Most marketing teams optimise for the wrong thing. They chase clicks, chase followers, chase impressions, and wonder why revenue stays flat. Growth marketing isn't about doing more marketing. It's about building a system that learns, compounds, and scales. The frameworks in this lesson have been used by companies like Airbnb, Dropbox, and Notion to grow from zero to billions, and they're entirely applicable to the brands you'll work with every day.
What Is Growth Marketing, Really?
Growth marketing is often confused with growth hacking, the scrappy, move-fast-and-break-things approach popularised during the early 2010s startup boom. But growth marketing is something more deliberate and more durable.
Here's what most marketing courses won't tell you: the brands that grow fastest are almost never the ones spending the most. They're the ones asking better questions. At Byter, we've onboarded clients who were running £30k a month in paid media with flat revenue, not because their ads were bad, but because their funnel was haemorrhaging value at the retention stage and nobody had bothered to look. Growth marketing is the discipline that forces you to look at the whole system, not just the bit you can put in a campaign report. It's slower to learn than performance marketing, and harder to sell to a client who just wants more Instagram followers, but it's the only approach that compounds.
According to McKinsey (2024), companies that embed experimentation into their marketing operations grow revenue 2.5 times faster than those that don't. That figure alone should reframe how you think about your role as a marketer. For context, a 2023 study by the Chartered Institute of Marketing found that UK marketing teams allocate, on average, 73% of their time to execution and just 11% to analysis and learning. That ratio is backwards for anyone serious about growth.
Growth marketers ask different questions. Not "how do we get more traffic?" but "where in our funnel are we losing people, and why?" Not "what's our monthly budget?" but "what's the highest-leverage experiment we can run this week?"
It is also worth distinguishing growth marketing from performance marketing. Performance marketing is primarily concerned with paid channels and measurable ROI on individual campaigns. Growth marketing is broader: it incorporates product thinking, behavioural psychology, data analysis, and cross-functional collaboration. A growth marketer might spend as much time talking to the customer success team as they do reviewing ad dashboards, because understanding why customers churn is just as strategically important as understanding how they were acquired.
This lesson introduces four of the most important frameworks in growth marketing, explains how to apply them practically, and shows you how to avoid the mistakes that trip up even experienced practitioners.
Framework 1: The AARRR Pirate Metrics Model
Developed by venture capitalist Dave McClure in 2007, the AARRR framework (affectionately known as "Pirate Metrics") remains one of the most widely used models in growth marketing because it maps the full customer journey into five measurable stages:
Acquisition: How do people find you?
Activation: Do they have a good first experience?
Retention: Do they come back?
Referral: Do they tell others?
Revenue: Do they pay?
The power of AARRR isn't the framework itself: it's the diagnostic discipline it forces. Most marketing teams instinctively focus on Acquisition because it's the most visible stage. But according to Bain & Company (2023), a 5% improvement in customer retention increases profits by 25–95%, depending on the industry. That's a staggering return compared to what most acquisition campaigns deliver.
When you map your funnel using AARRR, you're looking for the biggest leak: the stage where the most value is being lost. Fix that first, then move to the next. This prioritisation logic is what separates growth marketers from traditional campaign managers.
Real-world example: When Dropbox launched in 2008, their biggest leak wasn't at Acquisition. Their referral programme brought in enormous traffic. It was at Activation. New users would sign up, fail to install the desktop client, and never experience the core product value. Dropbox fixed this by redesigning their onboarding flow to guide users through a single critical action: putting one file in the Dropbox folder. That intervention, focused entirely on Activation, drove a 60% increase in retention. They didn't spend a penny more on advertising.
Similarly, early-stage e-commerce brands often discover their AARRR leak is at Revenue rather than Acquisition: thousands of people add items to their cart but abandon at checkout. Solving a checkout friction problem, whether that's slow load times, too many form fields, or limited payment options, can deliver more incremental revenue than doubling the paid media budget.
Practical application: Build a simple AARRR dashboard for any brand you're working with. Define one primary metric for each stage, benchmark your current performance, and identify which stage has the steepest drop-off. That's your growth priority. Review it monthly, not quarterly. Growth moves faster than most reporting cycles.
Framework 2: The ICE Scoring Model
Once you've identified where to focus, you'll typically have more ideas than time or budget. The ICE scoring model, popularised by Sean Ellis (who also coined the term "growth hacking"), gives you a fast, structured way to prioritise experiments.
Each idea is scored across three dimensions:
Impact: If this works, how significantly will it move the needle? (Score 1–10)
Confidence: How certain are you it will work, based on data or precedent? (Score 1–10)
Ease: How quickly and cheaply can you implement it? (Score 1–10)
Your ICE score = (Impact + Confidence + Ease) ÷ 3
An idea that scores 9 on Impact but 2 on Confidence and 2 on Ease gets a 4.3, lower than a more modest idea that scores 7, 7, and 8, which earns a 7.3. ICE stops teams from chasing shiny, ambitious ideas that will take six months to execute when a simpler experiment might deliver meaningful results in a fortnight.
ICE in practice: Suppose you're working with a SaaS brand and you've identified that Activation is the weak point. Your team generates four ideas:
Rebuild the entire onboarding experience with interactive walkthroughs (Impact: 9, Confidence: 5, Ease: 2 → ICE: 5.3)
Add a single welcome email sent 30 minutes after sign-up with one clear next step (Impact: 6, Confidence: 8, Ease: 9 → ICE: 7.7)
Add a progress bar to the existing setup wizard (Impact: 5, Confidence: 7, Ease: 8 → ICE: 6.7)
Host a weekly live onboarding webinar (Impact: 7, Confidence: 6, Ease: 5 → ICE: 6.0)
The welcome email wins, not because it's the most impressive idea, but because it's the highest-leverage one you can actually ship this week. Run it, measure it, learn from it, then move to the next. This is the compounding logic of growth marketing: lots of small wins add up to transformative results over time.
Warning
ICE scoring is a tool for alignment and prioritisation, not a substitute for strategic thinking. A high ICE score on a tactically sound but strategically misaligned idea can still lead you in the wrong direction. Always sense-check scores against your broader business objectives.
One useful evolution of ICE is the PIE framework (Potential, Importance, Ease), used frequently by conversion rate optimisation agencies. PIE weights "Potential", how much room for improvement exists at a given touchpoint, more heavily than pure impact estimates. For conversion-focused work, PIE can be a more precise tool; for broader growth work, ICE remains the more versatile starting point.
Framework 3: The North Star Metric Framework
One of the most common failure modes in growth marketing is optimising for metrics that don't actually reflect value creation. This is where the North Star Metric (NSM) framework becomes essential.
Your North Star Metric is the single number that best captures the core value your product or service delivers to customers. It sits between vanity metrics (followers, page views) and lagging indicators (annual revenue), and it acts as the compass for every growth decision.
Famous examples include:
Airbnb: Nights booked
Spotify: Time spent listening
Slack: Messages sent
HubSpot: Weekly active teams
Facebook (early growth): Monthly active users who visited at least 7 days in a month
WhatsApp: Messages sent per day
Notice what these metrics have in common: they all measure customer behaviour, not company outcomes. Revenue, profit, and market share are results of customers deriving value. The NSM framework asks you to measure the value itself.
The NSM framework, championed by Amplitude's growth team and detailed in their 2022 Product Analytics Playbook, works because it aligns marketing, product, and commercial teams around a shared measure of customer value. When everyone is pulling towards the same metric, experimentation becomes coordinated rather than chaotic.
A cautionary tale: A retail subscription box business might be tempted to use "monthly recurring revenue" as their North Star. But MRR is a financial outcome. It tells you that things are going well or badly, not why. A better North Star might be "boxes rated 4 stars or above by subscribers this month." That metric captures genuine customer delight, which is the underlying driver of renewal, referral, and expansion revenue. When that number drops, you investigate curation quality, packaging, or delivery experience, not ad spend.
For a digital marketing agency, a relevant North Star might be "number of clients hitting their KPI targets this month," because that's the truest measure of value delivered. Revenue is important, but it follows from this.
Practical application: Identify your client's North Star Metric before designing any growth strategy. If they can't articulate it, helping them define one is itself a high-value strategic contribution. A good NSM should be measurable weekly, directly influenced by the team's actions, and genuinely correlated with long-term business health.
Byter Tip
Byter Insider: We worked with a direct-to-consumer wellness brand based in Shoreditch that was tracking revenue as their primary growth metric. Everything looked fine on paper until we ran an AARRR audit and discovered their 30-day repeat purchase rate was sitting at 11%. The category benchmark is closer to 30%. We helped them redefine their North Star as "customers making a second purchase within 45 days," then rebuilt their post-purchase email flow using the Byter Retention Loop: welcome, engage, reward, remind, win-back. Within four months, their repeat purchase rate had climbed to 27% and their revenue per customer had increased by 34%, without spending an additional pound on acquisition. The North Star change is what made it possible. You can't fix what you're not measuring.
Framework 4: The Bullseye Framework for Channel Selection
Even the best growth strategy will stall if you're distributing your efforts across the wrong channels. The Bullseye Framework, developed by Gabriel Weinberg and Justin Mares in their book Traction (2014), helps growth marketers identify the one or two channels that will drive the majority of their growth.
The process works in three rings:
Outer ring (Brainstorm): List all 19 traction channels (SEO, paid social, content marketing, partnerships, PR, email, events, etc.) and generate one or two ideas for each.
Middle ring (Rank): Identify the six channels that seem most promising given your audience, budget, and stage of growth.
Inner ring (Prioritise): Run cheap, fast experiments on your top three channels to determine which has the best potential. Double down on the winner.
The Bullseye Framework is particularly useful for early-stage brands or new product launches, where spreading budget across every channel is the fastest route to mediocre results. According to Weinberg and Mares, most successful businesses find that one channel accounts for the majority of their growth at any given time.
Example application: A B2B SaaS startup targeting HR managers in mid-market businesses might brainstorm ideas across all 19 channels and rank their top six as: content marketing, LinkedIn paid ads, email outreach, partnerships with HR consultancies, webinars, and SEO. They run small experiments on the top three simultaneously: a blog series, a short LinkedIn campaign, and a cold email sequence to a list of 200 prospects. After six weeks, the email outreach generates 14 qualified demos at a cost per lead of £38. LinkedIn generates 6 demos at £190 each. Content marketing shows promise but needs six more months to compound. Decision: double the email budget, maintain content investment, pause LinkedIn.
This is how the Bullseye Framework works in practice. It's not about dismissing channels. It's about sequencing them intelligently and letting data, not intuition, guide allocation decisions.
The Bullseye Framework: a three-ring process for identifying and prioritising the traction channels most likely to drive growth
The Experimentation Engine: How Growth Compounds Over Time
Understanding the four frameworks above is valuable. But the real compounding effect of growth marketing comes from building an experimentation engine: a repeatable, documented process for generating hypotheses, running tests, and capturing learnings.
The most sophisticated growth teams, at companies like Booking.com, Netflix, and Monzo, run hundreds of experiments simultaneously. Booking.com famously runs over 1,000 A/B tests at any given time. That scale isn't achievable for most brands, but the underlying discipline is: every experiment should be documented, every result should be shared, and every failure should be treated as valuable data rather than a source of embarrassment.
A basic experimentation log should capture:
The hypothesis (If we do X, we expect Y to happen because Z)
The metric being measured
The sample size and test duration
The result and statistical confidence level
The decision made and the rationale
What to test next
Over time, this log becomes one of your most valuable strategic assets. Patterns emerge. You start to understand which levers move your specific audience, which messaging resonates, and which assumptions consistently prove wrong. That institutional knowledge compounds in ways that no single campaign ever can.
The 70/20/10 experimentation budget split is a useful heuristic for allocating effort: 70% of your resources go to proven channels and tactics, 20% to promising experiments with existing channels, and 10% to genuinely novel, high-risk bets. This ratio keeps the business stable whilst ensuring you're consistently discovering the next growth lever before the current one plateaus.
How the Frameworks Work Together
These four frameworks are not independent tools. They work best as an integrated system, and at Byter we run them in sequence every time we scope a new growth engagement. This is the thinking behind our Byter 3R Framework: every decision maps to Reach, Retain, or Revenue. AARRR tells you which R is leaking. The North Star Metric tells you how to measure it. ICE tells you what to fix first. Bullseye tells you which channel to use. The four frameworks slot together precisely because they answer four different questions that the 3R model surfaces.
Use AARRR to identify where to focus in your funnel
Use the North Star Metric to define what success looks like at that stage
Use ICE scoring to decide which experiment to run first
Use the Bullseye Framework to determine which channel will carry that experiment
In practice, this might look like the following for a direct-to-consumer skincare brand:
AARRR audit reveals a 74% cart abandonment rate. Retention (repeat purchase) and Revenue are the leaky stages
North Star Metric defined as "repeat customers placing a second order within 60 days"
ICE scoring surfaces two high-priority experiments: a post-purchase email sequence (ICE: 8.0) and a loyalty points programme (ICE: 5.7)
Bullseye analysis confirms email is the right channel to carry the experiment: it has the best signal speed and lowest cost to test
You run the email sequence for six weeks, measure repeat purchase rates, and use the learnings to inform the loyalty programme design. One framework feeds the next. That's growth marketing operating as a system rather than a collection of individual tactics.
ICE Scoring in action: four experiment ideas scored and ranked, the welcome email wins despite lower Impact, because Confidence and Ease are both high
Common Mistakes Growth Marketers Make
Even with solid frameworks in hand, practitioners regularly fall into the same traps. Here are five to watch for:
Optimising the wrong stage of the funnel. Pouring budget into acquisition while ignoring a broken onboarding experience is like filling a leaky bucket. Diagnose before you invest.
Running underpowered experiments. A/B tests that don't reach statistical significance teach you nothing, or worse, teach you the wrong thing. Always calculate required sample sizes before you start, not after. As a rough rule of thumb, most conversion tests require a minimum of 1,000 visitors per variant to be meaningful; email subject line tests typically need at least 500 recipients per variant.
Confusing correlation with causation. Your conversion rate went up the same week you launched a new campaign, but it also happened to coincide with a bank holiday. Isolate variables rigorously. If you can't control for a confounding factor, acknowledge the limitation explicitly in your reporting.
Abandoning channels too quickly. The Bullseye Framework calls for fast experiments, but "fast" doesn't mean "two days." Most channels require 4–8 weeks of consistent effort before returning meaningful data. Impatience is a growth killer. Content marketing and SEO in particular often require 3–6 months before meaningful signal emerges. Pulling the plug early is one of the most common and expensive mistakes in digital marketing.
Neglecting the compounding effect of retention. Acquisition is exciting and measurable in real time. Retention improvements are slower to surface but far more powerful over time. Build retention metrics into every growth review. A brand with a 90% monthly retention rate will have approximately 7× more customers after two years than one with an 80% rate, starting from the same acquisition volume.
Treating frameworks as checklists rather than thinking tools. AARRR, ICE, NSM, and Bullseye are lenses, not recipes. The practitioner's job is to apply judgement, challenge assumptions, and adapt these models to the specific context of each brand. Mechanical application without contextual thinking produces mediocre results regardless of which framework you're using.
Key Takeaways
Growth marketing spans the entire customer lifecycle, not just acquisition, and the biggest gains are often found in retention, not acquisition
The AARRR framework helps you diagnose where your funnel leaks and prioritise fixes based on where value loss is greatest
The ICE scoring model gives you a structured, democratic way to prioritise experiments, and prevents teams from defaulting to the most impressive rather than the most actionable ideas
Your North Star Metric should reflect real customer value, not vanity or purely financial measures, and it should be measurable weekly
The Bullseye Framework prevents the dilution of effort across too many channels simultaneously; most growth at any given stage comes from one primary channel
Building an experimentation engine, a documented, repeatable process for testing and learning, is what turns individual frameworks into compound growth over time
Statistical rigour, patience, and a focus on retention separate good growth marketers from great ones