An agency is the right call while the founder still runs most of the company. Once the business starts copying itself into new locations or markets, it needs a system that sets goals, tests hypotheses and owns the result.
A marketing agency fits a company at the 0 to 1 stage, when the founder makes most decisions, the team is small and the job is to find the first channels that bring customers. It stops being enough at the 1 to 10 stage, when the business model has to work in new locations, audiences and markets and stops copying itself one to one. From that point a company needs a growth system: a financial goal with its key indicators, a go-to-market strategy built on audience research, a three-month tactical plan, and a growth lab that tests hypotheses every week across channels. The clearest signs it is time to switch are a working financial model, known channels that bring the best customers, and a marketing team or agency whose iteration speed no longer matches the company's targets. An agency sells a service delivered to a script. A growth system is a company habit of setting hypotheses and changing how sales, marketing and operations work together.
Most companies hire a marketing agency at the right moment and keep it for too long. An agency is built for a clear brief: launch the ads, produce the content, bring leads. That is exactly what a young company needs. Trouble starts when the company’s question changes from “how do we get customers” to “how do we repeat what works in ten new places without losing money”, and the agency keeps answering the first question.
A growth system is the way a company sets a financial growth goal, breaks it into indicators with owners, tests hypotheses against those indicators every week, and scales only what proved itself. It covers sales, marketing and operations, not marketing alone.
I have built marketing and growth processes in FMCG, as marketing director of restaurant chains in Russia and the United States, including a franchise network that grew to 1,200 locations, then in a fintech company in Southeast Asia, a real estate agency and our growth agency, Pushers. Today we work with clinic chains across Russia and with startups looking for their first customers. The pattern below repeats across all of them.
At this stage the founder is the operating system. They make most of the decisions, the core team is a handful of people who understand each other without documents, and there are no written processes. Decisions follow the mood of the team more than a shared plan.
That is fine. The job is to find a product and the first channels that bring paying customers, and the founder holds the strategy in their head. An agency adds hands and channel experience. Metrics here are mostly performance metrics: cost per lead, customer acquisition cost, first sales.
The first serious problem of growth is that the business model does not transfer one to one. The same restaurant, clinic or store, online or offline, lives differently in every region, because the environment and the audience are different. No boxed solution repeats the first result exactly.
At this stage a company needs analysis more than campaigns: where the real growth points are, which parts of the business model hold in a new place and which do not, which processes have to be described and measured so that someone other than the founder can run them. A management company starts to form, with the first middle managers and the first signs of a top team running the network.
The company is also exposed. Bigger competitors are nearby, and the team tends to spread itself over new directions without knowing which one deserves to be scaled. The toolkit changes too. Performance metrics stay, and new ones join them: brand awareness, work with the existing customer base, managing customer flow, and visibility at every point of contact, from offline stores to marketplaces.
An agency hired to run ads is not set up to answer any of this.
From here a growth system is no longer optional. Its existence decides how fast the company grows. When the model is profitable, founders and investors need answers to concrete questions: in which locations to open, at what cost of capital, on which markets and audiences to work and which to leave alone.
The first input is the size of the goal. Growing 10 times and growing 100 times are different plans. They need different teams, different effort and a different approach.
The second input is speed. If the target is a thousand locations over a hundred years, opening about ten a year and keeping them from closing will get there. Active expansion toward the same number in five years needs extra financing, risk management and a system that runs growth as a process.
| Part | What it produces | Rhythm |
|---|---|---|
| Strategic goal | One financial target, its key indicators and the parts it breaks into, often as a KPI tree | Yearly, reviewed quarterly |
| Go-to-market strategy | Audience research, market demand, competitors and their channels, key success factors | Built once, updated with each market |
| Tactical plan | Budgets, campaigns, audiences and the hypotheses to test in the period | Every three months |
| Growth lab | Weekly tests of hypotheses across channels and niches, scaling of the winners | Weekly, on a HADI loop |
The go-to-market strategy rests on research, not opinions. That means qualitative interviews and customer personas, desk research on demand and trends, and a clear view of what competitors do well and which advertising channels they use.
We run the tactical plan as three-month strategy sessions with the client team, including operations. In healthcare this matters more than anywhere: changes take a long time to stick, and hiring and onboarding one specialist can take months, so a plan that ignores operations fails on the first hire.
The growth lab is where hypotheses become numbers. It checks them in different channels and audience segments and passes the ones that work into scaling. Our Growth Lab practice is built on this.
A clinic network sets a goal for the year and splits it into new patients, repeat visits and average check. The three-month plan picks two hypotheses per indicator. The lab tests them weekly; a call-script change that lifts the call-to-booking rate goes to every branch, a channel that misses its cost target is cut on the agreed date.
Behind the parts sits a habit. A growth system teaches the company to grow on purpose: to set uncomfortable hypotheses, to change things itself instead of riding the same rails for ten years and waiting for the market to shift, and to make sales, marketing and the team that delivers the product work on the same numbers.
The same method works for a solo founder or an independent expert. A self-managed growth system is small: one goal, a fixed budget for experiments, a short list of hypotheses and a weekly look at what moved the number. Putting yourself inside that frame shows quickly which actions bring the best results right now and which only fill the week.
If several of these apply, switching to another agency will reproduce the same limits. Build a growth system instead, inside the company or with experts who can build the competence quickly and hand it over.
Start with the goal and the numbers: write down the financial target, break it into indicators with named owners, and check which of them your current team can actually move. Our marketing operations audit checklist covers that first pass. If you want help building the system itself, tell us about your company and we will start with a two-week audit.
No. An in-house department can still work like an agency: take briefs, run campaigns, report activity. A growth system starts from a financial goal, connects sales, marketing and operations around it, and runs hypotheses with a decision rule. It can be built with an internal team, outside experts, or both.
Yes, in a reduced form. A solo founder can set one goal, a fixed experiment budget and a weekly review of what moved the number. That minimum version already shows which actions bring results and which only fill the calendar.
The first working version usually follows the first three-month planning cycle: goal and indicators, research, a tactical plan and the first round of tests. In industries with long change cycles, such as healthcare, where hiring and onboarding one specialist can take months, the system needs more than one cycle to settle.
Not necessarily. Agencies stay useful for production and channel execution. What changes is who owns the goal, the hypotheses and the decision to scale or stop a channel. That ownership moves into the growth system.
Most agencies own the click and nobody owns the front desk. How we structure clinic work in three horizons, the research behind each one, and one year of results.
Most paid channel tests do not end with a decision. They end when the budget runs out, the attribution model changes, or everyone quietly stops looking at the dashboard.
The two models solve different problems: how to tell which one your company needs, and what each costs in time and control.
