Lexi’s World started as an idea while my three-year-old played on a keyboard. I made a small cartoon of the genesis of the idea here. It is an app for kids who are just starting to get interested in letters and words. It is not exactly a spelling or reading app, more of a play space for pre-literate children who are letter curious. In 2018, I released it through the Apple Appstore and it received some critical acclaim — but little traction with parents and kids. I have made many changes, additions, and improvements since then, but I have not been able to shift the needle on user uptake. In this assessment, I want to critically evaluate what worked, what didn’t and present some hypotheses about what could work in the future.
What worked
The interaction at the heart of Lexi’s World is a keyboard that allows kids to make words that then come to life. If a child types the letter ‘p’ it will only let them type pig, because, for now at least pig is the only word starting with ‘p’ in the app.
Once they type ‘pig’ they can click a checkmark that will bring a pig to life in the game. Now that there is a pig the word ‘mud’ becomes available. When a kid types ‘m’ now they get the chance to write ‘mud’ or ‘mouse’ which is also available. Through this simple mechanism, they can create a world of pigs, mice, and many other animals. They can help Lexi fly with balloons, ride horses, and unicorns.
Librarians really appreciated it! Right from the start, Lexi’s World got some great press from librarians. Carissa Christner from Madison Public Library was on a local NBC affiliate station demoing it soon after the release in 2018. She got it and shared my enthusiasm for the idea.
In 2019, it was an Honor Recipient for the Association for Library Service to Children’s Excellence in Early Learning Digital Media Award. That was a big deal for me. I am a huge fan of children’s books and the ALA’s Newbery and Caldecott Awards are not only prestigious but also a really practical way to discover new books to read with my kid. I still hope that the Early Learning Digital Media Award will gain that sort of acclaim.
Lexi’s World received two more recognitions from children’s media awards. In 2019 it won the Kiri prize from Digital Ehon in Japan. And it received a mention in the Bologna Children’s Book Fair’s BolognaRagazzi Digital Award for Education. You can see a video of the judges considering it here, Warren Buckleitner says that from a pedagogical point of view, Lexi’s World is “pretty exciting.” I agree, of course! However, it is still unclear how much merit this has in the real world.
Aside from a few letters from teachers and some lovely reports from friends and family, I don’t have a great grasp of how it is being used. In the cases where I have heard positive feedback, it has often involved an enthusiastically engaged adult, which may not in itself be a failing — many children’s books would also fall into that category for example.
What didn’t work
During testing — some kids loved it — I definitely indexed towards them (bad UX research practice — mea culpa). One kid who was nine played with an early demo and got very excited by the notion that some words would unlock others, like dog and bone, bunny and carrot, and cat and box. Some kids were so-so into it but mostly wanted to play other things they already loved on their devices. Lexi’s World has more of a Daniel Tiger demeanor than Mario Cart, so kids who were used to more frenetic digital experiences were not always engaged.
When I released it, a few people bought it, but it was not going to allow me to make it my full-time job (despite it basically being my second full-time job already!). This led me to try to understand what was not working and make changes accordingly.
As a designer practicing design research and strategic design in my day job, I knew what needed to be done. I needed to listen to users, watch users, and try and understand their motivations and pain points. I needed to understand their journey from discovery through to deciding to download and buy the app and play it. It was imperative to understand what kept them playing it and even more crucially how they lost their interest in it.
I also knew that gaining this understanding would be a lot of work, and hard to achieve as an individual outside of the context of institutional support and outside of work hours. I had been testing in an ad-hoc fashion on playdates and it was not enough. In retrospect, I should have scaled that up to some more methodical testing and I still might, more on that later. Beyond the number of downloads and what few reviews I received I had very few ways to determine why it wasn’t working — but that didn’t stop me from having a lot of strong hypotheses.
It wasn’t like my hypotheses came from nowhere. I had watched kids play the game, and I had read extensively on kids’ game design (thanks Toca Boca magazine for such great content), played untold games, and watched my daughter play with games that we deemed were decent. From this experience, I created a list of hypotheses about changes I thought could make a difference, and I spent a lot of time making those changes.
Hypothesis 1. It needed a more explicit play structure.
I always thought of Lexi’s World as more of a play space than a game, but it never quite had the open-ended generative dynamics of a world-building game that a play space might really need to be compelling. Nevertheless, there was something that drove at least some kids to engage with the game. While adults would often be stumped about what to do in Lexi’s World, kids would quickly come to understand the basic mechanism, and those who enjoyed the game would create their own play dynamics that would drive the play. Creating endless unicorns and removing ones with undesirable colors was one slightly dark play pattern I observed for example. Other kids would try and discover every word in the game, something that is more challenging than it first appears. I wanted to keep the openness and not prescribe a specific way for a child to engage with the app, but I also knew that it wasn’t entirely working on its own as an open-ended play space.
I had an aversion to the badge and star reward systems that so many kid’s games seemed to rely on to drive engagement. There is an abundance of games in which a child is shown a picture of a cat and then given the scrambled letters of ‘c’, ‘a’, and ‘t’ to arrange to spell ‘cat.’ When they succeed the result is invariably a rewarding chime and an animation presenting the star you earned. In Lexi’s World, it is the other way around, after spelling ‘cat,’ the reward is a cat. There are stars in Lexi’s World too, to get them a kid must type ‘star.’ I am still proud of this intrinsic reward system, but it isn’t actually all that compelling.
Soon after the release of Lexi’s World a friend sent me an article titled B.F. Skinner: The Most Important Theorist of the 21st Century by Audrey Watters. Watters provided a context for my distaste, with her critique of Skinner’s behavioralism with its extrinsic rewards aimed at changing people’s behaviors and its saturation of digital experiences with ‘nudges’ and ‘hooks’. But she also gave a clue to another pedagogical approach when she wrote that, “Seymour Papert lost and B. F. Skinner won.” I googled ‘Pappert,’ and discovered his constructionist alternative to behavioralism. We can create tools for people to achieve their own projects with their own motivations and learn through the process. The reward for learning is that it enables you to complete your self-defined goals.
With this in mind, I did not want to overly determine the play with prescribed goals, but I did want to add some more hints for possible structured play. Watching kids constantly trying to interact directly with Lexi, and voicing their thoughts about what she wanted in her world. I added wishes in little thought bubbles. If you tap on Lexi once you can see what she is thinking about. If you tap again the keyboard highlights the first letter of that wish. If a child fulfills the wish — Lexi does a little happy jump.
It was a light touch, and had some positive feedback in my limited testing — by this point, it was hard to corral my usual suspects, and I didn’t have a good way to test beyond my friend and family. It had no impact on adoption. In the end, it neither really expanded the world-building and nor did it transform into goal-directed play, it was still in a limbo zone.
Hypothesis 2. It wasn’t pretty enough.
In early 2020, Lexi’s World had been out for over a year and it still was not getting much love in the app store. I was trying to work out if there was any magic string I could pull to change the trajectory and make it something more compelling for kids and parents. One potential string was the look and feel. I had chosen to use a flat rendering style in the game. I thought it made it feel a little more picture-book-like. But combined with some klutzy animation and modeling it didn’t have that connotation for everyone — I knew some people found it ugly. This usually came to me indirectly — another risk in using friends and family for testing — they don’t want to hurt your feelings!
During the course of releasing Lexi’s World, I had reached out to a number of people for advice. I was lucky enough that Björn Jeffery, the co-founder and long-time CEO of Toca Boca was willing to talk. He is a very thoughtful person and had a lot of great feedback and ideas, one of them was that he thought the game needed more texture. I took that as another polite way of saying it doesn’t look so good, he’s British, hence the understatement.
I played with textures, lighting, colors, re-modeling, animation, etc. Unfortunately, this coincided with an upgrade in Unity (the program I built Lexi’s World in) that broke all my models, so I had to reimport them all which was a lengthy process because of the way Unity ties assets into code and UI (grrr). I had to decide whether to spend another three to six months of my free time fixing Lexi’s World and giving it a new look or abandoning it. I put it down for a few months, demoralized, but I still felt like there was something to the project so came back and reworked a lot of the models and added some shading to give it a more traditional low-poly look.
And just because I didn’t have enough to do I added support for French (Thanks Mia and family!). This meant that I had to add more animals to ensure that there was something starting with each letter in French. I added goat, goose, deer, elk, sheep, boot, tiger, and more to the repertoire.
Although I am really happy with the result, it made little difference to the number of downloads from the app store. It reminded me that often the ugliest sites will succeed if they get the core interaction right. Although that isn’t the hypothesis I went with next…
Hypothesis 3. It needed a freemium model.
When I started making Lexi’s World in about 2015, most of the quality kids’ apps used the paid app model. There were a number of companies, notably Toca Boca, Sago Mini, Tinybop, and Originator that were making beautiful and engaging apps and selling them at reasonable prices. The core proposition of a paid app is simple — you pay for the app and you can own it forever, there are no hidden costs, no surprise in-app purchases, no advertising.
By the time Lexi’s World was published in late 2018, it was a different picture. Most kids' apps were free, and even some of the established studios making quality content were switching to a freemium model with additional content or subscriptions offered through in-app content. It is hard to compete with a paid app in a market where well-known and trusted brands are letting parents download their apps for free. There are many other things to consider in the kids’ app market, but giving parents a chance to try before they buy seemed like a necessity.
In a perfect world I could have just checked a box in the AppStore settings to: “Allow free 24 hour trial period,” and Apple would have used a centralized system to allow users to download the app, try it, and buy it to continue using it. In the world we actually live in, I had to recreate the patterns that most similar apps have settled on and create my own 24 hour trial period, a parental gate and store interface for purchasing the product in full via an in-app purchase, as well as account for those who had already downloaded it when it was a paid app. I wouldn’t bore you with that last sentence except that the thought occupied my mind for the three months it took to switch the model.
I wish I could say it paid off. This was undoubtedly a low point for me. Many more people downloaded the app, but no more ended up paying for it. It sunk in immediately, they could try it but did not want to buy it.
Ugh.
It has been a long journey and it might not be the end, but this is a good spot to take a pause and reflect before I continue.
What might work…
Of course, I have plenty more hypotheses. It is one thing to say you will pause and reflect, but immediately upon reflection, I start coming up with more ideas about what isn’t working and what could work. I am going to list some of them here as well as pose a question about how to proceed. What research can be done upfront to validate these hypotheses, or generate new ones?
Untested Hypothesis 1. All the things.
One thing I have witnessed and heard is that kids are really disappointed when their favorite animal is not there. The magic of the game disappears for them.
What if all the words they might expect to be there are there? There is a finite number of words that a three, four or five-year-old knows, and far fewer animals. On the animal front, Lexi’s World probably contains 80% of the animals that they are likely to want. When I list the exceptions, you can see that this is a compelling hypothesis, the game is ‘broken’ without them: giraffe, panda, bear, bird, snail, camel, hippo(potomus), ladybug — and ladybird, puppy, kitten, donkey, monkey, gorilla, crocodile, alligator, chicken, rooster, ox and frog. Maybe meerkats and sloths? And then a few more verbs like dance, crawl, sleep…
Maybe Lexi’s World would become magical if every animal a four-year-old could think of would be there.
Some skeptical thoughts: searching for animals requires parental involvement, it is hard for a four-year-old to discover that an animal is not there without asking what letter it begins with. I would expect that if this was the main blocker for adoption I would hear many more requests for specific words to be added because adults are often good at relaying those thoughts. I have heard some requests, but not many.
Untested Hypothesis 2. Build worlds with words.
Minecraft lets you build a world, Toca Boca Nature lets really young kids build a world. Lexi’s World contains that promise but it isn’t realized. What if the world-building aspect needs to be expanded? How might Lexi’s World expand to allow more world-building?
Make a city, add a ‘road’ and place it, add a ‘car,’ a ‘train’ etc. Make a farm, or go inside and fill a bedroom. Kids could return to these worlds and move things around in them.
Skeptical thoughts: Does the keyboard become secondary, or worse a barrier to this world-building? In the other world-building games the blocks are readily accessible. This seems like it could be amazing, but also maybe not flowing from the core proposition of the magic of words bringing things to life? How could this hypothesis be tested? Would I start adding more words, or does it require a different gameplay mechanism with more drag and dropping of elements?
Untested Hypothesis 3. A more animated Lexi.
When I started thinking about the app, there was no Lexi. I added a character to appreciate the animals being added, and to give them something to walk around. As soon as she appeared in the app, she demanded to be animated. She needed to look at the animals, be able to ride them, feed them and appreciate the addition of everything in the world. Some of the moments most appreciated by kids is when they can identify with Lexi riding a horse or flying with balloons. By contrast, some of the worst moments are when Lexi doesn’t react and just feels a bit detached. How might Lexi become a more engaging personality?
Does she get annoyed by animals? Have favorites? Play more games with them? Can she dance, jump rope, play games? Can she sleep? Wave? Yawn? Sneeze? What makes her really happy? Bored? Does she sit down when you don’t interact with the app? Does she have a backstory? Is she trying to get off this planet? Find her parents? Make it home?
Skeptical thoughts: how does this relate to or reinforce the other protagonist in the game — the child with the keyboard? This sounds awesome but does it actually make the gameplay mechanism any more compelling? How would this be different from just improving the graphics? What are the first steps in this direction that could be tested?
Untested Hypothesis 4. What about the phonemes?
If you have read any article on the science of learning to read in the last ten years you will have heard about the strong evidence for the importance of connecting graphemes to phonemes. While connecting letterforms to sounds is central to learning to read a partially phonetic language like English it is not the sole solution to learning to read. Lexi’s World is not necessarily an app focused on teaching children to read — it is more that it is a chance for kids who are becoming curious about letters and words to follow their interests and learn more about the relationships between letters, words and meaning. Clearly, phonemes are closely related to this realm of pre-literate investigation.
Sesame Street is a huge inspiration to any designer of educational media for small children and their animations of an onset and coda coming together to form a simple word are seminal and the legacy can be seen everywhere. Why not incorporate that into Lexi’s World? When a child makes the word they could tap different components of a word and hear the phoneme? It could be done with no extra affordances — or as one teacher suggested in an email to me- rather than entering the word with a checkmark it could be entered with a swipe across the letters. As the child swipes it could read the word. It might take a few attempts to get the affordances for this right but I think it could be nice.
Skeptical thoughts: I am not that skeptical. I think it could add pedagogical value to the game. Maybe it would help the kids who are a little further along in their pre-literacy explorations. Perhaps there would be a way to demonstrate that educational merit in a way that could have it included in the curriculum, although there are many, many competitors in that space.
Untested Hypothesis 5. Ludic loops.
Skinner won, right? When I directly ask kids what they would like to see in the game, they often say they would like to get stars or points and be able to spend them on Lexi and make here wear different clothes, etc. I have ignored them because I hate this pattern.
Could it be done well? I know that Sky: Children of the Light is a splendid social world exploration game with collecting light and spending it on your character as an important play dynamic. It is beautiful and charming and really, very compelling.
How might Lexi’s World implement a collect and spend system that honors the spirit of the game and still re-enforces the central motif of magically creating a world from words? What if it was tied to the wish-fulfillment? Every time a wish is fulfilled the player gets a gem/seed/wish/promise/point, and when they have enough they can use them to help Lexi — get new clothes, accessories, one-off things — like a house/bike/spaceship?
Skeptical: Is this how everything ends up?
Untested Hypothesis 6. Of course, it hasn’t worked out, there has been no testing!
Writing this last section has been cathartic, but also compelling. Despite my skepticism, I am often a blind optimist when it comes to my own projects. That is probably useful because otherwise, they wouldn’t ever happen, but it is also a terrible approach from a design perspective. I needed to test my hypotheses early and test the app itself a lot more to really understand what is working, what doesn’t work, and what could work — that I might not have thought of. So my last hypothesis — is test, test, test.
I saw some people testing a kid’s science app at the Academy of Science. Maybe I need to take it to the road.
Another wishful thought I had is that you lovely people who have actually made it to the end of this retrospective which has largely been for my own accounting will also find these questions compelling and might want to help. Perhaps you run a game design course, or a product design program, or a UX research program and think that this would make an interesting case study. Or perhaps you are a parent or teacher and think there is something to the idea of this app and would like to see it progress.
Here is how you can help.
This has been a largely solitary endeavor because there is overhead in involving others; I like doing everything; I don’t want to burden others; I want to claim all the fame — LOL. But I think the time has come to ask for help.
- Please leave any feedback about any of the hypotheses, how they could work or not. Clearly, some of these ideas could be merged, and all could be developed. Let me know your opinions! What am I missing? What would you do?
- Test Lexi’s World with parents and kids that you know, I have prepared a Google Form to submit feedback.
- You think one of these hypotheses is awesome and would like to sink a fortune into it.
- Please share this article with anyone you think might be interested.
- Promote Lexi’s World. Do you know a celebrity parent? Do you know a journalist who may be interested in writing about Lexi’s World? Or someone who cares about kids’ apps?
Lexi’s World in the Apple AppStore
https://apps.apple.com/us/app/lexis-world/id1434971401
Lexi’s World in Google Play
https://play.google.com/store/apps/details?id=info.poppoppop.lexi
Thank you!
