Growth as a model, not a channel
Ask a channel-led team how the business grows and the answer is a list: paid search, social, email, partnerships. Ask a growth team and the answer is a description of a mechanism: customers arrive this way, they get value when they do this, they stay for this reason and they bring others through this path. The list is a budget; the mechanism is a model, and only a model can be improved deliberately, because only a model says where the leaks are.
That shift has consequences for how the work is judged. A channel is judged on its own report, which is why channel teams argue about attribution. A model is judged on whether customers acquired through it are worth more than they cost, which is a question about lifetime value (LTVTotal value a customer brings over time.) against acquisition cost (CACCost to acquire one new customer.), and on whether the leading indicator of value, the north star, is moving. The growth economics section owns the first question; this section owns the model and the north star. The growth marketing hub sets both in the wider picture, including how growth sits alongside brand.
A model also changes what counts as marketing work. Onboarding, pricing pages, the referral prompt after a good experience and the email that brings a lapsed customer back are all parts of the mechanism, and a growth team will work on whichever part is leaking most, whether or not it sits in a traditional marketing remit.
Comparison, loops, cadence: the core guides
The core guides are best read in order.
Growth marketing versus traditional marketing sets the growth operating model against the campaign-and-channel model most teams inherit: scope, cadence, metrics, team shape and the role of brand. It is a comparison rather than a verdict, because most companies run both and the useful question is how they share the economics.
Growth loops in practice is the core of the section. It explains why a loop compounds where a funnel restarts, catalogues the common loop types (viral, content, paid, sales) with what each needs to turn, and shows how to draw a loop, find its leak points and measure its throughput and cycle time. It closes with loops for businesses that do not sell software, where the loop is usually there but slower and less instrumented.
Experimentation cadence covers the operating rhythm that turns a backlog of ideas into learning: prioritisation, weekly and sprint rituals, the trade between test velocity and test quality, and the learning log. A single A/B TestControlled comparison of two variants. is a CRO matter; the programme that decides which tests to run and what to do with the results is a growth matter, and the guide keeps to that boundary.
Further guides
Further guides take the cadence into the weekly meeting, the loop model into a business with two sides and into the arithmetic of how a loop compounds, the north star into the decision to change it, and the programme as a whole into buying help in:
- Running a weekly growth meeting — a tutorial for the weekly hour: who attends, a pre-read of the north star and its inputs, the test read-out, decisions, capacity and the learning log.
- Growing a two-sided marketplace — finding the side that limits matches, starting narrow, acquiring each side differently, protecting trust and expanding area by area.
- How a growth loop compounds — how the customers a loop produces add up generation by generation, when they arrive, what wears the factor down and what a loop is worth.
- Changing a north star metric — telling a north star that has stopped working from a loose definition or a weak input, and re-basing key results, inputs, the model and tests when it changes.
- Hiring and evaluating a growth marketing agency — what stays with the business when a provider runs part of the programme, the questions and warning signs, the reports to test and how to leave.
The north star metric
A north star metric is the single measure a growth team steers by: a count of the value customers receive, chosen because it leads revenue rather than lagging it. It is a KPIA metric tied directly to an objective. with a particular job, which is to stop every experiment being argued on a different number.
| Kind of business | Candidate north star | Why it leads revenue | Inputs beneath it |
|---|---|---|---|
| Subscription software | Accounts completing the core action each week | Active accounts renew; inactive ones churn before renewal | Sign-ups, activation rate, weekly active share |
| Online retailer | Repeat orders per customer per quarter | Repeat buyers carry most of the margin | First-order rate, second-order rate, order frequency |
| Marketplace | Transactions completed between matched parties | Both sides stay when matches happen | Supply listings, demand searches, match rate |
| Content or media | Returning readers per month | Returning readers are the audience advertisers and subscribers pay for | New readers, return rate, sessions per reader |
| Services firm | Qualified proposals sent per month | Proposals convert at a steady rate to engagements | Enquiries, qualification rate, proposal rate |
| Meal kits or consumables | Boxes or units consumed per customer per month | Consumption predicts renewal and referral | Activation, pause rate, cancellation rate |
Two tests tell a candidate apart from a vanity measure. It must move before revenue does, so that a change in it is a warning or a promise rather than a post-mortem; and it must be something the team can plausibly act on this quarter through the inputs beneath it. The north star framework in this site's library adds the objectives-and-key-results layer that connects the metric to quarterly targets, and the objectives and KPIs guide covers the wider discipline of choosing measures that match a business objective. When a north star chosen this way stops leading revenue or starts being moved without creating value, the guide to changing a north star metric covers the decision and the switch.
The growth model document
The growth model is a short document, rarely more than two pages, that a team keeps current and reads at the start of every quarter. It contains the model sentence (arrive, activate, stay, refer), a drawing of the main loop with its measured leak points, the north star and the input tree beneath it, a summary of the economics with the definitions used for lifetime value and acquisition cost, the current assumptions the team knows to be shaky, and the date of the next review. The growth economics section supplies the numbers; this document records which numbers were used and why, so that a new analyst can rebuild them and a sceptical stakeholder can see the assumptions before arguing with the conclusions. The spreadsheet growth model tutorial carries the same inputs forward month by month, so that a target can be checked against what the inputs can deliver before anyone commits to it.
A language-learning app's growth model on one page
The app's model sentence reads: learners arrive from app-store search and from friends' shared streaks, activate when they complete a lesson on three separate days in the first week, stay while their streak is alive, and refer when they share a milestone. The loop drawing shows four leak points, with the measured drop at each: 38% of installers never finish a first lesson; 51% of those who do never reach three days; 12% of weekly actives lapse each month; and only 6% of milestone reachers share one. The north star is "learners completing a lesson on five or more days this week", with inputs of installs, first-lesson rate, three-day activation rate and weekly lapse rate. The economics line reads: blended CAC £3.40, LTV on contribution £9.10 for the average install and £26 for a three-day activator, payback inside two months for paid search installs and over five months for paid social. The assumptions section flags that the lapse rate is measured only on learners who allowed notifications. Review date: first week of next quarter.
Common fundamentals problems
Most difficulties in a growth programme trace back to a missing or wrong fundamental. The table maps the symptom to the guide that treats it.
| Symptom | Usual cause | Guide |
|---|---|---|
| Growth stalls whenever paid spend is paused | The model is a funnel fed by paid, with no loop behind it | Growth loops in practice |
| A strong loop is valued at its first round of invitations, or credited with customers still to come | Later generations left out, or counted before they have arrived | How a growth loop compounds |
| Experiments win individually but the business does not grow | Tests judged on local metrics, not the north star or economics | Experimentation cadence |
| Growth and brand teams compete for the same budget | No shared economics; two operating models run as rivals | Growth versus traditional marketing |
| Every meeting argues about which number matters | No north star; each function reports its own | This page, then the objectives and KPIs guide |
| The north star keeps rising while revenue and retention do not | The north star no longer leads revenue, or is being moved without creating value | Changing a north star metric |
| Test velocity is high but nothing is remembered | No learning log; results live in slide decks | Experimentation cadence |
| Acquisition grows while profit falls | Loop throughput measured without lifetime value and payback | Growth economics |
| The weekly meeting reviews dashboards and decides nothing | No standing agenda or pre-read; decisions not written down before the hour ends | Running a weekly growth meeting |
| A provider reports wins and a falling acquisition cost that the business's own cohorts and ledger do not show | Results on definitions the provider controls, tests called by their builders, or a cost that leaves out the provider's fee | Hiring and evaluating a growth marketing agency |
| Buyers find no supply in some areas while acquisition spend rises | Buyer spend aimed where supply limits matches; liquidity read as one average | Growing a two-sided marketplace |
The guides in this section assume the measurement foundation exists; where it does not, the growth marketing hub describes the first plan that puts cohorts and a north star in place before the experiments begin.
