Little Data, Big Results

I remember a few years ago sitting at a table with 15+ managers and executives at our yearly summit, discussing what our company goals were going to be for the next one, three, and five years.

“We want to be the number one player in our space.”

“Ok, how do we define ‘number one’?” I asked.

What followed was a three hour debate. We started with the goal of establishing metrics, measurements, and milestones to help us reach such a lofty goal. But the conversation soon devolved into gut-checks, guesses, and opinions that, in the end, left us without any clarity or definition.

It was disheartening, to say the least. And from talking to my peers, it’s a common scenario that teams face all the time. It’s most likely that you’ve experienced this first hand — big, nebulous goals with a lack of transparency and understanding on how to reach them.

Which is perhaps why Big Data is so popular these days. Big Data suggests you track anything and everything with the promise that one day — when you finally define what “being number one” means, for example — all the data will be there for you to sift through to see if you hit your goal.

Of course, you won’t be the one doing the sifting. You’ll need to hire a team of data scientists to come in to read the tea leaves, partly because they are the only ones trained on how to use the complicated tool you put in place.

And your team most likely won’t be making effective data-driven decisions along the way. Big Data tools rarely trickle down to the team — they don’t have permissions to access the data or it’s too expensive to sign them up individually, so they must rely on top-down reports to guide them.

If this sounds unproductive and time consuming and resource-intensive to you, you’re right.

Which is why we focus on Little Data.

Little data doesn’t require data warehouses connecting all your systems. Little data doesn’t involve large scale processing or data scientists focused on sophisticated statistical modeling. Little data isn’t about special permissions for creating reports, security features limiting access, or advanced training to operate the system.

Little data is about the team getting back to the basics and focusing on actions that make your business successful.

Little data builds on the notion popularized by Paul Graham at Y Combinator: focus on a single metric to define your success. In the case of YC companies, Graham always pushes for growth.

“Focusing on hitting a growth rate reduces the otherwise bewilderingly multifarious problem of starting a startup to a single problem. You can use that target growth rate to make all your decisions for you; anything that gets you the growth you need is ipso facto right.”
- Paul Graham, Start Up = Growth

If you expand this idea from its core, you can start to see how tracking just a little bit can turn into big results. You don’t need to track everything, rather focus on a few key performance metrics that define your goals.

But of course, just like the round-table discussion I had years ago, the hardest part teams often face is defining those goals and metrics.

Listed below are simple guidelines to follow to help your team become an effective data-driven team.

  1. Start with an objective
    Take Graham’s advice to heart and find the one objective that aligns with your team. It might be growth, or it might be customer acquisition, activation, retention, or churn. If you are creating objectives for a smaller team, perhaps your goals are focused on improving the speed, quality, or accuracy of your deliverables. The most important factor is making sure the entire team is in agreement.
  2. Engage the entire team
    Top-down goal setting alone and a lack of transparency can be a hindrance to your team’s output. So explore ways to meet or exceed your goals as a team. This is a great way to both surface new ideas, share perspectives, and get buy-in from everyone involved.
  3. Set a specific goal
    If your object is to improve product quality, “deliver more and better features” is not a good goal to set. “Reduce escaped bugs by 20%” is better because it is measurable. A defined goal gives your team something tangible to track and drive to. Don’t be afraid to make adjustments as you go either. If the target is easily met, set it higher for the next quarter. If it appears to out-of-reach, break it down as a team and reset expectations.
  4. Increase team transparency
    There are a few ways to get your team rallied around the metrics you are tracking. First, you can make them responsible for reporting it. Assigning ownership helps keep the data front and center to your team and will influence their decisions. The second thing you can do is be open with the data. If the team is responsible for hitting your targets, they should know how they are doing along the way. Share data with team dashboards or reports that everyone has access to.
  5. Encourage evidence-based decisions
    If your team is tracking but then not using data to inform your decisions, then you are all bark and no bite. Lead by example and backup your decisions with data. Help your team become aware of how their actions and decisions impact key results, and they will start offering recommendations for improvement.
  6. Avoid vanity metrics
    Vanity metrics are easily manipulated and can give false hope that you are doing better than you really are. Total Downloads or Page Views can be manufactured through an increase in ad spend. But that won’t necessarily increase your Weekly or Monthly Active User rates. And active users equals returning users, which equals paying customers.

If this approach sounds productive and manageable and actionable and to you, then you are right again. It’s an approach we can verify first hand.

Looking back at my story at the top of the post, some teams — confused by the outcome at the summit meeting — decided to take this little data approach to measure their metrics. Quarter to quarter, those teams saw increased quality and accuracy in their output and could point to where their work was contributing to the overall health and success of the company.

Our only failure with these teams came from our inability to provide visibility across the entire organization and align everyone with the same goals — a hard task to achieve when you are tracking in spreadsheets and power points and big screen TVs full of vanity metrics.

We built Notion as a direct outcome of these shortcomings. And we’ve already seen teams apply the little data techniques detailed above while using Notion to bring their teams together to capture the story behind their data, and ultimately drive better decisions.

We can’t wait to see what your team will use it for.

Kevin Steigerwald is the co-founder of Notion, focused on helping good teams build, measure, and learn with data.

If you liked this article, click the ♥︎ to help promote this piece to others.

Have tips or suggestions on how to build a data driven team? We’d love to hear them. Comment below or reach out on Twitter.