Data solutions should not require code

Data Solutions: 3 Important Lessons I’ve Learned While Working at Startups

Over the past 10 years, the volume and availability of business data and data solutions has exploded. This has largely been driven by SaaS platform adoption and the democratization of data beyond engineers. Where data used to be locked away in databases—available only to those with credentials and SQL skills—it’s now available in every tool your team uses. 

I’ve been in leadership positions at fast-moving startups during the course of this evolution and watched as teams have worked to adapt. Understanding and leveraging the data your business creates gives you a leg up on your competition. As a result, every team I’ve been on has worked diligently to take advantage of their data.

Here’s what I’ve learned. 

My Three Data Solutions Takeaways

1. Build data solutions for your team, not yourself.

At Quantopian, we used Python in Jupyter notebooks to analyze data. Everyone in the company knew and used Python—every day. I loved how Jupyter enabled me to easily access data in multiple databases and create repeatable reports. When I left Quantopian to join Owl Labs, I took this knowledge with me and set up a Jupyter reporting server.

This worked fine until our team started to grow. After a few years, I was the only person on our 30-person business team that knew Python. By building a reporting platform with a significant barrier to entry for everyone on the team, I effectively became the data analyst for our company. 

There were two major problems with this:

  1. My team had to come to me to get the data they needed, and this slowed their ability to look for insights and make decisions
  2. I actually had another job to do! I couldn’t spend all day building and running reports for people.

I quickly learned that by implementing a data analytics solution others couldn’t use, I was slowing down everyone on the team. 

With so many different data solutions your team can use to access data across systems and databases, make sure you have a clear understanding of the skills your team has—and the skills you expect to hire—and implement a tool that will be useful to as many people on the team as possible.

2. Avoid data barriers.

My experience at Owl also taught me the importance of not restricting people’s access to data. Data discovery is an iterative process, a flow. You come up with a question, look for the data to try and answer it, and typically come up with 3 more questions. Often, the data doesn’t look exactly how you expected, and you need to keep digging. With each new iteration, you have to go back and get more or different data. 

When my team at Owl had to ask me to get data for them, I hindered their discovery process because I couldn’t respond to their queries fast enough. And I had been in their shoes: getting into a creative discovery flow and having to stop because I ran out of a key ingredient – data – and it killed my motivation and forward momentum. Ugh. 

In my experience, it’s key that your team has easy access to business data beyond what’s available in the tools they use every day. The magic often happens when data from different parts of the business come together under a fresh set of eyes. By making it simple for your team to get into the flow of independently asking and answering questions, you empower them to get the best results. 

3. Don’t expect everyone to learn to code.

Earlier in my career, I was a product manager at HubSpot. One of HubSpot’s key assets was terrific data from our customers about how inbound marketing could improve business results. Our marketing team wanted to leverage this data to help others understand the benefits of blogging, landing pages, and effective nurturing. The problem was, the data was locked away in our customer databases and the engineering team didn’t have time to pull reports for the marketing team. 

Our solution was to give the marketing team read-only access to a copy of the database and then teach them to write SQL to get the information they needed. Ridiculous? For 2021, yes, but this was 2011. Our options were limited.

Fast forward to Owl Labs in 2019 where I had a Python reporting server and a tenacious Head of Customer Success who wanted to understand our customers intimately so that she could provide the best experience. Did I force her to learn Python? Of course not. Instead, we invested in a business intelligence tool and started the process of unlocking our data. 

It’s not realistic to expect everyone to learn to program in order to do their job. In fact, I would say in many cases it’s detrimental. What was required in 2011—when the tools didn’t exist—isn’t the solution today. It’s imperative that you invest in no-code tools for your team to be able to access the data of your business and that those tools be broadly accessible. 

Don’t Give Up on Data for Everyone

As the data within your business continues to grow, and your headcount does as well, democratizing data access for your team might seem daunting. Thankfully, there are companies building data solutions enabling you to do just that. I’ve been fortunate to work at fast-growing companies where I can learn from my mistakes and then apply those learnings in new ways.

Hopefully, some of what I’ve shared here will help you and your team make sure that everyone has access to the tools and data they need. 

Want to learn more about breaking down the silos of data access and insights?

Share this post

Share on facebook
Share on google
Share on twitter
Share on linkedin
Share on pinterest
Share on print
Share on email

Kloudio empowers you to do more with your data

Recent Posts

elt vs etl

ETL vs ELT: Which Is Better? [+ 5 Pros and Cons]

The main difference between ETL vs ELT is the process order and transformation stage, as well as the time it takes to perform both of these aspects. It’s up to you to decide what process is best for your data needs. This guide will help.

Gut Feel vs. Data Driven Decisions: Which Is Better?

“Trust your gut" is the trademark advice of mentors and grandmothers alike, and it remains a trusted mantra for many successful entrepreneurs. When making data driven decisions is preferred, is trusting your gut still a good idea?
All articles loaded
No more articles to load
Scroll to Top