I am always on the lookout for good digital frameworks, especially when it comes to metrics, as they really help simplify the complexity of our modern business environment. Complexity can lead to infinite choice, which often leads to paralysis.
Dave McClure’s AARRR framework (or Startup Metrics for Pirates) is an excellent antidote to this situation. The following essay outlines it in detail.
This framework was originally created for startups. A startup is essentially a search for a repeatable and scalable business model, so testing a series of hypotheses about the various parts of the business. It needs to adapt over time, and tell if the business model is worth scaling into a company.
Established organisations already have a repeatable and profitable business model. Any business school will tell you that the numbers to track are Income Statements, Balance Sheets and Cash Flow Statements, however more and more as platforms are becoming the backbone of these organisations, adopting the startup mentality to product is increasingly worth paying attention to.
An example would be a large organisation switching to offer their product through a managed e-commerce website. Traditional activity switches from broad scale mass marketing to targeted and measurable activity on the platform, with new tactics required to evolve to shifting customer needs.
To this end, I’ll phrase this essay to be inclusive of both startups and established businesses who are managing a product - as a definition product will refer to any sort of digital platform (e.g e-commerce website, widget or service or sharing based platform).
AARRR: 5 Steps to Success
The AARRR Framework is broken down into five simple categories to measure the customer lifecycle.
Each section is broken out in detail below.
Users come to the product from various channels.
In the current landscape, there are increasingly more and more channels available to acquire customers. Existing methods are often becoming increasingly programatic, and new and “shiny” options emerge almost daily. Again, choice paralysis.
Our goal is to refine this down into a few channels that work the best.
John Wanamaker, billed as the father of advertising once said “Half the money I spend on advertising is wasted: the trouble is I don’t know which half”. Right now we have the tools in digital to find the right half! Just look at Google. A huge part of their revenue stream and success has been letting people figure out if their advertising actually works by calculating direct ROI.
There are three simple ways to uncover channels that work for you:
- What is the highest volume channel (#)?
- What is the best performing channels (%)?
- What is the lowest cost channel ($)?
With this in mind we can test them out, find the sweet spots and focus on and optimise these channels going forward.
As Peter Thiel has been quoted as saying, founders often fail because they don’t understand distribution and acquisition. Since they haven’t thought about what works, they try some sales, advertising and viral marketing - everything but the kitchen sink. This is a bad idea, because often it is likely that only one channel is optimal. Poor acquisition - not product - is often the number on e cause of failure.
The users land on our product. Our goal is to ensure they enjoy their first visit and have a “happy” user experience.
Take a deep dive and try and define what this happy experience will equate to, and you should get a pretty good idea of what to measure here. This could be time on site, page views, clicks etc.
Make sure you A/B test a lot of landing pages and content to achieve thew optimal experience.
Our goal is to make the user come back and visits the product multiple times.
Retention activity on a digital product usually includes email or blog/RSS on housed content.
Easiest way to not suck at this is to automate emails - 3, 7, and 30 days (or the individual lifecycle of the product), triggered by activity (or lack of) and events.
Metrics should be pretty simple to come up with.
Our goal is for the user to like the product enough to refer others. This should largely be natural and organic.
This next bit is really important. Don’t do any sort of viral marketing until your product doesn’t suck! Because then people will tell other people that your product sucks. Positive experience matters.
Do qualitative testing on your product until you get really high numbers, and then you can start thinking about testing viral ideas.
Users conduct some monetising behaviour. Either it is your established organisations way of doing business, or as a startup you have come up with a monetisation idea.
Defining Measurement Metrics
What to measure falls into four categories:
Traffic Analysis / User Engagement. Track what users do, usage and conversion % from an empirical sample of users. Look for best performing, largest volume, and lowest cost.
Usability Testing / Session Monitoring. Watch what users do, and figure out any problems or barriers to solve.
A/B Multivariate Testing. Figure out what users do in one scenario versus another and pick what is better, or what copy / graphics / User Interface (UI) is most effective in any given scenario.
Monitoring and Tracking competitors against what you offer. Hopefully things like demographic targeting, keyword traffic, user satisfaction, and marketing channels.
Find a tool that can help you measure each, and think about investing in them for analysis.
With this framework applied, you can create a metrics dashboard applying AARRR to measure results moving forward. An example is provided below for a fictitious company.
Role Specific Questions
The AARRR framework should be approached across the organisation, so individual business areas should have input and be thinking about specifics.
Founder / CEO / Strategist / Manager
Analytics and data can be overwhelming if not specific to the needs of the business. The goal then is to be the head editor, and ensure you are approaching things with a less is more mindset.
Hypothesise the customer lifecycle and refine it for the business.
Try and choose 5 - 10 conversion steps - but ensure you iterate (we need to be constantly evolving and moving with forward momentum). Guess what you think people are going to do and measure against that.
Focus on conversion improvement. This cannot be overstated!
Delegate each metric to someone to own and empower them to improve it.
Marketing / Advertising
Your goal is to find out what activity works best for your business - think about designing and testing multiple marketing channels.
Select for the trifecta - High Volume, High Conversion, Low Cost. Once you find it, scale up and optimise.
Make sure you are measuring deeper down the conversion channel, not just the website / landing page. This should be an end-to-end analysis. Segment and select your channels and customers by conversion at the lowest possible level (ideally after purchase).
Again, monitor and test, but always be doing what drives the best results.
Product Managers / Engineers
Again, less is more! Don’t just build lot’s of features, think about optimising what you already have for the best results.
- Spend 80% of your time optimising existing features.
- Spend 20% of your time developing new features.
Create hypotheses, then A/B test the hell out of them. Measure any improvements, and apply what works. Rinse and Repeat.
There you have it, Dave McClure’s AARRR “pirate” framework, a great solution for both startups and established businesses. Apply this to your business for a deeper understanding of your customers and to make more profit down the road.
This post continues my series on Mental Models.