Shaping Focus with Amplitude’s North Star Framework


I recently watched Moneyball. It’s an 11-year-old movie about how the Oakland Athletics changed the way to structure winning baseball teams. In one scene, the A’s are fretting about how to pay good players with a shoestring budget. Jonah Hill’s character, Peter Brand, sums up the movie’s premise like this: “Your goal shouldn’t be to buy players. Your goal should be to buy wins. In order to buy wins, you need to buy runs.”  

This leads Brad Pitt’s character, Billy Beane, to define the team’s guiding North Star metric; on-base percentage: “You get on base, we win. You don’t, we lose.”  

A North Star metric is your organization’s big bet. 

A North Star metric is a succinct way to describe your strategy. It’s the thing to watch to tell if you’re going to win. John Cutler, Product Evangelist and Coach at Amplitude, describes the North Star as your Level 0 bet. It’s the most fundamental bet your organization makes.  

For the A’s, the big bet is, “If we maximize the probability that we have someone on base when we get a hit, we’ll score more runs than the other team.” 

Notice that when the A’s picked on-base percentage, they also implied some other stats weren’t as important. The A’s are not going to try to maximize the number of home runs they hit per game. They are not going to maximize steals to get into scoring position sooner. While steals and home runs are viable strategies in baseball, they are not the A’s strategy.  

If we try to play like the Yankees in [the front office], we’ll lose to the Yankees [on the field] – Billy Beane 

Product strategy isn’t baseball: how to win the game you’re playing 

Most of us are not playing baseball. Since baseball has a set of well-defined rules and variables, shouldn’t this fall apart when you apply it to a market, where there are so many more unknowns? As it turns out, Amplitude has studied a lot of businesses and has grouped all products and services into one of three games. 

  • Attention Game – You win when customers spend time in your product.  
  • Transaction Game – You win when customers making transactions in your product. 
  • Productivity Game – You win when customers get their work done efficiently and effectively using your product.  

Most of the stats are already well-defined for these games in one form or another. The trick isn’t to identify novel things to measure. The trick is to identify which measures predict the success of your product strategy. (Don’t try to beat the Yankees by doing exactly what the Yankees do.) 

Influence your North Star by influencing its inputs 

With our North Star defined, we move to Level 1 bets – Inputs. Inputs are the ingredients that make up your North Star, or the elements of a formula, as Amplitude’s documentation describes it. When you change an input for the better, you see an improvement in your North Star.  

For Oakland’s North Star of on-base percentage, there are 3 ways to get on base: a hit, a walk, or getting hit by a pitch. Those three ways to get on base are the three ingredients in the formula. To increase on-base percentage, the A’s players need to increase hits, walks, or the number of times they get hit by a pitch.  

But again, product strategy isn’t baseball. Let’s take WeTransfer, a file-sharing company, as an example. Right away, you should notice two things. The first is file-share services is a crowded market. Dropbox, Box, Google, Microsoft, and Adobe are all in this market space. The second is file-sharing is a productivity game. You win when someone can efficiently and effectively share their files with another person.  

WeTransfer’s core product provides a way for creatives (especially advertising agencies) to share large files with clients simply. In other file-sharing software, creative agencies often struggle to separate client accounts from employee accounts, which require per-user fees. Real-time collaboration features often go unused, as creative work is done in specialized software and only the result shared with clients. WeTransfer focuses on lightweight sharing, avoiding these pains and delivering simplicity. 

The inputs to realize WeTransfer’s product strategy might be the following: 

  • Shares to Different Domains – Customers are sharing large files (> 25 MB) with email addresses on at least 2 different top-level domains, implying multiple client relationships, a characteristic of creative agencies. 
  • Completed Receptions – Receiving parties are completing the download of a shared file within 24 hours, signaling easy reception and completed transfer.  
  • Weekly Prior Sharers – Number of customers who shared a file in the last week that have shared at least 2 files before, indicating higher likelihood to convert to paid features like organizing prior shares or sets of files.  

This might produce a North Star metric for WeTransfer titled “Weekly Completed Receptions,” or how many times this week someone downloaded a file shared from a different email domain within 24 hours. If this North Star metric improves week-to-week, WeTransfer tends to see better revenue. It is a leading indicator of productivity game wins, specific to WeTransfer’s market position and strategy.  

Prioritize the input(s) by looking at “appetite.” 

Now that we know the inputs that are driving our North Star, we’re able to make smaller, faster bets to improve them.  

Product teams thrive on these Level 2 bets, as Cutler calls them. New possibilities arise quickly, as you crowd-source opportunities to improve your Level 1 inputs across your teams. The difficult part becomes prioritizing which bets to work on first.  

To help with this, I like how Ryan Singer of Basecamp thinks about it in his book, Shape Up. Decide what to bet on next based on your appetite, or how much your organization wants to spend on it. For Basecamp, their standard bet appetite is 6 weeks. Amplitude’s documentation suggests a more flexible timeframe, anywhere from 1 to 3 months.  

In the WeTransfer example, one of their level 2 bets might be the following: We think that sharing hard-to-guess links is enough security for most of our customers, allowing us to avoid accounts for every client. This bet will primarily impact our Completed Receptions input, since we’re aiming to minimize friction to receive a large file. It might also impact our Shares to Different Domains metric. 

This bet pays off for WeTransfer because of their target audience. Creative agencies are often producing things designed for the public eye, especially in advertising. WeTransfer can deliver simplicity by avoiding mixed employee and client permissions management causing pain for advertising agencies in other platforms. Divorced from their strategy, it might be difficult to explain how hard-to-guess links is both minimum and viable.  

In Oakland’s case, their competitive analysis showed they couldn’t outbid other teams for free agent players who got a ton of hits, one of their North Star inputs. But they recognized that they could easily afford players who walked a lot or got hit by a pitch, so they prioritized bets that would increase walks.  

“I believe that there is a championship team of 25 people that we can afford because everyone else in baseball undervalues them.” – Peter Brand 

One of their high priority Level 2 bets became “With our budget, I bet we can sign and start players with mediocre hits per game but high walk rate, because the rest of the league doesn’t value walks as highly as we do. We think these players will improve our walks input without compromising our hits input, improving our overall on-base percentage” 

Notice how there is still time to make an adjustment before deciding to pivot the bet. Once these new players play a few weeks of games, Oakland can look at the stats to figure out if the bet is paying off. If things don’t look quite right, they have room for a few Level 3 bets, what I call adjustment bets, before making a larger pivot.  

Find the habits, activities, and features that correlate to improved inputs 

Level 3 bets are the fastest feedback loop in your product strategy. These are the activities and habits that correlate to improvements in Level 2 bets. For some digital products, this might be features developed on a 2-week timeline. The important thing is to revisit the adjustments to see if they’re really improving an input.  

For the A’s, some level 3 bets might be coaching players to crowd the plate to walk more often or avoid swinging unless the pitch is precisely where they hit the ball well. It’s easy to see if these adjustments are having an effect. Over the last two weeks, when a player crowds the plate, do they walk more?  

For WeTransfer, a level 3 bet might be that we need to hide the option of password-protecting a transfer initially, to see what percentage of visitors seek out that feature. If it is a sensitive creative asset, we’re also betting the customer will pay a small fee to password-protect the share. We should see an increase in Shares to Different Domains with a negligible difference in Completed Downloads After Share. That gives confidence we captured more of our customers with sensitive projects, without causing too much friction to prevent their recipients from opening the file. 

Notice how we’re always learning something from these features, even if they don’t end up correlating to improvement. If a we have a feature that isn’t having any impact on our North Star inputs, we can safely remove it from the product, allowing us to focus on the features that matter and avoid zombie initiatives, as Tom Kerwin calls them.   

Don’t wait on lagging metrics 

Notice how none of the measures above focus on revenue, despite revenue being the true measure of winning in business. The reason is simple: at each checkpoint where you measure revenue, you have very little influence over your performance during that time period. Amplitude’s documentation highlights the need for leading North Stars, predicting success rather than observing it. It’s a tough balance but for a successful product strategy, look for ways to get from your North Star (Level 0 bet) to your smaller Level 3 bets that allow you to measure success early, before it is no longer realistic to pivot or adjust.  

Need help connecting your strategy to your digital efforts? Give us a call. 

A lot of times, our clients can articulate their strategies really well in words, but their digital products don’t quite match their approach. UDig can help by working together to define leading inputs that match your strategy, and then help set up your digital products to drive improvement in those inputs. If you’d like to talk more about how UDig can help your team, don’t hesitate to contact us. 

About The Author

Hugh Dillon is a Principal Consultant, focused on product strategy and design.