Introduction

Early 2023 I left Ento after having built the company as a co-founder for four years. In short, Ento is a startup company applying machine learning to various data sources related to energy consumption in buildings with the goal of optimizing energy efficiency and minimize CO2 emissions.

I joined as a technical co-founder, right after handing in my PhD, utilizing my recent experience from the energy industry and prior experience as an IT consultant.

Having developed a curiosity about the fundamentals of running a business, it felt natural to me to build a startup from scratch. This curiosity was similar to how I got drawn to study physics based on a curiosity about the fundamental principles about the world and the universe in particular.

During the first three years I played a key role in product discovery and product development: machine learning, frontend and backend development using the scientific Python stack and Django.

As the company grew, I transitioned to build and manage all company operations from zero to post-seed. Including Finance (invoicing, accounting, budgeting, reporting), Legal (IP, compliance, customer and vendor contracts) and People (recruiting, onboarding).

We raised a $3.6 million seed round in 2021. In 2023 after having set up and streamlined all business operations I exited to join a moonshot project.

How I spent my time

I was inspired by how Sam Corcos meticulously tracked how he spent his time during the first two years as CEO of Levels. Never having tracked my time in any way, I found a simple method for doing a high-level analysis to look back on the last four years.

There are generally three categories to how I spend my time: building the product, processing emails, meetings.

In the sections below I look into my coding and email activity. I haven’t looked into meetings, but expect there’s a high correlation between time spent in meetings and number of emails as there’s usually a thread of mails related to schedule and follow up on meetings.

Building the product

Anyone who uses GitHub to version their code is familiar with the contributions calendar on their profile. It shows how many contributions were made in the form of code (commits), pull requests and issues per day for the most recent 12 months. In my case it’s close to 100% commits.

It's straightforward to get GitHub activity for previous years by simply changing the year in the url. My daily code contributions during the four years building Ento look like this:

Figure 1: 90-day rolling average of daily contributions to Github.

Emails, emails everywhere

If you're using Gmail, it's possible to extract all emails via the Gmail API. Doing so and filtering system-generated emails like chat notifications, my email activity looks like this:

Figure 2: 90-day rolling average total daily email activity (received + sent).

It seems the email activity is going exponential. This is mostly caused by increasing number incoming emails, not all of which requires a response. As the team grows the number of cc-email for information purposes increases significantly.

Note that this graph shows the sum of all emails: incoming and outgoing. The factor between incoming and outgoing emails only ever increased.

For everyone who uses Gmail, I recommend the framework by Andreas Klinger to get a grip of your inbox. Following this framework you won't miss any emails or forget about following up.

Overview

Combining the data, this is what my activity looks like:

Figure 3: 30-day rolling average of daily code contributions and emails sent (normalized).

Note that this graph shows emails sent and not total email activity, a more accurate display of my work activity.

The first year was bliss: focusing on building. No customers, no fundraising.

Leading up to the seed round in November 2021 there were several months of due diligence and negotiations.

The funding resulted in a phase transition in all aspects of the company operations. One immediate change was that we were now able to hire employees for the first time. Something I spent a lot of time on for several months following the funding round. I remember spending more than 60 hours on interviews and preparation during two weeks early in 2022.

Most other aspects of the company operations focused on communicating with various vendors, setting up systems and outsourcing contributing to the high email activity in 2022.

Our initial target market was municipalities, which have long periods of downtime during the summer holiday season. This meant we could also afford to take some time off in this period without loosing leads.

Writing this I realize it would be interesting to analyzing the correlations between the different activities. Unfortunately, the figures were created half a year ago and I no longer have the data, so that is left as an exercise for the next time I look in to my work activity.

What I learned

It’s no coincidence that people refer to being part of a startup as accelerated learning and the equivalent of doing an MBA or similar. The following are the most general things I learned during my four years of building Ento. They are presented in no particular order. These topics do not represent every important aspect of building a startup. To some this might seem insightful while obvious to others. It depends on where you are in your journey.

Mentoring

The right mentor at the right time can significantly reduce the time it takes to solve a problem. This is similar to recommending books: the right book for the right person at the right time.

It's a good investment to spend hours talking to different people before finding the right match compared to spending days struggling to figure out a solution on your own.

This is also an argument that at least a small amount of time should be spend on networking. For people like me (introverted, slightly autistic) it seems tedious and tiring and the signal-to-noise ratio is very low. But still it has resulted in a few long-running connections that have been very helpful and made the tediousness worth it.

Have an open mind to what mentoring means. It doesn’t have to be formalized in any way. It can simply be an informal conversation over a cup of tea.

Sometimes the best mentoring advice is not an answer to your question but a rephrasing of your question that enables yourself to answer the original question.

When it comes to finding a mentor, don’t be impressed by accomplishments and CVs. Sometimes the best mentor for a given problem is a seasoned CEO with decades of experience. Other times it’s a fledgling entrepreneur with just a few more months of experience than yourself.

Making time

I recommend blocking instant messages, emails and other distractions for hours at a time every day for focused work on your top priority. The Eisenhower Matrix is a great tool to help prioritizing.

Organizing - Todo list vs calendar schedule

I've seen discussions on whether to use todo lists or calendar scheduling to stay organized.

A challenge regarding todo list is the optimism about how many tasks can be done within a single day. Scheduling a realistic amount of time for a task in the calendar gives you a better overview. However, a challenge with the calendar approach is how to determine a realistic time requirement for each task. Also, is your brain in the right mode for a specific task when it comes up on your schedule?

My usual approach is to have one big todo list in a plain text file. Not using any app or framework in particular. Just a bullet list with tasks and sub tasks that gets updated on an ongoing basis. The order of the list defines the priority and I update it whenever necessary. Usually once per day. Usually remove all done tasks once a week and spend a few minutes looking back at the week that passed. A challenge of this approach is that you have to be comfortable with an endless list of tasks that never is fully completed.

I've recently changed my approach so that I now keep a daily note. A list of the most important tasks for the day and a section below for my notes and thoughts of the day. That allows me to go back in time and reflect on the past days and weeks.

A feature of this approach is that whenever something pops into my mind outside working hours, I simply spend a few seconds adding it as an item on my list for the following day, so I don’t have to remember it until then.

One technique to get used to this approach is to follow the advice on strict work hours and wind down time as proposed in Sleep Hygiene For Software Engineers by Scott Klum.

Hiring

Expanding the team from founders to employees takes a lot more effort than just hiring people. Your role and responsibility as a founder changes from the moment you hire the first employee.

You realize that keeping everyone synchronized and aligned on company and team goals is not something that happens automatically, but requires continuous effort.

Regarding culture: the initial implicit or explicit agreed way of doing things might no longer be optimal when the team grows. Experienced employees will suggest changes to or entirely new processes. Especially in case of inexperienced founders.

I'll summarize with a quote from The Making of a Manager: "A manager's job is to build a team that works well together, support members in reaching their career goals, and create processes to get work done smoothly and efficiently. ... to get better outcomes from a group of people working together.”

Learning by doing

I’ve read a large number of books and blog posts as well as taking evening classes on business administration both before and during my startup journey. All of that prepared me for what to expect according to the current canonical startup advise: customer discovery, customer obsession, lean startup and failing fast.

I recommend all of these activities, but they can only take you so far. There is no learning like learning by doing. To accelerate your learning, realize that there's no speed limit and whatever scares you, go do it.

Effort and results

In the beginning when there are few people working on the project there is a direct relation between hours worked and progress. At some point this relation starts to decouple. You can spend all your waking hours working, but the effect on progress decreases.

By then your priorities should change to set up well-defined processes and automated systems. Avoid becoming a bottleneck and enable the rest of the team to execute effectively. Delegate, delegate, delegate. "Unblocking others is your top priority" – Sam Corcos.

Think of this as transitioning from an individual contributor (or maker) to a manager regardless of whether you are managing people or systems.

Looking ahead

It was a very difficult decision to leave the Ento team and everything I’ve been building for the past four years. Having spent a great deal of time streamlining, automating and outsourcing all processes related to business operations, I'm leaving the team in a good position to take over, so it’s an appropriate time for me to leave even though there’s probably never a perfect time to do so.

I left the company because I got an exciting opportunity to join a new moonshot project. Something so exciting that I just have to do it. In my new role I'm back to building the product and engineering organization. I look forward to building on and expanding my learnings. It's time to build!