How Growth Actually Works: Acquisition, Activation, Retention, Revenue and Compounding Growth

You cannot understand growth from one framework.

A useful way to think about it is through five different views of the same business:

• Value: Do customers get enough value to stay, pay, expand or return?
• Lifecycle: Where do people enter, activate, retain, refer and generate revenue?
• Accounting: What is adding to and subtracting from the customer or revenue base?
• Economics: Can the company afford to keep growing this way?
• Compounding: Does today's output help create tomorrow's users, revenue or distribution?

Across all five sits one decision:

What is the current constraint on durable growth?

That constraint might be demand. It might be acquisition. It might be activation, retention, monetisation, capacity or market saturation.

The goal is not to maximise every metric at once.

The goal is to understand the system well enough to know what is currently limiting it.

"Growth" is one of those words that becomes less useful the more broadly it is used.

A team might call any of these growth:

• more website traffic
• more leads
• more app installs
• more signups
• higher transaction volume
• more monthly active users
• more revenue
• more customers
• more referrals
• more accounts expanding

Some of those may be valid measures of growth for a specific business.

But activity is not automatically growth.

A better starting point is:

Growth is a sustained increase in the customer, usage or economic base that matters to the business.

The important words are sustained and matters.

If a campaign doubles signups for three days but almost none of those users activate or return, the campaign created activity.

If a discount doubles transaction volume for a week but the users disappear when the incentive ends, the business experienced a spike.

If new customers stay, generate value, and make the next period start from a stronger base, the system has grown.

This is why growth needs more than a top-line chart.

The first growth question is not:

How do we acquire more users?

It is:

Is there enough customer value here to justify acquiring more users?

That sounds obvious, but growth teams can spend months optimising distribution into a product that is not yet creating durable value.

Recent product-market-fit guidance from J.P. Morgan makes the same distinction. Returning, referring, paying and expanding behaviour are stronger signals of sustained demand than initial activity alone, and scaling before demand is established can become expensive quickly. Source

This does not mean every company must wait for some magical moment called "product-market fit" before doing any marketing.

It means growth spending multiplies the system you already have.

If the product creates value, distribution can amplify it.

If the product leaks value, distribution can amplify the leak.

Useful signals depend on the business, but they often include:

• cohorts that keep using or transacting
• repeat purchases
• expanding accounts
• willingness to pay
• referrals or organic discovery
• customers returning without constant incentives
• qualitative evidence that the product solves an important problem

There is no universal retention number or one survey result that proves a company has found product-market fit.

The question is whether behaviour shows that value persists.

One of the most useful distinctions in growth is between flow and stock.

Acquisition is flow.

New users or customers enter the system.

Retention builds stock.

Some of those people remain part of the active customer, usage or revenue base.

Imagine a business that acquires 10,000 new customers every month.

That sounds impressive.

But if 9,500 customers disappear each month, the business has to keep replacing almost the entire inflow just to move forward.

Now imagine another business acquiring only 4,000 customers a month, but losing 1,000.

The first company wins the acquisition chart.

The second may be building the stronger base.

This is why a growth chart needs to be decomposed.

Suppose a product has:

100,000 active users in January

and:

105,000 active users in February

The dashboard says:

+5% growth

Useful, but incomplete.

There are many ways to arrive at 105,000.

Illustrative example, not benchmark data.

Both grew 5%.

But they are not the same growth system.

Business A is replacing a huge amount of lost activity every month.

Business B needs far less new acquisition to produce the same top-line growth.

This is the basic idea behind growth accounting.

Amplitude's growth-accounting framework separates active users into four groups:

• New: active for the first time in the current period
• Retained: active in both the previous and current period
• Resurrected: active before, inactive in the previous period, and active again now
• Churned: active previously, but inactive now

For active users:

Ending active users = Retained + New + Resurrected

And the change in the active base can be written as:

Net change = New + Resurrected - Churned

Amplitude's growth accounting explanation

This is much more useful than saying:

MAUs grew 5%.

It tells you what produced the 5%.

Revenue growth accounting adds two more movements

Revenue can expand or contract without the customer disappearing.

So a revenue view may include:

• new revenue
• retained revenue
• resurrected revenue
• expansion
• contraction
• churn

Using the conventional decomposition:

Current-period revenue = Retained revenue + New revenue + Resurrected revenue + Expansion

To explain the change from the previous period, separate gains from losses:

Net revenue change = New + Resurrection + Expansion - Contraction - Churn

Here, retained revenue is the portion carried forward without expansion. Contraction and churn sit on the loss side of the change equation.

The exact accounting should match how your business defines activity and revenue.

The principle is more important than the notation:

A top-line growth rate tells you what happened. Growth accounting starts telling you why.

A quick ratio can help, but it is not a health score

For a user-based model, one simple diagnostic is:

Quick ratio = (New + Resurrected) / Churned

If the ratio is above 1 for a period, user gains exceeded user losses in that period. If it is below 1, losses exceeded gains.

Amplitude uses this ratio as a quick pulse on the composition of growth. Source

Do not turn it into a universal benchmark. Two companies with the same ratio can still have very different retention, economics, market size and growth quality.

Growth becomes harder to interpret as the base gets larger.

Going from 100 customers to 200 is 100% growth.

Going from 100,000 customers to 110,000 is 10% growth.

The second business added 9,900 more customers than the first one.

Yet its growth percentage is far lower.

So a falling growth rate is not automatically evidence that the engine is broken.

Sometimes the denominator simply became much larger.

The useful diagnostic is to ask what changed underneath:

• Did new customer volume fall?
• Did acquisition quality deteriorate?
• Did retention decline?
• Did resurrection improve?
• Did expansion slow?
• Did churn increase?
• Did the business enter a more saturated market?

Growth rate is an output.

The components underneath it are where the decisions live.

Growth accounting describes what changed in the base.

The lifecycle view asks how people move through the business.

A widely used framework is AARRR, introduced by Dave McClure:

Acquisition → Activation → Retention → Referral → Revenue

McClure's original "Startup Metrics for Pirates" framework was designed to focus startups on a small set of actionable customer behaviours rather than vanity metrics. Original framework overview

The stages are useful because they force different questions.

AARRR is useful.

But it is not the complete growth model.

Lean Analytics makes an important point: these stages do not necessarily happen in a strict sequence. A user might refer before paying, return several times before registering, or monetise before creating a referral. O'Reilly overview

That is one reason to treat AARRR as a diagnostic lens, not a law of how every customer behaves.

AARRR answers:

Where in the lifecycle are people moving forward or dropping out?

That is valuable.

But it does not fully answer:

What changed our active base?

That is what growth accounting helps explain.

It does not fully answer:

Can we afford this growth?

That requires unit economics.

And it does not fully answer:

Does this cohort help create the next cohort?

That is where growth loops and compounding become useful.

These frameworks should not compete with each other.

They answer different questions.

Lifecycle view: Where is behaviour breaking?

Accounting view: What is adding to and subtracting from the base?

Economics view: Is the growth financially sustainable?

Compounding view: Does today's output create future input?

The mistake is trying to make one framework answer all four.

Acquisition usually gets disproportionate attention because it is visible.

Traffic increases.

Leads increase.

Installs increase.

The chart moves immediately.

But a useful acquisition review needs at least four dimensions.

1. Volume

How many relevant users or customers did the channel bring in?

2. Quality

What happened after they arrived?

Did they activate?

Did they retain?

Did they pay?

Did they expand?

3. Economics

What did that acquisition cost?

And what did those customers eventually contribute?

4. Durability

Can the channel keep producing similar quality at similar economics?

A channel that sends 50,000 signups can be worse than one that sends 8,000 if the second group activates and retains substantially better.

This is why acquisition metrics should be cohorted by source whenever possible.

The question is not:

Which channel gave us the most users?

It is:

Which channel gave us the most useful future customer value relative to its cost?

Activation is one of the most abused words in growth.

A user registered.

Are they activated?

A customer completed KYC.

Activated?

They installed the app.

Activated?

They finished onboarding.

Activated?

Maybe.

But those are often process milestones, not necessarily value milestones.

A better definition is:

Activation is the earliest behaviour that credibly indicates a user has experienced meaningful product value.

That definition is deliberately contextual.

For a collaboration tool, activation might involve completing meaningful work with another person.

For a marketplace, it may be a successful match or transaction.

For a remittance product, it may be completing a successful transfer and receiving the expected outcome.

For a B2B platform, activation may require multiple users or an account-level workflow.

A good activation metric should help you predict meaningful future behaviour.

But one warning matters:

Prediction is not causation.

If users who complete a certain action retain better, that action may be a useful activation signal.

It does not automatically mean forcing every user to complete the action will cause retention to increase.

More motivated users may simply be more likely to do both.

Use behavioural data to find candidate activation signals.

Then use product analytics, research and experiments to understand the mechanism.

Two users can eventually reach the same value but have completely different experiences.

One reaches it in two minutes.

Another reaches it in two days.

The longer the path, the more chances there are for uncertainty, friction, distraction or abandonment.

So activation should often be paired with:

Time to Value

Ask:

• How long does it take a new customer to experience meaningful value?
• Which steps are necessary?
• Which exist because of internal process?
• Does time to value differ by channel, segment or use case?
• Which delays correlate with abandonment?

The goal is not always "shorter."

A financial product may require necessary compliance steps.

An enterprise tool may require implementation.

The goal is to remove unnecessary delay between intent and value.

Acquisition tells you that someone was interested enough to arrive.

Activation tells you that they experienced some initial value.

Retention asks whether that value was strong or recurring enough to matter again.

This is why retention is such an important signal.

But "retention" needs a denominator, an event and a natural time interval.

A payroll product, food delivery app, tax-filing product, B2B workflow tool and international remittance app should not all be judged on the same D7 or D30 retention metric.

The right question is:

Retention of whom, doing what, at what interval that matches the real use case?

For some businesses that means:

• weekly active use
• monthly transactions
• quarterly account activity
• annual renewal
• repeat purchase within a category-specific window
• revenue retention rather than logo retention

This is also why benchmarks should be used carefully.

A "good" retention number without business-model context can be worse than useless because it encourages teams to compare unlike products.

Retention is not only a downstream metric.

It changes what the company can afford to do upstream.

Better retention can:

• increase customer lifetime value
• reduce how much new acquisition is required to maintain the base
• improve payback economics
• increase opportunities for expansion
• create more referrals or word of mouth
• create more reviews, content or network participation

A16z makes this point directly in its discussion of retention and go-to-market investment: increasing spend into weak retention simply adds more customers to a leaky bucket. A16z retention analysis

That does not mean retention is always the bottleneck.

It means poor retention puts a ceiling on how efficiently many other growth investments can work.

Many growth discussions divide the world into:

active users and churned users.

That hides a third possibility.

People can come back.

A resurrected customer is someone who used the product previously, became inactive, and later became active again.

Amplitude's retention lifecycle explicitly separates new, current and resurrected users because each group represents a different growth opportunity. Retention Lifecycle Framework

Resurrection matters particularly for products where usage is naturally intermittent:

• travel
• ecommerce
• financial transactions
• marketplaces
• seasonal products
• infrequent workflows

A dormant customer who already understands the product may be very different from a completely new prospect.

But resurrection needs the same discipline as acquisition.

Ask:

• Why did they come back?
• Did they return because the underlying need reappeared?
• Did a product improvement matter?
• Did a lifecycle message work?
• Did they become retained again?
• Did they immediately disappear after one transaction?

A reactivation spike is not automatically durable resurrection.

A product can have:

• strong acquisition
• good activation
• high engagement
• impressive retention

and still be a weak business.

Eventually, the company has to capture enough economic value from the value it creates.

That does not always mean charging every user directly.

Revenue may come from:

• subscriptions
• transactions
• advertising
• commissions
• usage
• enterprise contracts
• financial spreads
• marketplaces
• expansion
• cross-sell

The growth question is:

Does the monetisation model support the customer experience and the economics required to keep investing?

Revenue is not merely the last stage of a funnel.

It changes what the company can afford to build, support and acquire.

Imagine this simple case:

It costs ₹2,000 to acquire a customer.

Over the customer's economically relevant lifetime, the company earns only ₹1,300 in contribution from that customer.

Acquiring more of those customers makes the top line grow faster.

It also destroys more value.

That is why growth needs an economics view.

Important measures can include:

• Customer Acquisition Cost (CAC)
• contribution margin
• gross margin
• lifetime value (LTV or CLTV)
• CAC payback
• expansion
• churn
• cost to serve

CAC payback asks how long it takes to recover the cost of acquiring a customer. Bessemer describes it as a core measure of sales and marketing efficiency because capital tied up in acquisition cannot be fully reinvested until it has been earned back. Bessemer on CAC payback

You will often see universal-looking benchmarks such as a 3:1 LTV:CAC ratio.

Treat those as context, not laws.

Business model, gross margin, cash availability, customer segment, contract length, retention and growth stage can all change what is acceptable.

The principle is simpler:

The value captured from a customer eventually needs to justify the cost of acquiring and serving that customer.

We will cover CAC, LTV and payback in a dedicated guide rather than turning this article into a finance textbook.

The mechanics are shared.

The metric that represents durable growth is not.

This matters because copying a growth dashboard from another company can create false confidence.

A marketplace may look healthy on buyer growth while supply quality deteriorates.

A subscription product may grow logos while revenue retention weakens.

A fintech product may grow registered users while repeat transaction behaviour stays flat.

The growth model should start from how the business creates and captures value, not from whichever dashboard template is fashionable.

Once the system is visible, the most useful question becomes:

What is currently preventing materially more durable growth?

Not:

Which metric is lowest?

A metric can be low without being the limiting factor.

The constraint is the part that, if improved meaningfully, would unlock the greatest system-level improvement.

Possible constraints include:

Demand constraint: The market does not want the product enough.

Acquisition constraint: Value and retention are strong, but too few qualified people discover the product.

Activation constraint: People arrive, but too few experience meaningful value.

Retention constraint: People reach value, but it does not persist strongly enough.

Monetisation constraint: Customers get value, but the company does not capture enough of it.

Capacity constraint: Demand exists, but operations, supply, risk, compliance or service capacity cannot support more volume.

Market constraint: The current addressable segment is becoming saturated.

Compounding constraint: The business grows, but every new cohort has to be acquired almost from scratch.

This is the difference between measuring growth and operating growth.

Imagine you are running growth for a marketplace.

Customer acquisition is not accelerating fast enough.

Which lever do you pull?

At TaskRabbit, the underlying problem was closer to D.

Former growth leader Mike Rothenberg later described a "leaky bucket" issue: when a new customer's first task was not successfully completed, that customer almost never returned.

The company spent too much time and money trying to brute-force growth through acquisition instead of fixing the matching problem underneath it. TaskRabbit later relaunched its platform, and Rothenberg said growth reaccelerated afterward. First Round Review

The lesson is not:

Always fix retention before acquisition.

The lesson is:

Do not assume the visible growth metric identifies the real constraint.

Acquisition was where the growth problem appeared.

Successful value delivery was where the problem lived.

Funnels are useful.

They help teams ask:

• Where are people dropping?
• Which step has the biggest conversion loss?
• Did a product change improve progression?
• Which segment moves differently through the lifecycle?

The limitation is that funnels usually move in one direction.

Input → conversion → output

A funnel does not naturally show how an output becomes a future input.

That is the purpose of a growth loop.

Reforge describes loops as closed systems where an input passes through a process, creates an output, and some of that output can be reinvested into the next cycle. Reforge on growth loops

A simple referral loop might be:

New customer → receives value → invites friend → friend becomes customer → friend receives value → invites another friend

A content loop might be:

User creates useful content → content becomes discoverable → new visitor arrives → visitor becomes user → creates more content

A revenue loop might be:

Customer generates contribution → contribution funds acquisition → acquisition produces customers → customers generate more contribution

Funnels tell you where conversion breaks.

Loops tell you whether output can feed future growth.

Use both.

When people hear "growth loop," they often think:

referral program

or:

viral invite

Those are only two possibilities.

Growth can compound through:

Word of mouth. A good experience creates future acquisition without directly buying every impression.

User-generated content. Users create pages, assets, templates, listings or other content that attracts future users.

Network effects. More participation increases product value, which attracts or retains more participation.

Revenue reinvestment. Healthy contribution funds more distribution, product investment or sales capacity.

Expansion. Existing customers increase usage, seats, spend or scope.

Data and product improvement. More usage can improve recommendations, matching or product quality, which improves future value.

Sales learning. More customers create proof, references and operational learning that can improve future conversion.

The important question is not:

Do we have virality?

It is:

Does today's output improve tomorrow's ability to create value, acquire customers, retain them or monetise them?

Quizlet is a useful contrast to the TaskRabbit case.

When growth leader Javier Ortega joined Quizlet, the company already had major growth from word of mouth and organic search powered by user-generated study content.

But the team still needed a clearer model of how those mechanisms worked.

They mapped four core growth loops:

• word of mouth
• UGC-driven SEO
• paid acquisition
• re-engagement

and identified product actions such as studying, creating content, inviting friends and upgrading that powered those loops. Reforge case study

The important part is not the exact set of loops.

It is the operating behaviour.

The team moved from:

We know SEO and word of mouth are important.

to:

We understand the actions, inputs and outputs that make each loop work.

That turns growth from a collection of channels into a model the company can diagnose.

Working in growth has made me suspicious of spikes that look excellent in weekly reporting.

A promotion can increase transaction volume immediately.

A campaign can create a wave of signups.

PR can produce a sudden traffic jump.

An influencer can generate a large cohort.

All of those can be valuable.

But the harder questions come later:

• Did these users reach the same value?
• Did they retain?
• Did they return without the incentive?
• Did the unit economics still work?
• Did the cohort create any future acquisition, expansion or revenue?
• Did the base remain stronger after the campaign ended?

This is the distinction I use:

A spike is what happens when you push the system. Growth is what the system preserves or compounds after the push.

That does not mean every campaign must create permanent users.

A launch can be strategically useful even when some of its effect fades.

The point is to know whether you are measuring a temporary event or a stronger growth system.

At this point, the pieces fit together.

I find it useful to look at the same business through five views.

1. Value

Do people want this enough to continue?

Look at:

• repeat behaviour
• retention
• successful outcomes
• willingness to pay
• expansion
• qualitative pull

This is the foundation.

2. Lifecycle

Where are people moving or dropping?

Look at:

• acquisition
• activation
• retention
• referral
• revenue

AARRR is useful here.

3. Accounting

What is adding to and subtracting from the base?

Look at:

• new
• retained
• resurrected
• churned
• expansion
• contraction

This tells you the composition of growth.

4. Economics

Can the business sustain this growth?

Look at:

• CAC
• margin
• LTV
• payback
• cost to serve
• cash required to scale

This tells you whether growth is financially supportable.

5. Compounding

Does the output of one period help the next one?

Look at:

• referrals
• word of mouth
• UGC
• network effects
• expansion
• reinvestment
• product/data improvements

This tells you whether growth can become more self-reinforcing.

Across all five:

Find the current constraint.

That is the operating layer.

Use this table as a starting point when growth is underperforming.

The table does not tell you what the answer is.

It helps you ask the next useful question.

When a growth metric drops, the temptation is to brainstorm solutions immediately.

I would start with diagnosis.

First 10 minutes: verify that the change is real

Before building a theory:

• Did the metric definition change?
• Did tracking break?
• Did a data pipeline delay occur?
• Is seasonality involved?
• Did the reporting window change?
• Did one large account or cohort distort the number?

Never spend a week solving an instrumentation problem with product work.

Next 10 minutes: decompose the growth

Break the top line into whatever applies:

• new
• retained
• resurrected
• churned
• expansion
• contraction

Ask which component actually changed.

Next 10 minutes: cohort the change

Look by:

• acquisition source
• signup period
• customer segment
• geography
• product version
• platform
• pricing plan
• relevant use case

Aggregate numbers hide where the problem lives.

Next 10 minutes: inspect the lifecycle

If new customers declined, inspect acquisition.

If acquisition stayed healthy but the base slowed, inspect activation and retention.

If customer growth is healthy but economics deteriorated, inspect monetisation and cost.

If the loop weakened, inspect the product action or distribution mechanism feeding it.

Final 5 minutes: write competing hypotheses

Do not write:

Retention is down.

Write something closer to:

Retention decline appears concentrated among customers acquired from Channel X after onboarding step Y changed.

Then add at least one competing explanation.

For example:

It may also reflect a seasonal shift in customer mix rather than the onboarding change itself.

Now you have something testable.

A healthy growth system does not mean every metric rises every week.

It means the important parts of the system reinforce rather than quietly cancel one another.

Typically:

• acquisition brings customers with plausible fit
• activation gets them to meaningful value
• retention shows that value persists
• resurrection recovers some dormant value where appropriate
• monetisation captures enough value to sustain the model
• unit economics permit sensible reinvestment
• outputs create some future acquisition, retention, expansion or efficiency
• measurement makes the next constraint visible

The exact balance varies by business.

A marketplace and a B2B SaaS company should not look identical.

A seasonal financial product and a daily consumer app should not have the same retention expectation.

A mature company and a pre-product-market-fit startup should not optimise the same constraint.

That is why growth strategy cannot be reduced to one universal funnel.

A growth model improves diagnosis.

It does not remove uncertainty.

It cannot tell you:

• whether the market will change next year
• whether a competitor will redefine the category
• whether customers will accept a future price
• whether a correlated activation behaviour is causal
• whether a campaign will work before it is tested
• whether a company's current product-market fit will persist indefinitely

It also cannot make a bad product good.

A precise dashboard can describe weak value very accurately.

The purpose of the model is to help teams connect evidence to decisions.

Not to make uncertainty disappear.

This cornerstone is the system map.

Several topics deserve their own focused guides because compressing them here would make this article less useful.

AARRR

How to define acquisition, activation, retention, referral and revenue for different business models.

Funnels vs growth loops

When a funnel is the right diagnostic, when a loop is the right model, and why both can coexist.

North Star Metric

How to identify a measure of customer value that can guide a company without becoming another vanity metric.

CAC vs LTV vs Payback

How acquisition economics work, where common formulas fail, and why benchmarks need context.

Growth equations and models

How to turn growth components into a quantitative model that helps teams forecast and diagnose.

Acquisition vs activation

Why adding more users can be the wrong answer when new customers are not reaching value.

Growth experimentation

How to run experiments that change the growth system rather than produce isolated metric wins.

These supporting guides will connect back to this page as the cluster grows.

What is a growth model?

A growth model is a structured explanation of how a business creates, retains, loses and compounds customers, usage or revenue over time. A useful growth model connects customer value, lifecycle behaviour, growth accounting, economics and the mechanisms that produce future growth.

What is the AARRR framework?

AARRR is a startup metrics framework associated with Dave McClure. It stands for Acquisition, Activation, Retention, Referral and Revenue. It is useful for examining customer lifecycle behaviour, but it does not by itself explain growth accounting, unit economics or compounding loops.

What is the difference between a growth funnel and a growth loop?

A funnel shows how people progress through stages and where conversion drops. A growth loop shows how an output from one cycle can become an input into the next cycle. Funnels are useful for conversion diagnosis. Loops are useful for understanding compounding.

Why is retention important for growth?

Retention preserves value from previous acquisition. Better retention can reduce the amount of new acquisition required to grow the base, increase customer lifetime value, improve payback economics and create more opportunities for referral, expansion or other compounding effects.

Should a startup focus on acquisition or retention first?

There is no universal answer. If demand and retention are strong but too few qualified users discover the product, acquisition may be the constraint. If large numbers of users arrive but do not stay, scaling acquisition may amplify the leak. Diagnose the current constraint rather than choosing a function by default.

What is activation?

Activation is the earliest behaviour that credibly indicates a customer has experienced meaningful product value. The correct activation event depends on the product and should not automatically be equated with signup, onboarding completion or another process milestone.

What is growth accounting?

Growth accounting decomposes top-line growth into the components that created it. For active users, that often means new, retained, resurrected and churned users. For revenue, expansion and contraction may also matter.

What causes growth to compound?

Growth compounds when outputs from today's customers or activity improve the business's ability to create future value or acquire, retain or monetise future customers. Examples include referrals, word of mouth, user-generated content, network effects, expansion and reinvestment of healthy contribution into further growth.

How do you identify a growth bottleneck?

Start by decomposing the top-line metric, then inspect lifecycle movement, cohorts, economics and compounding mechanisms. The bottleneck is not simply the lowest metric. It is the constraint whose meaningful improvement would unlock the largest increase in durable growth.

What metrics should a startup track for growth?

The exact metrics depend on the business model. Most teams need some combination of acquisition quality, activation, retention, resurrection, revenue, churn, unit economics and the metrics that describe their most important growth loops. Track the smallest set that helps you make decisions rather than collecting every possible metric.

1. Dave McClure, Startup Metrics for Pirates
2. O'Reilly, Analytics Frameworks for the Startup Lifecycle
3. Amplitude, Growth Accounting
4. Amplitude, Retention Lifecycle Framework
5. Reforge, Growth Loops Are the New Funnels
6. Reforge, How to Build an International Growth Loop Model
7. First Round Review, From Zero to IPO: How Growth Needs to Evolve at Every Startup Stage
8. J.P. Morgan, Product-Market Fit: Why Lasting Demand Matters for Startups
9. Bessemer Venture Partners, Scaling to $100 Million
10. Andreessen Horowitz, Retention Is All You Need

I’m Saurav Mishra. I work across growth, product, marketing and analytics, and I also build and experiment with digital products using modern AI and development tools.

LinkedIn · Explore my case studies →