A brief timeline of events, numbers and learnings on our side-project that went viral. More important; what value did we get from it?
The hype around Pokémon Go was hard to miss in the past week. As fans ourselves,
we wanted to play after launch. However, it wasn’t available in the Dutch App
Store. Not only did we create accounts in New Zealand App Store to play Pokémon Go, Joshuamade a simple Pokémon Go alert service. Sign-up and we’ll send you an e-mail when Pokémon is available in your App Store.
As a reference, this is what we’re talking about. Simple region list, with ability to register for a notification e-mail.
Yes, it went viral.
The past week is one full of learnings and big numbers for us. We wrote up a
brief timeline of events, insights and will share numbers that show you how
crazy this thing really is. For us, this project doesn’t have a business model, so what value are we getting from it? A lot actually! Not only are we testing our own tools, we’re also growing the audience for one of our ventures in Firmhouse, Skill Nation.
Virality is a numbers game, so let’s start with the fun part.
After the first ‘viral day’ we just stopped making predictions on what the
maximum numbers would become. Every time when we hit a record, it felt like it
would be an unbeatable peak. Appears it wasn’t.
Although we did start out by simply tweeting and engaging the Pokémon Go audience, this is not how it got big. Let’s take a look at a brief timeline of events.
6th & 7th of July — Pokémon Go launches
Pokémon launches down under, one day later hits the App Stores in the United States. Joshua checks if there is any way of getting a notification when an app launches in the app store, we didn’t find one.
After discussing making something ourselves during lunch, Joshua decides to just build it. His first version ships two hours later.
A twitter account was set-up, we started engaging with people using the pokemongo hashtag. This led to some early traffic and attention.
Joshua’s actions on Twitter already let to some early traction.
Our results by the end Wednesday.
The early numbers showed that we’re on to something. Getting something picked up on Twitter normally isn’t a matter of hours.
July 11th — Let’s give Product Hunt a try
One of the go-to places to launch a product is Product Hunt. As a community interested in products, having a top position in the rankings guarantees a lot traffic and attention.
We posted the website to Product Hunt just before lunch.
At this time, for every sign-up, we still received Slack notifications. During lunch we could slowly hear the frequency of notifications go up, Product Hunt was working. We didn’t expect much to be honest, as the chances or reaching a top spot are small and in the end up to the community. We did receive enough votes to get to the top-3 fairly fast.
Sign-ups were increasing, fast. As the notification in Slack started to drive us nuts, we disconnected the integration. Joshua would regularly update the whole team on the number of sign-ups.
So, here we are, the effect we were hoping for is happening. It popped the question what we wanted out of this. Seeing the number of sign-ups going up, we will be spending around $1.000 on sending e-mails through Sendgrid.
One of our projects is a gaming platform, focussed on competitive skill gaming. Considering the overlap of target audience with Pokémon Go, we decided to promote the platform. Jokingly, we discussed a bunch of beta sign-ups would be a good result.
Our metrics for Monday, 11th of July.
Monday ended up with a total of 12.360 sign ups for a notification when Pokémon Go would be available. Skill Nation received about 30–40 beta sign-ups.
July, 12th — Extra bump because of hype and media coverage
We steadily have 400–500 concurrent visitors. Until some spike from Taiwan, concurrent visitors shot up to 750+ concurrent visitors. Our server was humming along just fine, but now chokes on the requests coming in.
Oops! Dropping requests, let’s fix this.
With the help of the rest of the team, resources were scaled using Intercity and Dokku). Within 30 minutes, and without further downtime, everything was back on track.
At 10:50 we already hit 100.000 pageviews for that day. A total of 58.573 people signed up that day.
July 13th, 14th & 15th — Surfing the hype
Over the course of four days, we break record after record. With Pokémon Go hitting mainstream news, bringing in more traffic. Thursday referral traffic overtook majority of the traffic via Product Hunt.
Every day ended with a new record for sign-ups, with July 15th being the busiest day. 101.152 signed up, which means we got at least 1 sign-up every second (1.17 to be exact).
July 16th & 17th — Pokémon Go launches in additional countries
Saturday morning marks the launch of Pokémon Go in many European countries. We needed to blast out 20.000 e-mails as fast as possible. Anticipating the big batches of e-mails we needed to send, we already setup a Sendgrid account.
But right at primetime, we got blocked by Sendgrid, with our e-mail being flagged as potential spam in their compliance filter. We discovered that 24/7 support by Sendgrid doesn’t count for compliance issues. The only thing we can do is wait for the Sendgrid compliance team to start their shift and unlock our account. Over three hours later we are finally able to send out the notifications.
We decided it would be fun to share some of numbers. This will give some context on how crazy the past week was!
Sign-ups per day, Friday 15th averaged 1 sign-up per second.
Signups per country, we almost had 1% of Norwegians sign-up for a notification!
Build up of traffic between 11th — 17th of July.
When opportunity strikes, don’t stare it in the face. We shifted gears and made sure we’re getting most out of this.
Adding the link to one of our ventures, Skill Nation
Gaming is big. Punching through the noise is hard. We’ve seen this first hand while working on first traction for Skill Nation. With that in mind, we added a link to our beta landing page to the top of the Pokémon Go regions. This lead to 37.000 visitors to Skill Nation and of 750+ beta sign-ups! We didn’t expect that many and this gives us a great number of people in our target audience.
Stress testing our own infrastructure
The past year we’ve developed a set of tools we use to run our projects and products on. Intercitypowers server configuration and changes, Formbox processed registrations. An influx of traffic and sign-ups is always a make- or break moment for technology. Seeing our own tools coping and being flexible enough to make changes fast, was super valuable. We rolled out updates without downtime, while having 1700 people live on the site.
Starting up AvaibleYet.com
Pokémon Go is not the first product to launch in stages across the world. Apple does it time and time again. Refreshing the manufacturers website to see if a product can be ordered is nobodies hobby. Could we make it easier for people to be notified when a service or product launches in their country? We think we can.
In the coming weeks we’ll be launching AvailableYet.com. Making it easier for startups to announce the availability of their product. Potentially increasing sign-ups, as users only have to sign-up once at AvailableYet.
We believe that solving simple problems well is what sets healthy businesses
apart from the average startup.
Simple problem, high ambition and low expectations
Joshua started the project in his 20% time, time that he can devote to anything he wants that contributes to the mission of Firmhouse. Sometimes this is fiddling around with new techniques, contributing to Open Source projects or building something simple.
We believe that solving simple problems well is what sets healthy businesses apart from the average startup. Our internal ventures and projects are focussed on clear problems. We do have high-ambitions for anything we start, but we manage all projects with low expectations of success. Having a high ambition serves as a drive and source of energy, meeting results with low expectations makes it easier to move on from failure. We just found a way in something didn’t work and are able to move on. Don’t confuse this with accepting mediocrety.
Shifting gears when traction comes is team effort
Obviously a success like this is super exciting to the rest of the team. What really showed was how everybody could contribute. Not only with the hard parts when something failed, but especially with bringing extra energy & excitement making it an even bigger success.
It’s the external dependencies that increase risk & failure
Almost everything on our end went well. We got to scale the server without too many issues. We operated normally as a team, took care of things that needed to be done while enjoying this project going viral. Only external dependencies failed us, mainly on the e-mailing part.
As maker of a project, it’s really hard when you can’t deliver your value, when an external service fails you. Especially when an e-mail service is designed to send big batches of emails, but blocks you when you do so.
Currently we’re still doing good numbers, 500–700 concurrent visitors. And receive about 1000 sign-ups per hour. We’ll keep the site live until Pokémon Go is available in all regions, we’ll make sure everybody gets their notification!
Thanks for reading this far! We hope you enjoyed this breakdown and story on our Pokémon Go launch tracker. It would really help if you like (using the lovely heart button below) or share the story!