5 Things We Learned While Building Landbot

Illustrator: Xelon Xlf
landbot product journey

Being responsible for a young product is like playing Age of Empires: you need to prioritize your goals and distribute your scarce resources to get the maximum benefit in the shortest time –while a bunch of enraged enemies is burning down your walls. You always have that feeling of making the wrong decision. Always. And it’s probably true.

If I went back to that initial time at Landbot, there would be a few things I would love to know to save more of a headache. Here they go, Fer from the past, 5 learnings we got while building the product.

#1 Agility is Your Competitive Advantage

In the beginning, your main asset is agility. Please, avoid making any decision that sacrifices it. If you are lucky and your product resonates in the market by chance, you have a couple of years until the big players peek over there. Agility, focus, the ability to take risks are your most precious advantages, and you must protect them at all costs.

How we did it at Landbot

The first version of our company was born in 2015 as a B2C delivery service through WhatsApp. Two years later, we launched what we currently know about Landbot, with tons of continuous iterations in-between, blind steps, and experiments like a B2B2C model, a kind of helpdesk, or an NLP-based chatbot.

Just a month after writing the first line of code for Landbot, when we almost ran out of cash, we launched into Product Hunt. That night was one of the scariest of my career. But finally, amid hundreds of bugs and crashes, Landbot got more upvotes than we could dream of, becoming Product of the Day. Going all this way was just possible because of the incredible team we have.


Things that Worked

  • Many speed problems come from the inability to give up secondary benefits that end up taking us away from our focus. This framework will help you a lot to make decisions based on “A single decisive reason”.
  • Alignment with your team is essential. Although it seems obvious, it is not that simple. Spend time double-checking alignment over goals or direction.
  • And regarding changes in direction: they will happen (because you don’t know what you don’t know), and it won’t be easy for everyone. Throwing work into the trash, taking risks, or questioning entrenched ideas is not easy to digest. Prepare yourself and your team for this situation from the very beginning and root your relationship in trust.

#2 Find Out Who You’re, and Do It on Purpose

One of our feedback meetings, in a cold 2018 winter.

If something worked (or didn’t), face it: you have no clue about it. In that first stage, the data usually becomes a hieroglyph or a trap, and the need to justify numbers to stakeholders can become a dangerous end instead of a means. On the contrary, a practical and qualitative approach to data can make a difference in knowledge through the first years.

How We Did it at Landbot

One thing that has always worked particularly well at Landbot is that regardless of our data assets, we’ve always tried to start our analysis with plain and simple questions.

Who is our client? Why do they use Landbot? Who finds more value? Answering now, with far more resources and a first-class brand-new Data team, is much easier. But five years ago, with hardly any technicians or experience, we needed a few doses of creativity and persistence.

Things that Worked

  • Automated user testing. Mechanical Turk is an interesting (but quite creepy) concept that allows low-cost manual tasks to be distributed to thousands of people in a sort of human-driven API. We used it to send testing instructions and questions to those testers. It allowed us to perform massive fast tests without a large user base. Today there are more specific tools like UserTesting that work extremely well.
  • Fake automated onboarding. For some time, we took advantage of Landbot’s ability to create human-powered registration chatbots and intervene during our clients’ onboarding process to collect real-time insights.
  • Product and subscription analytics. Mixpanel, Amplitude, Chartmogul, and others are really great tools that allow anyone on our team to solve almost any product question. Even complex topics like correlation, retention, and experimentation can be easily solved with these amazing tools.
  • Feedback river. Another fav classic. If someone in the company acquires a user insight, this information must flow through the different departments. Since 2016, every two weeks the Landbot product team meets to exchange views on the insights of the feedback river, and many of our most differential features (such as Bricks) come from there.

User insights helped us turning our old builder (left) into the new one (right)

#3 Make it Fun

We live in the attention economy. Once we get a new user, we know that there are only a few seconds to capture her attention. And lately, Gen Z got enrolled into work, so things are becoming harder. Only a truly memorable product, which delivers tangible value clearly and conveniently, will be able to win the user’s attention.

Traditional activation no longer works and now is replaced with concepts like Habit, Aha, or Setup moments. That new habit-based activation only will happen if the product generates satisfaction, motivation, and cadence enough, as it happens in video games.

How We Did it at Landbot

Several parts of the product were designed with “fun” in mind: from the builder to the user onboarding flow or the chatbot’s interface. But I want to zoom in an interesting product marketing case that happened last year.

During the 2020 lockdown, I personally built a full no-code escape room using Landbot. It was a team-building initiative to help in those difficult days; a fun and unpretentious game to spend an hour fighting a mad AI with the help of your teammates. But the result was exciting and the team suggested extending it to our users. How? As a showcase for the new Landbot version about to launch.

Things that Worked

  • Basically, such an out-of-the-box launching strategy was insanely successful. Thousands of players spent almost 20 minutes in the game (vs. 2 avg. minutes in landing pages), experimenting with what Landbot is capable of. The difference in conversion was not that big, but every lead that came from the game showed far better engagement.
  • More important: we learned the importance of showing vs. telling.
  • The game involved several viral loops with real-life participation within Twitter and Product Hunt. This helped by pushing traffic without a budget.
  • The whole launch was a source of joy and unity for our own team.

#4 Product Vision

You’ve probably heard hundreds of times about the importance of product vision and roadmapping. You know the theory, you’ve read the books, but I’m sorry to tell you that reality will be far different. Let me explain what I’ve learned over the years by being responsible for setting up the product vision and roadmaps at Landbot.

Roadmaps Are Not About Time

Your favourite quote will be from a young but clever product manager:

This means in rough that you must stop wasting time on hard-to-maintain timelines. Instead of this, win focus with a Notion document where you throw light on priorities and reasons.

Roadmaps Are About Alignment

A roadmap is a tool aimed to win alignment with your team: no more. If you over-optimize it you are taking over your team’s role in the solution space. If you don’t manage to solve your team’s main concerns, you are failing. Your role here is to get all the information from them and agree on departure points and directions. Roadmaps must explain why we are here and why we want to dive there.

Our Approach to Strategy

One thing I really like about Landbot is how we unlock complex problems like in this case, product strategy. We don’t reinvent the wheel nor have a brilliantly innovative approach, but, hey, things do work.

Basically, we have five strategic pillars that drive our product direction. Everyone at the company is aware of them and could answer why they are so important and what we are doing there. Both the teams we have and their OKRs are shaped around those pillars so, in the end, it is easy for us to check its status, constraints, or relevancy.

What those pillars are about may be the subject for a future post...

#5 Culture Is Everything

One of the biggest challenges for startup teams is the transitioning stages. One day your company is small enough to need you as a swiss-army-knife, in charge of design, product management, front-end development, branding, sales, or product marketing. And the other day your company starts growing, your team multiplies by 4 and you haven't yet had time to even understand your new role and processes.

Keeping the right team culture in such a terrifying scenario sometimes seems hard, but it pays off in the end. Not doing it means far more nightmares in the future.

How We Did it at Landbot

Even at the very beginning, we cared a lot about people and culture. One person on the team particularly contributed to it: Esther Valiente. She, as responsible for HR, did an incredible job by picking the right people with the best culture, even when we had no big budget in our pocket. Today those people who joined us in that phase (Vicente Arinyo, Osama Nehme, Edu Ruiz) remain as a perfect sample of what makes Landbot unique: our team play.

Precisely those three persons were the ones that saved that first Product Hunt launch night. It was 12 am and the launch was scheduled for 08 am and we had several critical product issues. They took ownership of the situation and managed to solve the problem. The next morning we were live in Product Hunt, they still were in front of their screens –with no sleep– and the problems had just disappeared.

The product wasn’t perfect, for sure, but we got more than 2K upvotes from the community and it was thanks to those guys.

Things that Worked

Shortly after hiring our VP of Engineering, we started implementing several interesting changes to the product organization. Those changes are based on three basic people principles that I learned from him and right now forge my creed:

  • Trust. A feedback loop that flows in both directions (team > manager > team) with communication and candor as a ground.
  • Accountability. Decisions will be owned by the teams. The leader will be responsible for guiding, training, and removing bottlenecks.
  • Motivation, as the opposite of control and micromanaging, motivation happens once the person feels control over herself. Nothing bursts motivation more than that lack of control.

Last Word

I must say that one of my favorite questions when interviewing candidates is: What do you know now that you would have liked to know when you started working as [insert position]?”

As a frustrated philologist, I like to see this question as a talk with myself from a few years ago, where a mature Fer bore the younger Fer with lessons and regrets. I suppose it is the law of life: certain things must be learned through experience. Young people need to fail.

And it may sound curious, but I realize now that there’s one more important learning that has been here for decades. Something I learned playing role-playing games as a child is that the best Game Master is not the one who spends most time studying the rules or writing paper-based plans because the game always evolves in an unexpected way. The best Game Master learns by doing and listening to the players, and this is how he creates true value for them in the form of unique stories—the same way as we do in products. If just a piece of Landbot aims to achieve that same goal, I guess this whole journey will have been worthy enough.