Case Studies > Umami

How Umami uses GameIntel to train, research and test their ideas

Discover how GameIntel helped Umami to improve not only their games, but also their team’s knowledge on hyper-casual games.

Umami

Challenge

When Umami almost tripled their workforce – adding five new members – they wanted to transform them into super hyper-casual developers. They needed to train them, teach them the industry and show them the best practices. And fast. As a hyper-casual company, they need to move quickly: running three brainstorms a week and developing around four prototypes every week. They needed a system that could help them learn, research and optimize their games.

Solution

Umami used a combination of GameIntel’s features to solve these challenges. They trained their people using Market Insights to learn about specific games and what works, Top Charts to discover trending fads for their brainstorms and Benchmarks+ to see what metrics they should be aiming for. By combining their own data with ours, they were able to get accurate, reliable information and decide which ideas would work. And all together, it helped them hit their target cost per impressions 20% of the time.

Interesting facts

About Umami

Founded in 2019, Umami Games is a Danish-based Independent Game Developer, focusing specifically on the hyper-casual market. Since day one, they’ve focused on making fast prototypes, testing within 14 days. They aim to make at least 10 hyper-casual games each month, and they leave the decision-making down to the metrics.

Just a few of Umami's highlights

ABC Runner

ABC Runner

9+ million installs
Released: Oct 2020

CarCraft.io

CarCraft.io

200+ thousand installs
Released: April 2021

My Little Baby

My Little Baby

100+ thousand installs
Released: July 2020

Hiring a new team

Umami started with a small three-person team, but were recently ready to expand and grow their business. Their strategy was to hire newer, entry-level members – new to hyper-casual, at least – and then train them up. They started by hiring four new developers and a marketing person.

“But how do we get them into the mindset fast?” Riley Anderson, CEO of Umami, said. “We wanted to teach them the user journey, but also more specific and practical techniques. How do we know which keywords to pick? What are our competitors doing? Which icon would be best?”

GameIntel allowed them to walk their employees through the process and show them what worked and what didn’t. In one example, their video game artist was able to look at what was popular, compare different app store descriptions and learn which style of icon would be best to use. Not only that, but they could look and see whether the logo they had designed was unique and distinct (or whether it just looked like everyone else’s). All this helped make sure that their tests had the best chance at success as quickly as possible.

Creating swathes of prototypes

The hyper-casual industry is extremely fast-paced. If you’re looking for a successful idea, you need to jump onto it as quickly as possible. As such, Umami creates 10 to 15 prototypes a month, ready for testing. But to develop that many games in quick succession, you need plenty of ideas.

“We brainstorm at least three times a week. So we use GameIntel a lot for research,” Riley explained. “We’ll hear from our publisher that parkour games are popular, for example. So we type parkour into GameIntel, see what’s been done and look at what’s working and what isn’t. Similarly, we might use it to get an overview of what’s trending right now. For example, if you go on GameIntel right now, about half the top 50 are fidget games.”

Using GameIntel like this helps give Umami a push in the right direction. And lets them focus on the brainstorm itself.

Creating the best game possible

It’s not only about quantity, of course. It’s about quality. At the end of the day, those prototypes need to be games that will feel satisfying to play. Otherwise, players will immediately leave.

“We keep track of a lot of different metrics: retention, playtime, cost per impression,” said Riley. “But at this stage, we’re looking to get below – say – 0.30 cents CPI on our iOS Facebook tests. By combining GameIntel with the rest of our technology, we’ve had a 20% hit rate on passing that goal so far.”

Predicting what that figure should be differs for every company. But with GameIntel, Umami was able to make sure that they set realistic targets for the sub-genres they’re working on.

And once a game passes that target, it’s time to look at different monetization options. Using GameIntel, they were able to search for competitors, check how they’re monetizing similar games and find out how much they’ll likely earn.

“We follow the 40/10/10 rule. Get 40% retention on day one, 10% on day seven and a playtime of around ten minutes,” Riley said. “If the prototype hits those targets, we can move onto the next stage and really push the monetization. We think about how much it will actually make us and whether it’ll be profitable. That’s when we really focus on Benchmarks+.”

Improving and optimizing games

Once a game passes the tests, Umami can look at optimizing it to get the most out of it. They use level funnels to break down all the levels and see how many they need and whether players drop off at any point.

“We’re getting closer and closer to the truth,” said Riley. “We don’t want to lose time on the boring stuff. Instead, we want to focus on the fun stuff – like making the games and doing the research. So GameIntel helps us scale up and optimize our games quickly.”

Using a combination of GameIntel, their own tools and our A/B Testing service, Umami can easily make sure that they’re getting the most out of their games. It helps them throughout the entire process: from teaching their people about the industry, through their research to optimizing the final game.

“Your tools just know what the market wants,” Riley concluded.

Try GameIntel for yourself

You can get access to the base version of GameIntel for free or subscribe to unlock all the extra features. Make the best game possible.

Still not convinced?
Hear it from the best