Growth systems
Growth loops vs campaigns: building growth that compounds
A campaign spends to create activity. A loop is built so the output of one user's action helps produce the next user. Both are useful — but only one of them keeps working after the budget stops.
Campaign thinking is the default because campaigns are legible. There is a start date, a budget, a creative set and a report. System thinking is harder to sell internally, because the value arrives later and belongs to no single quarter. It is also where durable growth comes from.
Linear spend against a closed cycle
In a linear model, attention is rented. Budget produces reach, reach produces conversions, and when the budget stops, activity returns to its baseline. Nothing about the third campaign is cheaper than the first, apart from accumulated creative knowledge.
In a loop, each completed cycle leaves something behind that helps the next one start: a page that ranks, a profile that attracts, a review that reassures, a colleague who was told.
Linear campaign vs circular loop
The four loops worth building
Most businesses can build at least one of four loops without inventing anything exotic.
- Content and search loop — work produces genuine expertise, expertise becomes indexed answers, answers attract people with the same problem, and their questions determine the next piece of work.
- Referral loop — a delivered outcome earns a recommendation, the recommendation brings a similar customer, and similar customers are cheaper to serve and more likely to refer again.
- Product output loop — using the product creates something public: a listing, a profile, a portfolio, a shared document, a review that becomes a discovery surface.
- Retention loop — returning customers raise lifetime value, which increases what acquisition can afford, which widens the surfaces the business can compete for.
Four loops on one product
Diagnosing a loop that is not turning
A loop that produces nothing is usually not the wrong model; it is a model with one broken step. The output is created but never discoverable. The referral is earned but never asked for. The content ranks but answers a question nobody buys against.
The content and search loop breaks most often at structure rather than at writing. Individual articles rank in isolation and never accumulate authority because nothing connects them — the reason content clusters and internal linking architecture are part of the growth model rather than an SEO detail.
Where campaigns are the right tool
Loops take time to turn, and there are situations where time is the constraint. Campaigns remain correct for a product launch with a fixed date, for seasonal demand, for entering a new market or region, for testing an offer before committing to build around it, and for reaching an audience that does not search for what the business sells.
| Situation | Better fit | Why |
|---|---|---|
| Fixed launch date | Campaign | Attention must exist on a specific day; loops cannot be scheduled |
| Seasonal peak | Campaign | Demand exists already and only needs to be captured |
| Established, recurring demand | Loop | The same question is asked continuously and can be answered once |
| New region or segment | Campaign first | Seeds the participants and evidence a loop needs to start |
| Expertise-led services | Loop | The work itself produces the material the loop runs on |
| Unproven offer | Campaign | Buys demand evidence faster than organic surfaces can |
Using campaigns to feed the system
The two models are complementary when campaign output is treated as an input rather than as a result. A campaign produces three things worth keeping: evidence about which messages and audiences respond, conversion data about where the path holds, and a first cohort of users who can start a loop turning.
Campaign output feeding the system
Capturing that value requires the campaign to run on the same architecture as everything else — the same modules, the same events, the same definitions. That is the practical case for landing-page systems.
What this changes in planning
Planning against loops changes the questions asked at the start of a quarter. Not only what will we run, but what will still be producing in twelve months. Not only what did this cost per acquisition, but what did this cycle leave behind for the next one.
Both questions belong to the same system. The wider argument sits in digital growth as a system.
Frequently asked questions
- What is a growth loop?
A repeating cycle in which the output of one user's activity becomes an input that helps acquire or activate the next user — a published page, a referral, a review, a shared artefact, a signal that improves discovery.
- How is a loop different from a funnel?
A funnel is linear: attention enters at the top, some fraction converts, and the next cycle starts again from paid or borrowed attention. A loop feeds its own input, so the cost of the next cycle falls as the loop turns.
- Are campaigns bad?
No. Campaigns are efficient at creating a starting condition and at buying information quickly. They become a problem only when they are the entire growth model, because they stop producing when the spend stops.
- Which loop should a business build first?
The one closest to something the business already does well. A firm producing genuine expertise should build the content and search loop; a service business with satisfied clients should build the referral loop.
- How long does a loop take to compound?
Long enough that most organisations abandon it early. Content and search loops typically need two to four quarters before the compounding is visible. Referral loops can move faster, but only where the product experience justifies a recommendation.
- Can a campaign start a loop?
That is usually the best use of one. Campaign spend can seed the first participants, produce the evidence a loop needs, and pay for the surface the loop will later fill for free.