found. Proves you can execute.
- Monthly retainer — ongoing management, reporting, iteration. Price on the value of the outcome, not on your hours.
Roughly half of well-run pilots convert into retainers if the audit surfaces something concrete and you propose the retainer before the project ends, not after.
The lean stack you actually need
Resist the urge to buy tools before you have clients. This is the whole list for the first ten:
| Job | What you need | Rough cost | |---|---|---| | Pipeline + follow-up | A CRM with automation | $0–97/mo | | Email sequences | Dedicated ESP or bundled | $0–30/mo | | Scheduling | Booking calendar (often bundled) | $0–15/mo | | Proposals + contracts | E-sign tool or a good template | $0–20/mo | | Reporting | Platform-native dashboards | $0 |
Everything else — analytics suites, project management, brand design — can wait until a client’s money is paying for it.
Get started with an agency CRM once outreach volume is high enough that you’re forgetting follow-ups. Before that, it’s overhead.
What actually stops people
Almost nobody fails here for lack of skill. They fail because they:
- Stay too broad, so no message ever lands hard enough to get a reply.
- Send too little, treating 20 emails as a real test of a channel.
- Quit after one follow-up, missing most of the replies that were coming.
- Under-price and over-deliver, ending up with ten clients they resent and no margin to hire.
- Build instead of sell — a fourth website revision is procrastination wearing a productive hat.
Ten clients is not a marketing problem to be solved cleverly. It’s forty warm conversations, then a few hundred cold ones aimed at a niche narrow enough to be memorable, tracked well enough that nothing gets dropped. That’s it. The agencies that get there are just the ones that kept sending after the first quiet week.