AARRR Metrics for eCommerce: The 5 Stages Explained

The first time you see "AARRR metrics" written down, it looks like a typo. It is not. AARRR is a growth framework that splits a business into five stages, each with its own small set of metrics: Acquisition, Activation, Retention, Referral and Revenue. Say the acronym out loud and you sound like a pirate, which is why it is also known as pirate metrics. That is the only pirate-related thing about it.

The framework was built for startups, so using it in an online store takes translation. This guide does that translation stage by stage: the question each stage answers, the one metric that answers it, the formula, and a worked example. It also says plainly where the model fits eCommerce badly, because it does.

What the AARRR pirate metrics framework is

McClure's argument was that most young companies drown in data and still cannot say what to fix. Five stages, five questions, one primary number each. Each stage answers a question about a group of people:

• Acquisition — how many people find us?
• Activation — how many take a first meaningful step with us?
• Retention — how many come back?
• Referral — how many tell someone else?
• Revenue — how many pay us?

The value is not in the acronym. It is in the discipline of assigning one primary number per stage and agreeing, as a team, which stage is currently the constraint.

Why eCommerce reorders the AARRR funnel

That reordering changes what each stage is about. In SaaS, retention is usage and revenue is the conversion of a habit into a subscription. In a store, revenue is the first transaction and retention is the whole repeat-purchase problem: getting a one-time buyer to become a customer.

Rewritten for a store, the five questions become:

• Acquisition — how many people do we attract, and what does that cost?
• Activation — how many of them show real intent?
• Revenue — how many place a first order, and how large is it?
• Retention — how many order again?
• Referral — how many recommend us to someone else?

The five AARRR stages as eCommerce metrics

The table below is the short version. The worked examples all come from one fictional store with 150,000 sessions, 3,000 orders and $360,000 in revenue in a quarter.

Acquisition

Visitors and potential customers discover your brand, your content and your products. The job here is not more traffic, it is better traffic at a defensible cost.

Two things make this stage useful rather than decorative. First, calculate CAC by channel, not blended, or the good channels subsidize the bad ones invisibly. Second, judge CAC against lifetime value, not against first-order revenue. A $30 CAC is excellent for a customer worth $400 and ruinous for a customer worth $35. Our guide on how to reduce customer acquisition cost covers the full calculation.

The best acquisition targeting starts from your existing best customers: the people who order frequently, at high value, and stay. Build lookalike audiences from that list rather than from all buyers.

Activation

Activation is the first committed action, before any money changes hands: an account, a newsletter subscription, a wishlist, an add to cart, a started checkout. Marketers call the underlying moment the "aha moment", when the visitor understands what your product does for them.

Pick one activation action and stick with it, or the number means nothing quarter to quarter. Add to cart is the usual choice for a store, because it is the earliest action that reliably predicts a purchase.

Improving activation is a messaging problem before it is a design problem. Run jobs-to-be-done interviews with your best customers to learn what they were actually trying to accomplish, then make sure the landing pages and product pages say that back to them.

Revenue

The revenue stage is the first order. Two numbers describe it: how many sessions turn into orders, and how much each order is worth.

Most of the work at this stage is removing friction and doubt at checkout: transparent shipping costs and delivery dates, visible return policy, payment options, and a mobile checkout that does not punish thumbs. If you do not know where to start, ask recent buyers what nearly stopped them.

The first order also sets the tone for everything downstream. A confusing checkout, a surprise shipping fee or a late delivery does not only cost you that order, it costs the second one. See digital customer experience for the wider picture, and how to increase average order value for the AOV half.

Retention

Retention is where eCommerce growth actually compounds. It is also the stage most stores cannot report on, because their analytics are session-based rather than customer-based.

Supporting metrics: churn rate, average days between transactions, customer lifetime value, and RFM segments. The single most useful of these is average days between transactions, because it tells you when a customer is late rather than merely quiet, which is what makes a win-back campaign timely. Our guides on calculating repeat purchase rate, reducing churn and the KPIs that move CLV go deeper.

The qualitative half matters just as much: ask loyal customers why they stay. Their answers usually name one or two operational things (delivery, packaging, a specific product) that no dashboard would have surfaced.

Referral

Happy customers who tell other people are the cheapest acquisition channel there is. They are also the hardest stage to measure honestly, which is why it comes last here. To get more of them, run a referral program, which can sit inside your loyalty program, and reward the customer each time they recommend you to someone new, as Melanie Balke describes in her AARRR framework overview.

Trackable referrals — a referral program, a shared code, a review left on the site — are a fraction of real word of mouth. That is why Net Promoter Score is normally used alongside referral rate: it measures stated willingness to recommend, which is the leading indicator, while referral rate measures what you can actually attribute.

Which KPIs to track at each stage

The table below is the diagnostic version: what a weak stage looks like, and what to do about it first.

Where AARRR fits eCommerce badly

Being honest about this is what makes the framework usable.

• Referral is weak for most stores. Unless you sell something people naturally talk about, or you run a well-designed referral program, the referral stage will report a small number that barely moves. Measure it, but do not let it take priority over retention just because the acronym gives it equal weight.
• Activation is fuzzy without accounts. A store with guest checkout has many visitors who go from landing page to purchase with no intermediate commitment. Defining activation as add to cart keeps the stage meaningful; defining it as account creation does not.
• The stages are not sequential. A loyal customer arriving from an email re-enters at retention. A referred customer enters at acquisition and often converts in one session. Drawing AARRR as a strict funnel double-counts people and hides repeat behavior.
• Revenue is a stage, not the goal. Treating first-order revenue as the finish line is exactly the habit that produces expensive one-time buyers. Judge acquisition against lifetime value, which lives two stages later.
• Five numbers can still be vanity numbers. Traffic and add-to-cart rate look like progress. If neither moves revenue or repeat rate, the framework has just organized your vanity metrics more neatly.

AARRR vs RARRA and other reorderings

The disagreement between the two models is not about which metrics matter. It is about which stage you fix first when you can only fix one. AARRR implies you start at the top of the funnel; RARRA says start with the people you already have.

For most established stores, RARRA describes the better priority order and AARRR describes the better reporting order. Report on all five stages in the eCommerce sequence, then spend your improvement budget from the bottom up.

How to set up AARRR reporting for your store

1. Write down the five stages and one metric each

   Acquisition: CAC by channel. Activation: activation rate. Revenue: conversion rate and AOV. Retention: repeat purchase rate. Referral: referral rate. Resist adding a sixth stage or a second primary metric.
2. Agree on your activation action

   One action, defined in writing, that everyone reports on. Add to cart works for most stores. Newsletter subscription works if your buying cycle is long and email drives the second visit.
3. Connect data at the customer level

   Session-based analytics cannot answer the retention question, because it counts visits rather than people. You need order history keyed to customers before repeat purchase rate, CLV or RFM mean anything.
4. Set a baseline over a full quarter

   One quarter, all five numbers, plus the same quarter last year if you have it. Anything shorter and seasonality will decide your priorities for you.
5. Pick the weakest stage, not the most interesting one

   Compare each stage against your own trend rather than against an industry average. The stage that has moved backwards, or that you could not measure at all, is the one to work on.
6. Run one experiment and re-read the same five numbers

   One hypothesis, one stage, one measurement window. If the stage metric moves but revenue and repeat rate do not, you improved a vanity number and you should say so out loud.

Two of these steps are tooling problems. Customer-level retention data comes from Nexus by Omniconvert, which computes repeat purchase rate, cohort retention, RFM segments and CLV from your order history and pushes those segments into your ad and email tools. The experiment in step six belongs in Omniconvert Explore, which handles A/B testing, personalization and on-site surveys for the acquisition and activation stages.

If you want the wider framework rather than the metric set, the CVO Academy covers Customer Value Optimization end to end, including the retention work that the referral and retention stages depend on.

Frequently asked questions about AARRR metrics

Alexandra Panaitescu

Content Marketing Specialist

Alexandra Panaitescu is a B2B content marketing specialist with over 8 years of experience building data-driven content strategies and inbound campaigns that help businesses grow, from generating qualified leads to establishing brand authority and revenue.