6 years building Planning Poker Online: from a weekend project to 147K users (and 3 mistakes I had to fix)

Miquel Las Heras — Mon, 11 May 2026 11:05:22 +0000

In early 2020 I built a real-time planning poker app over a weekend, almost by accident. Six years later, 147,000+ people use it every month, including teams at Google.

This is a story about Firebase, Hacker News, a €60k mistake, and what I'd do differently if I started today. Especially relevant if you're building a SaaS in 2026 and worried about LLM-era discovery.

The 2020 origin

In 2019 I was working as a frontend developer in a distributed team. At the start of every sprint we would always do the same estimation meeting: how many difficulty points each task should have, how to split them, etc. The "sprint planning meeting".

To not influence others, each team member has to say their estimation secretly, and on the count of 3 we all show our estimations and discuss the results. But… how do you do that in a video call?

Those sessions felt quite unprofessional and boring. To cast the votes we all wrote our estimations in the meeting chat at the same time. I am sure some team members cheated and waited until they saw other estimations to cast theirs xd. In a previous company we were using fingers in the camera — imagine trying to see all the fingers of a whole team and count estimation points xD.

Back in 2019, Google Firebase was becoming very popular. I was especially interested in the real-time capabilities — they had just launched Firestore and I wanted to give it a try. I thought it would be fun to build a small tool for my team so we could do proper estimations.

After shipping the first version of the planning poker app (early 2020), I was a bit shy or concerned about presenting it to my team and company, so I decided not to. Instead I posted it on Hacker News, Product Hunt, LinkedIn, etc. To my surprise the post on Hacker News gained big traction, and in 1 week I had 2,000 users!

Early decisions

A few early decisions shaped everything that followed:

Choosing the domain for SEO

While many companies pick their brand name for the domain, I decided to pick a domain name in SEO terms. I saw that "online planning poker" or "planning poker online" were highly used to find tools like mine, and planningpokeronline.com was available, so I went for it.

This had a great benefit in the early days: very quickly we ranked #1 in search results, which brought us a lot of traffic. But for the long term, it has been a problem, as we don't have a strong brand name for our app.

This has been even more painful now in AI times. LLMs are good at identifying tokens: when they find a brand name they know it is a brand and they store every occurrence in their models. We are usually mentioned as "planning poker online", which is often not related to our brand but to the technique itself.

Posting on Hacker News, 2k users in a few days

Posting on Hacker News got big attention. In just 2 weeks we went from a freshly shipped prototype with 0 users to 2k users.

Those users gave me really good feedback directly. Having a dedicated tool to run those meetings made teams feel more engaged and professional, and they were having fun. That's when I realized I was onto something.

Onboarding the designer, the professional touch

The app was useful, but you could tell it was a prototype made by an indie developer.

When Oriol, our designer, joined the project, the app became a real professional app. I remember looking at the site and thinking, OMG this is a real app. It wasn't the first app I'd launched publicly, but this one felt really mine, as I had started everything from scratch.

This is how our landing page looked in the early 20s:

Our landing page in the early 2020s.

Simple & Fun: The Real-Time Poker Table

This has been the key decision point in our success: while other apps felt like filling boring spreadsheets, we made a tool that felt like a real game.

You can see when a player joins a room because their avatar appears next to the table. When they cast their vote you see their cards turning blue, and when showing the cards, a nice flipping animation comes in revealing the results of the voting round.

We even made it possible to throw stuff at other players around the table, like you would shoot Nerf darts in an office. You can throw whatever emoji you want, or predefined objects like Nerf darts, paper balls, or paper planes.

This is a screen of our landing page from 2022 where you can see how our table looked back then — it hasn't changed very much:

Our landing page in 2022, the real-time poker table design we still use.

The no-signup decision

Making the app simple has been a top priority for us.

Create a game in 2 clicks: trying our app with the team just needs 2 clicks. The first one brings you to the create game form, the second submits the form with default configs and redirects to the game ready to play.

Developers just join and play: when they click the game link shared by the facilitator, they just see one small popup asking for a display name, and they can start estimating issues right away.

Making it easier for the facilitator has been key, but making it easy for developers has been even more important. We tried very hard to stay out of their way and let them just run their estimation meetings with ease.

This has some downsides, obviously. For example, it makes remarketing harder: as we don't collect email addresses we can't contact our churned or inactive users. But I think it has been a good decision so far.

6 years timeline

As the app became more popular, some people in my company saw on my LinkedIn that I was sharing stuff about my Planning Poker app, and they wanted to try it.

This became crucial to fine-tune the app: I was a real user of the app in every sprint. In every meeting I could see something to improve and write it down in a note.

The app started to grow massively. By April 2020 we released the premium plans. I still remember the first customer — it was amazing seeing someone willing to pay for something I built. This was the "proof of concept": we had not only a good tool but also a potential business.

The business kept growing really organically for years.

Listening to our users: The Jira Plugin

Some users asked if the app could integrate better with their Jira. So we decided to build a dedicated version of the app for Jira.

Having our app running inside Jira has great benefits for companies:

The Planning Poker table runs directly inside Jira, so there is no need to switch to another app.
Estimations sync seamlessly with Jira issues.
They don't need to set up a new vendor in their company, which usually means dealing with new data providers, new user accounts, an approval process, and so on.

We released the plugin in March 2024, and many of our web users switched to it. It was a big success. They were giving us great feedback, so it became a priority to make it the best planning poker app on the Atlassian Marketplace.

The Jira plugin today: the poker table running inside Jira.

The Retrospectives app

In February of this year (2026) we launched the final version of our Retrospectives app; both the web app and Jira plugin versions launched together. Our customers are loving it; the feedback has been really good.

A retrospective is an Agile ritual that runs after each sprint, where the team analyzes what went well and what didn't, and defines actions to improve their performance.

Our retros app follows the same principles: it feels like a game, and it is simple and fun. You can also throw stuff at teammates here! xD

We are actively working on improving this app. Don't hesitate to send feedback!

What we changed our minds about

The "big tool"

After the success of the Planning Poker app we thought we could create a full project management tool (something like our own Jira for Agile companies). We changed our minds and decided to make standalone apps for Planning Poker and Retrospectives. Here's why:

Each app keeps the maximum simplicity. Just one button to create the planning poker or retro and done.
Teams can plug and play apps easily. They can purchase the app they need for the time they need it.
Companies can keep using their existing project management tools: Jira, Asana, Linear, Trello, Todoist, etc.
It makes it easier for us to improve and simplify each app.

Paid Media

When I talked to other founders, many of them advised me: "Now that you have a working business model you need to invest money in scaling up", so that's what I did. I spent more than €60k in just one year on Google Ads only to see it didn't work. "Keep waiting, it will improve," I was told. You know what? Deleted money.

I realised that, at least for our case, we need to appear on the screens of our potential customers in the moment they are searching for the tool. So SEO was our best organic channel. Which brings me to the third thing we had to change.

Relying only on SEO

Our SEO was so good (we are the first result for "planning poker" in many search engines) that we thought it was the only thing we needed. And this worked well for many years, but now in AI times it has been a disadvantage.

To people starting now I would recommend picking the brand name for the domain and making sure LLMs know about your brand and recommend it. Although nowadays there are so many things changing that who knows which would be the best choice in the future :)

What we got right and stuck with

"Our app has to feel like a game, not like filling bureaucratic forms"

This has been a key differentiation point for us, and it has been our best feedback since day 1.

In our feedback tool, users report things like:

"team loves it and promotes conversations"
"Stays out of the way in a really nice way. Only does its job. Great work!"
"Easy to understand :) nice little robot <3"
"Feels really simple to use and can be overviewed quite nicely"

And the most loved feature, by far, is throwing stuff at colleagues xD:

"Throwing emotes to colleagues"
"I like throwing things at people"
"throwing the emojis to people is hilarious"
"it's super easy to use, team really likes the ability to 'throw' things at each other"

Our poker table today, same philosophy as day 1.

Closing reflection

I always loved coding and crafting the best stuff I could for others to use. Seeing more than 147,000 people using the apps we built feels amazing and keeps me wanting to improve them every day. (Don't worry, we won't start adding features without any sense xd.)

I am always happy to talk, so feel free to reach me on LinkedIn or try our apps: Planning Poker Online and Retrospective Online.

DEV Community: Miquel Las Heras