Growth Step #1: Pick a Goal

We’re kicking off a new series on this blog. Starting from first principles, we want to outline how to grow. This is Part 1 of 5. This first post originally published on my personal blog about product and growth.

If you want to get the full series, subscribe to get future YesGraph posts here.



The very first step in growing your startup is to pick a goal.

It is surprising how often this isn’t mentioned as part of the process. When you hear people from Facebook’s growth team describe their process, they keep it simple: measure, test, deploy. This shows how deeply ingrained the growth ethic was at Facebook. Every employee knows they want the whole world on Facebook to the point that their advice skips goal setting.

To understand the importance of goal setting, just look at applications where the goal isn’t obvious for external observers. Before Twitter focused on monetization as a media company, they floundered with platform decisions along with the tension of getting regular users to publish vs consume. Is Foursquare a location API, a consumer local search utility, or a social network? This ambiguity is part of their current growing pains.

Deciding what the goal should be is hard. Startups are flooded with numbers: web traffic analytics, product event data, site speed & perf metrics, app store metrics, cohort data, and on and on. People see dashboards full of dozens of numbers, with additional complexity because you never just care about a number but also how it changes too.


Which numbers matter? Let’s explore the different aspects of a good goal.

Stage Appropriate

Goals change depending on your stage. In this post by Brian Balfour about different stages for a company, he explains how retention is the most important metrics at the start, as a measure of approaching product-market fit.

Easy To Understand

People tend to have a hard time understanding graphs and complex statistics. Your goals should become diffused throughout your team, which means ease of communication is essential. The goal should be a single number, and it should be something people understand. Active users and revenue are two that come to mind that would fit most companies, and everyone understands them. This also helps in PR if you’re going to announce milestones with your goals, like Facebook announcing one billion active users.


It is really tempting to have multiple goals. Signups and retention both matter, right? The problem is that you need a guide to help prioritize your product pipeline. Ambiguity here can cause you to work on the wrong things. It also makes communicating goals harder.

No Bullshit

It is surprising how often you continue to hear statistics that don’t matter, like pageviews. Some call these vanity metrics, because they don’t map to what matters for your business. Companies often publish the number of signups instead of the number of active users, when obviously the latter is what really matters for most businesses.


Teams within a company can certainly have their own goals. I’d recommend that they all have the same characteristics as a company goal. Ideally, the relationship between the company goal and the team goals are well understood. A good example breakdown can be found in this post by Tomasz Tunguz about structuring a sales team and the associated metrics of each component.

For my new startup, YesGraph, we went through this process. We’ve built multiple products, so have been through it more than once. Our new product helps companies grow, so the only reasonable metric is how much growth we’re causing. How many quality signups can be attributed to our tools?

There are other metrics for YesGraph, like the number of developers, data scale, and revenue. The good news is that they are all going to be proportional. So when building a new product, we’re focusing on a key metric that doubles as a quality metric.

Step #2: Build a Map

Need help? Just ask YesGraph. If you want to review your metrics, you should email us at or ask in the comments below. We obsess over picking quality metrics, and would be happy to help.



Subscribe to get future YesGraph posts here.