Color Water Sort (CWS) is a classic water puzzle game. The goal is to sort tubes filled with multiple colours of liquid into single colours, like this:

The core loop is excellent. Highly intuitive, with the right amount of challenge to feel satisfying when you successfully solve the puzzle. There isn’t a lot beyond that, however. The meta is very light and revenue comes almost exclusively from ads, meaning retention and monetisation are both below where TapNation would like.

These are classic hyper-casual problems, which is why this experiment is so interesting; it’s a case study in taking a game from hyper-casual to hybrid-casual.

Goals, goals, goals

When we started talking TapNation set out their goals:

Increase LTV: by either increasing playtime & retention and/ or increasing the number of rewarded ads seen per user.
Increase the amount of IAP and purchaser rate: by testing/ proposing new/ different IAPs opportunities.

Given this, we decided to start with player goals.

Experimentation has consistently shown that providing players with clear goals they care about increases engagement. They add a sense of purpose and accomplishment that can turn an aimless experience into something fun and rewarding.

If the idea was correct we’d be likely to see sessions, session length and retention increase. In turn, this would increase LTV, revenue being driven by ads.

The other benefit of starting with goals was that CWS already has three under-optimised goals peppered into the experience:

Earn a mystery reward by completing 10 levels.
Earn 500 coins by collecting 50 stars.
Open three chests by collecting three keys.

The experiment

The experiment was pretty simple. We gathered each goal and their associated rewards onto a screen that would be shown between each level attempt. This would amp up the current design where only the progression goal was regularly shown to players and rewards were only mentioned on the Stars Chest goal.

By optimising the UX we set out to understand how far the current setup could be stretched. If we found a significant uplift, we could iterate on goals and vanity rewards to grow LTV. If not, we could mark that “not engaging” and consider other features.

TapNation did an amazing job bringing this together to:

Call out each goal, e.g. Collect 50 stars
Highlight the reward, e.g. the chest containing 500 coins
Show progress, e.g. the animated progress bar showing 2 out of 5 levels have been completed

The results

LTV increased. Very slightly, but consistently. Retention was flat, however given TapNation finds that adding a new screen to the UX decreases retention, this may be taken as a small positive.

The driver of LTV was increases in sessions and session length, which led to a small increase in the amount of ads watched.

However, let’s not kid ourselves, the uplifts were small. If we want to grow the game this doesn’t look like the place to look.

Looking more deeply at the data suggests a reason we saw such a tiny effect; very few players made use of the rewards they earned:

Might it be that players don’t value the rewards they earn here? Or maybe they don’t know how to claim them? If players don’t find the goals motivating, is gameplay the heartbeat we’re looking for?

We decided to look at other, similar games for clues about what creates engagement in water puzzle games.

Competitor Review

We looked at three leading titles in the water puzzle genre and found a lot of interesting ideas:

Events

Firstly, all of the games we looked at use events. Some of them integrate events directly into gameplay. A great example is Get Color, which has an event that asks players to complete five levels to feed a cat. I absolutely loved this. The cat brings some emotion to the experience via a much loved animal in need of the players help. It’s present on the gameplay screen at all times, almost like an in-game pet. It’s fed once a day and otherwise sleeps in the corner.

In terms of integrating goals with gameplay this is really clever. It looks pretty light to add but I bet had a decent impact on the metrics we’re interested in. Kudos, Zephyr.

Water Sort Puzzle also has a cat related event (who knew the internet is into cats?) with a reminder on the gameplay screen. This is slightly different in that the challenge is to complete levels to uncover a puzzle. At sixteen fairly difficult levels for each puzzle and three puzzles in total, that’s a big ask. The player has three days to complete the challenge. The event begins at level 50 and is trailed from level 25.

These design decisions provide interesting hints at the metrics Guru Puzzle might be seeing. The average player is playing around 16 levels a day, retention is sagging from day 2. They’ve also made similar UX decisions to Zephyr, but ramped the game design via time period and challenge.

UX

An interesting idea that became apparent was around the concept of failure.

Unlike in CWS, in each of the competitor games, it isn’t technically possible to fail a level. If a player gets stuck (i.e. there is no available tube to move water to) their choice is to play a power-up or restart the level. There is no screen that tells them they’ve failed, instead they must make a proactive choice.

This keeps the player focused on gameplay and moves them into the economy in a positive way. That is, they must either use a power up (earned or paid-for) or watch a rewarded video to continue playing. This feels like a more natural way to integrate spending into gameplay and therefore improve both spend and engagement metrics.

Exacerbating this, in CWS there is no way for players to play an “undo” in a fail state. So, even if there are “undo’s” available, they can’t be played, which feels like a piece of UX that should be cleared up.

It was also clear there are UX changes we could make to CWS to better lead players to their rewards, if that’s the thing that is stopping conversion.

In general the game screen could be updated to integrate key elements of the game. This would have the dual benefit of making the game screen the hub of the game (and therefore maximising attention there) and increasing exposure of valuable features like the Daily Booster Wheel and Shop.

Gameplay

Finally, if gameplay is what CWS players love, each of the other titles mix up the level types where CWS keeps it very consistent.

All the competitor titles include “?” levels. These are slightly more challenging than standard levels as the colours in each tube are hidden behind question marks.

There are two versions, one in which the top colour is visible and another where every colour is hidden unless the tube is tapped. The latter type is a lot more challenging as the player has to remember the top colour in each tube. As the number of tubes and colours grows, it becomes increasingly difficult to remember where each colour is.

Similarly a couple of the games have Daily Challenge levels. These are more challenging than standard levels and outside the main progression, however the basic idea is the same:

And there are special levels. Some of the variance here is supplied by long tubes, some by mixing up the arrangement of tubes. As with the other ideas, they provide a slightly different challenge for the player to engage with.

Summary:

In summary we made the following conclusions:

The goals we tested were not sufficiently engaging to suggest they are worth pursuing to drive LTV or IAP
Players do not seem to be motivated by vanity rewards
- Both these conclusions suggest gameplay is the most compelling element of Color Water Sort
There are UX improvements we could make that may have sustained, incremental improvements to performance

Do Hyper-Casual Players Care About Goals?