The Pirate Metrics Model or AARRR is a 5-step model for creating a metrics framework, created by 500 Startups founder Dave McClure in 2006/7.
We recommend the model to our clients as a means of getting them thinking about desired user behaviour and essential metrics. You can read about how we run the model here.
I caught up with Dave McClure to chat about how the pirate metrics model holds up nearly 10 years after he first proposed it…
Coming up with pirate metrics
The model was a response to something Dave had been trying to figure out during the end of his time at PayPal in 2004, SimplyHired in 2005-2006 and Mint in 2007…
“I was trying to figure out a rational framework to make product and marketing decisions for startups and it seemed very confusing” says Dave. “There’s lots of different variables and [its] really not clear how to make decisions that made sense. Looking at the different kinds of things you’re trying to figure out I came up with that 5 step model.”
Before working out the model, Dave was paying attention to entrepreneurs like LinkedIn Co-Founder Reid Hoffman. “Engagement, distribution, modernisation was a simple three step way to think about it, but I tried to incorporate a little bit more about retention and referral. Drawing pictures on napkins, I finally came up with something that seemed like it made sense to people when I talked about it.”
Does it still hold up?
I asked Dave if there was anything he’d changed about the model over the last 10 years. “I think the major change that I’ve subsequently tried to communicate to people for the last couple years is that it doesn’t necessarily happen in that order…
“It seems like for most businesses, [they should] focus on acquisition and retention first and then consider how to scale the model and/or make money (although sometimes you need to figure out how to make money right away)… Then depending on the economics and depending on your financing you might proceed directly to distribution and growth or you might work on modernisation and unit economics.”
The first to admit that there are “probably flaws in the model and are better or different ways to think about it” Dave thinks the model’s value lies in its simplicity. “For a lot of people it’s easy for them to get their heads around [and think about] what type of metrics fall into each category.”
Another thing to consider is under understanding where you are in the MVP process, or whether you feel like you’ve got traction says Dave. “I think there’s a little bit of an emphasis on building product and building features when you’re trying to figure out what people actually want to use or what people want to pay for. Once you get to some level of usage then maybe you don’t need to keep focusing on the feature set as much and you can focus a little bit more on scale and distribution and maybe retention becomes more of an important part.”
Product versus marketing
Another thing Dave has considered more in the last 5 years or so is the split between product and marketing….
“I think people tend to look at the model and think it applies to product a lot, but they don’t always think about how does it apply to marketing efforts. In general, when people think about lean startup, they think about iterating on product because that’s what they know or what they do. Sometimes, it’s actually a lot cheaper and easier to iterate on marketing. They’re definitely interrelated, but both are parts of the puzzle, it’s not just about writing code and building product, marketing is an important part of that as well.”
Choosing metrics and making them work for you
Dave usually recommends that founders choose 3-5 key metrics to focus on, but thinks even that’s probably too many. “There’s a lot of people who plug in Google Analytics or some other metrics product and think ‘okay I’m done!’ and that’s really not at all done, that’s just the beginning. Its really not just about gathering information or even a small number of metrics, it’s about what type of questions are you trying to answer with the metrics, what type of decisions you end up making in product or marketing as a result of that data gathering.”
Gathering data simply isn’t enough. You need to be thinking of the right questions to ask and using the data to answer those questions says Dave. “You still need to set up experiments in a way that test for cause and effect and that you have a good question that you’re asking that’s tied into some type of customer development process or product strategy or marketing strategy that you’re trying to address.”
- The order (acquisition, activation, retention, referral, revenue) isn’t important – most startups will focus on acquisition and retention first
- Bear in mind where you are in the MVP cycle
- Consider product AND marketing
- Don’t get bogged down in too many metrics
- Use data to answer pre-determined questions
The Pirate Metrics model is still, 10 year on a valuable tool for startup teams to start thinking about important user behaviours and metrics. It’s by no means a comprehensive “how to” of metrics, but it’s a really great starting point.
Image Credit: HeisenbergMedia
If you’d like to discuss your startup or project, get in touch with Simpleweb today.