Shopmania is a downloadable game title that Gamelab created for casual game publisher iWin. In Shopmania, the player is cast as Lewis, a hapless worker at the $pendmoore super-store, a huge and exploitative department store chain. As Lewis, the player's job is to fill the carts of shoppers with objects that roll by on conveyor belts. (The motto of $pendmoore is "We do the thinking, so you don't have to!") Rather than covering the entire production process of making the game, this article takes a close look at one crucial moment in its development.

Filled with characters, cut-scenes, multiple play modes, lots of special items, and 60 unique levels, Shopmania represented a great deal of hard work from the project team. When iWin first released the game in May 2006, we all had high hopes that the game would do well on portal sites. When the game didn't perform up to our expectations, we decided to strategically redesign the game. This essay is about that redesign: what we did, why we did it, and what we learned. Within our experience on this particular game, there are lessons here about the design and the development process of casual games in general. By sharing our experiences here, we hope you'll find something useful for your own projects.

Inventing Play at Gamelab But first, a bit of context. Gamelab is a game development company based in New York City. Founded by Peter Lee and Eric in 2000, Gamelab currently has a staff of 35 and focuses on creating original downloadable games, including Diner Dash, Plantasia, Egg vs. Chicken, and Plantasia (and several more games launching soon). We also create multiplayer online games, educational games, and do strategic design consulting. Gamelab is a creatively-driven company, which means we always look to invent something new in our games. That "something new" includes new kinds of aesthetics, new content themes, and new forms of gameplay. And because this means we often give ourselves problems that we don't exactly know how to solve, our production process has to be flexible enough to accommodate experimentation and iteration.

Our process methodology is based on rapid prototyping and constant playtesting. Design decisions are based on our ongoing experience of playing the game, as well as from playtester feedback. At the same time, Gamelab has to survive as a commercial studio, and we can't afford to endlessly iterate our ideas. The moment in Shopmania's development that we write about here is a great example of the tension between innovation and project completion. The content and gameplay of Shopmania deviates from the norm: early on in the project we changed the game theme from fast food to shopping, and the Tangrams/Tetris/Bingo/resource management gameplay was not a direct copy of an existing game. So as with other Gamelab titles, there was much about Shopmania that we had to invent from scratch. And it was these very innovations that proved quite challenging in the design and development of the game.

All the Makings of a Hit Game Many aspects of Shopmania represented an attempt to do something new. The theme and storyline of the game (an ironic narrative about a domineering megastore dedicated to unapolagetic consumerism) was closer to Ren & Stimpy than the more neutral casual game content. Instead of the comic book style cutscenes we used in our last four games, Shopmania had animated interstitials. The gameplay, if not wholly unfamiliar, mixed inlay gameplay with resource management and color matching in new ways.

Of course, we weren't just making a vanity game. While Gamelab believes that the key to capturing the attention of the casual market is to create genuinely new game experiences for players, the newness of the game – its departures from expectations – should not alienate the very players for whom we've designed it. For us, one of the goals of making something special is to reach as many players as possible.

A few months prior to the pre-launch rush, the Shopmania development team had met with Gamelab's Design Advisory Board (a collection of local design and industry professionals outside our company that we use for expert critical feedback) for a reality check about the direction of the game. The Board assured us that Shopmania had a shot at being a hit title: it had fun content, intuitive and innovative gameplay, as well as great looking original art and a unique audio palette. In Shopmania, we had a game with something new and also "all the makings of a hit title!" We began to consider Shopmania's future market success a given – an overconfidence that turned out to be premature. We were eager for the time when players could download the game, play it, and see all of the silly consumer products we'd made for the five departments and the gameplay elements we'd created and then designed levels around. There were interactive, visual, and audio punchlines throughout the game, and the core mechanic seemed solid. Peter Nicolai, a programmer at Gamelab who was working on another title, remarked about Shopmania, "If this game isn't a hit, there is something is wrong with casual games."

Excitement Prior to Launch As development of Shopmania was wrapping up in April and May 2006, there was a bubble of excitement around the Gamelab office. After months of brainstorming, questioning game mechanics, playtesting, revising gameplay, endlessly revising the game interface, reworking the narrative, writing and revising in-game copy, improving our approach to the audio, creating animations, recreating animations, and generating playful visual and audio effects, the launch of the game had become highly anticipated. All that the Shopmania team really had left to do was to get through the Quality Assurance testing process with the help of our publisher, iWin. The game would shortly thereafter be available to the public! Then we would finally be able to send emails to our friends and family, proudly announcing the culmination of our hard work in the realization of another fun game with a 60-minute trial and a reasonable price tag. The end was in sight.

Gamelabbers both on and off the Shopmania development team snuck in superfluous play sessions to play the part of Lewis, addictively filling up shoppers' carts and fulfilling their requests. Staff and advanced playtesters alike were genuinely having fun making strategic choices about which upgrade powerup goodies to use and when to use them: Increase the size of a shopper's cart? Make another angry shopper patient again with a sweet? Upgrade the conveyor belt? Force the shoppers to make requests by investing in advertising? Or try a new combo of approaches? We were using the data generated from all of these play sessions ensure that point goals were properly balanced for both novice and, particularly, expert players. We felt that we had made something special and we couldn't wait to share it with the world.

Something is Wrong On May 22, 2006, iWin announced the release of Shopmania on iWin.com in its newsletter. This was it! Both iWin and Gamelab had been looking forward to this day for so long! We anxiously watched the reviews on iWin.com and, later, Big Fish Games and Yahoo! Games, hoping to bask in praise. After all, we were convinced that Shopmania had all the makings of a hit game. But Shopmania was not a hit in the casual game market. We watched it receive lukewarm reviews on iWin.com. When it launched on Yahoo!, the reviews were polarized. By June 6th, Shopmania was already entirely off the front page of Big Fish. People either loved the game or hated it. Worse than that, nobody was buying the game. According to iWin, Shopmania was generating a lot of interest and had an unusually high volume of downloads. But Shopmania apparently wasn't following through on casual gamer expectations and so they weren't "converting" – that is, they weren't buying the game.

Something was wrong with Shopmania. As designers, it is our job to anticipate a player's needs. Though gamers can be a bit finicky and seemingly unpredictable, we know that they will reward a truly well-designed game with lots of playtime and a high purchase rate. Diner Dash was a bit unusual for the market when we first released it in 2004, but players loved it and its success story goes without saying. Shopmania on the other hand wasn't succeeding, and we couldn't just blame the fickleness of the casual market (as tempting as that may be after spending months of hard work making something wonderful for them). Gamelab had somehow failed on a design level. We hadn't properly anticipated our players' needs.

A Second Chance After Shopmania fell off the front page of Big Fish Games, iWin contacted Gamelab via e-mail to begin a dialogue about the game's poor performance. iWin assured us that they still felt that Shopmania had "huge potential" and asked Gamelab to "evaluate what steps can be taken that can result in a performance bump." Since Shopmania hadn't launched yet on most of the major portals, there remained time and opportunity to make revisions. Everyone at Gamelab was disheartened by Shopmania's low conversion, tired of working on Shopmania, and already in development on other games. But we had a second chance. And it is rare to have a second chance to fix something that is wrong. Gamelab began approaching its second chance by working with iWin to coalesce useful, critical reviews and data. Players seemed to be having difficulty with the conveyor belt. They seemed to either get the game 100% or not get it at all. Yahoo! reviews included statements like:

• After 20 minutes, felt like I saw the whole game...
• ...after about five minutes it bored me.
• This game seemed a bit simple at first and then took off to where I found myself wanting to get to the next level to see what other obstacle was going to be put in my way!
• Could entertain for hours, very fun, great for almost all ages.

The positive reviews seemed to be from people who had invested longer play sessions. We were certain that we had successfully provided addictive gameplay and content rollout for advanced players (like ourselves) who'd already bought into the game. During production, it had been a priority for the Shopmania team to reward long term players with an excellent game experience. But had we done this at the expense of the early player experience?

Gamelab Co-Founder Peter Lee (who was not on the Shopmania development team) thought that we had underestimated our players' ability to learn the resource management and strategic elements of the gameplay. In one email to iWin, Gamelab summarized Peter's concern: "The game doesn't quite feel like a full game until about 30 minutes in. The carts stay too small for too long, and the game doesn't get interesting soon enough." We decided to focus on tightening the first 30 minutes of gameplay.

The Shopmania team met to discuss the changes that we could quickly make to the game. After sharing our solutions with iWin on June 9th, we began implementing them. On June 13th, Gamelab sent a revision of Shopmania to iWin with small but what we hoped were significant changes. We immediately received generally positive feedback from iWin about the changes. We also received more information from iWin via e-mail about the people downloading and playing Shopmania: People played the game for around 25 minutes per session before quitting. A good number of people did play multiple sessions – meaning they played the game a second time, either after quitting to the main menu or quitting altogether. But a good number quit mid-way through the trial. This confirmed our suspicion that we had not yet solved the problem of the first-time player experience. iWin put the revised game through another round of Quality Assurance testing and the Shopmania team resumed its work on other projects. During the week of June 26th, a full month after the initial soft launch of Shopmania on iWin.com, representatives of Gamelab and iWin met at Causality in Seattle. It was agreed that the changes Gamelab had made were good, but probably not enough to actually improve the game's market performance by earning the trust of first-time players.

A Second-and-a-Half Chance iWin again worked closely with Gamelab to understand player reviews, evaluate related download data, and tease out specific design flaws. While folks at iWin were having their own internal meetings about possible design solutions to these flaws, they empowered Gamelab to tackle the challenge independently. iWin's ideas were discussed only after Gamelab had sufficient time to brainstorm internally. We often identified the best solutions as the ones that both groups – publisher and developer – had gravitated towards.

Gamelab had designed the early parts of the Shopmania for a first-time player we hadn't actually met and that apparently didn't like our approach. What we were offering in these first moments of gameplay didn't earn the players' trust that they would enjoy the later moments. Why pay $20 for something that you doubt will be fun? Our first round of post-launch changes weren't enough to earn that very important trust which motivates a player's purchase. The Shopmania team had tried to respond to iWin's concerns by implementing small changes in order to save on development costs. This time around, the development team sensed that it was time to bring out the big guns and think in terms of broader, perhaps more costly, changes to the game. A company-wide email was sent around Gamelab imploring everyone to play the game and offer critical feedback:

Sent: Thursday, June 29, 2006 12:44 PM
To: staff
Subject: pretend you're a casual gamer with $20 to burn or not burn...
...and go to iwin, download shopmania, and play. After 5 minutes, ask yourself "do I want to keep playing? Why? Why not?" After 30 minutes, ask yourself "how about now?" After 45 minutes, ask yourself "what will I do when my trial runs out?" After an hour ask yourself "should I buy this? Why? Why not?" After this experience, reflect, brainstorm, create a list of ideas – wild, mild, and in between – of what we can do to turn this very-popular-to-download but not-very-popular-to-buy game into The Next Big Thing and a commercial success. There's nothing wrong with the gamers or the game – just the first 45 minutes or so… PLAY IT!

Although we had a build of Shopmania available for easy access on Gamelab's internal network, it was important that Gamelabbers go through the actual process of downloading and installing the game onto their computers to get a more authentic perspective on where we were letting down Shopmania's first-time players. The request for feedback was immensely successful and there was a spirit of solidarity around the Gamelab office. Staff dropped everything to go to iWin.com and pretend to be a casual gamer for an hour. Everyone knew that Shopmania by all means should and could be performing well. Gamelabbers were excited about the design challenge and acutely aware of the novelty of having the privilege to revise the game.

We were lucky: lucky to have the chance to revise the game; lucky that our publisher was still looking to us for design solutions rather than assigning them; lucky that Shopmania hadn't yet launched everywhere. Feedback about Shopmania came from all arms of Gamelab: visual designers, programmers, the office manager, project managers, interns, and friends. Catherine, as the team project manager, coalesced all of the various feedback that came in via e-mail and IM conversations and led a company-wide meeting just one day after the call to arms e-mail was sent around. The meeting began with a recitation of everyone's feedback, just to set the tone for "anything goes" and "critical feedback is good feedback." As a company team, we narrowed in on the design flaws in the game, and we agreed upon solutions. Many of the solutions required pretty significant additional production, but we knew that all the solutions were imperative for an effective revision. We had a second chance, and we didn't want to blow it.

What we did The end result of our process, of course, was to change the game itself. Eric, as lead game designer on the project, worked with Catherine and the rest of the project team to select the actual new features to implement. We didn't have the time to do a whole new round of redesign, implementation, and then testing to make sure that we had gotten it right (both for reasons of cost and because of our desire to get the redesigned game out to portals as quickly as possible). Instead, we had to be quite decisive. While we did do a bit of user testing during the redesign process, especially to see if we were getting the game tutorial and the first 30 minutes right, for the most part we jumped right in, both feet first. In the next several sections, we detail the primary changes that we actually made to the game. This list leaves out lots of small fixes and polishes that were made to the visuals, audio, and text. But it summarizes the most important aspects of the redesign we implemented. Each set of changes concludes with the "design takeaways" that we learned through the analysis and redesign.

The New Player Experience In the original version of Shopmania, we approached the first several levels of the game as a gradual tutorial that introduced the player to the basic game elements and the core gameplay. This approach was based on the generally held casual game wisdom that downloadable games should be very easy to play, and that the frustration of losing a level should be minimized. However, the problem with going too far in this direction is that the game ends up feeling like interactive muzak: you can play forever and not really lose, and the essential tension and challenge of a good game are lost.

From our analysis, players were telling us that the first seven or eight levels felt like a tutorial. By the third or fourth level, we had playtesters exclaiming out loud, "I get this game. Can I skip the tutorial?" This is bad news for a title that has to live or die on whether the first hour of play is truly compelling. Our mantra for the new player experience became, "The first level has to be FUN!" Rather than feeling like a hand-holding tutorial, the very first level of the game needed to provide enough challenge and excitement so that it felt like a real game. Because Shopmania's gameplay was not a copy of an existing game, it was a delicate balancing act to make sure the player knew what to do right away while also giving them an enjoyable and engaging level from the get-go. Our adjustments to the game help and tutorial (which we describe in the next section) also worked into this design goal. So we made the first level and the several levels that followed busier and more hectic. We introduced special objects and items into the game much sooner. This resulted in a more challenging game overall, and to compensate, we ratcheted down other aspects of the gameplay to make it more forgiving. For example, we made the shoppers that the player is assisting more "patient" so that they wouldn't penalize the player by leaving quickly. We even superficially varied the way each level looked. We had thought that new players would want a familiar, standardized level layout – at least until they became more familiar with the overall gameplay.

So the first 10 levels really looked very similar, and we didn't start playing with the level layouts until much later in the game. Luckily, levels were structured modularly, and so it was easy for us to play with the arrangement and format of different level elements. In this way, the early game felt less like a monotonous tutorial and more like a fully rich game space that the player was exploring. If there is variety in the early levels, then players will assume that variety continues throughout the whole game. One other small change we made was to modify the opening main menu so that first-time players knew exactly what to do. Although the arrangement of the options on the main menu highlighted the button for PLAY LEWIS' STORY (the main game mode), we tweaked the design to emphasize this button even more, including the addition of strategically placed animated sparkles. The last thing we wanted was a player confused about what to do even before the game started! The changes to the main menu, combined with the idea of making the early levels genuinely enjoyable, were meant to get the player right into the fun of the game as soon as possible.

Design Takeaways:

• It's OK to make games challenging. Don't make games – even casual games – too easy.
• The first level of a game should be FUN, right from the very start. Don't make early levels feel like an extended tutorial.
• Providing variety in early levels is important. Make players feel like the game has a lot to offer in the short- and the long-term.
• Players should know exactly what to do the first time they reach the main menu.

Helping Out Help Making sure that players quickly and easily learn how to actually play is incredibly important for casual games. Our approach in the first version of Shopmania was to minimize preliminary instructions screens and instead put players right into the first level. When the players needed to learn something new, we paused the gameplay and displayed a tutorial message in a pop-up bubble showing them what to do. In order to be complete and to minimize confusion, we made sure that every new action and feature was spelled out for players in this way. Although well intended, our analysis showed us that this approach wasn't working. Instead of feeling like they were jumping right into things, players were getting annoyed by the constant interruptions.

They weren't reading all of the tutorial text, and they ended up just as confused as they would have been without instruction. Most importantly, they weren't having fun because of the interruptions. Our response was to go for the approach that we had initially resisted. We added a non-interactive instructions screen to the game, which the player sees right before the first level starts. The purpose of this screen was to give the player a "big picture" idea of how to play, focusing on the main game elements (take objects from the conveyor belt, place them in customer carts, and ring up the carts). We animated the screen to clearly show the sequential aspects of this process and we designed the screen so that players had to view the sequence at least once before moving on.

At the same time, because this general screen covered so many of the game's basics, we were able to greatly reduce the interruptive tutorial messages. We still used them throughout the first few levels, but very sparingly – only when we really needed them or when playtesters were getting confused. We also made a heroic effort to cut down all of the game text that appeared in the first several levels, including the help text, the tutorial bubble text, and the level intro text. Shopmania has a lot of character and content, and even though we wanted to use the game text to establish the game world early on, we decided to radically reduce the text so that people would actually read it. Later on in the game, we kept the level intros a bit longer, so that players who had gotten into the storyline and world could enjoy them, while others could click on through.

The results were positive. Players were able to "grok" the general flow of the gameplay from the animated help screen. This meant that even though they might skim through the later pop-up tutorial windows, they were comfortable with the basic shape of the game and could figure out the details themselves. Streamlining the tutorial interruptions and shortening the text also kept players in the game, which was key in making the first few levels fun and engaging, as discussed above. More than any of the other changes, these redesigns of the new player experience were instrumental in our strategy for increasing conversion.

Design Takeaways:

• Help players understand the "big picture" of the gameplay as soon as possible. Don't be afraid of an instructions screen.
• Keep tutorial interruptions to a minimum.
• Reduce the amount of text whenever possible.

Ironing Out Gameplay Wrinkles One of the areas of player concern that bubbled up through our analysis of game portal message boards was interaction with the conveyor belt. Although we had tuned and refined this aspect of the game, some players were still mentioning that the interaction with the objects and the belt was awkward. This troubled us. Most games have a core mechanic, a repeated action that players do over and over again as they play a game. In the case of Shopmania, this mechanic was using the mouse to pick up items from the belt, rotating the items with the right mouse button, and placing them in a customer's cart. If there was something about this interaction that didn't feel right to a player, it was something that could possibly cloud the whole gameplay experience.

One challenge with this interaction was that the game was by nature a little fussy. Players had to pick up objects, a few of which were quite small, and place them carefully in an exact spot in a customer's cart. We had already made many adjustments to this core mechanic to make it easier on players. For example, the game supported both "click-and-stick" interaction, in which a player clicked and released the mouse button to grab an object, as well as "click-and-hold" interaction, in which a player clicked and held down the mouse button. We also had a "snap to grid" feature in the customer carts that helped players place objects. But as our testing showed, there were still problems. After brainstorming, playtester observation, and detailed analysis of some highly ambiguous message board posts, we decided that players were having trouble putting objects back on the belt (a strategy some players liked to use), and also that they weren't figuring out how the belt's "sensor" worked, which can be used to quickly fill up an empty belt with new objects.

So we implemented two changes to deal with each of these problems. Putting objects back on the belt was awkward because belt objects sat on an invisible grid, and it wasn't always clear where an object might fit. In our redesign, we implemented a simple "physics" system that let players put an object anywhere they wanted, with any blocking objects pushed out of the way automatically. The result was a fresher and more open-feeling interaction, for at least that aspect of the core mechanic. The belt sensor problem was trickier. As in many games when objects roll out on a belt to be grabbed, there are times when the belt gets empty and the game needs to speed up the procession of objects in order to give the player more options and keep up the flow of the game. We had implemented a sensor system that worked like a belt sensor at a supermarket checkout station. If no objects had reached the sensor point on the belt, the belt moved more quickly until objects did reach the sensor spot - and then it slowed back down again. It was not a perfect system, since the belt could be mostly empty and still move slowly if there was a single object on the far end of the belt beyond the sensor spot. However, players usually learned the subtle trick of picking up objects past the sensor to "trick" the sensor into speeding up and getting more objects into play.

Playtesters we observed who figured this out were often pleased with their own cleverness. However, it seemed that some players never figured it out and would consequently end up with one object at the far end of the slow-moving belt, painfully waiting for the game belt to bring more objects into view. Our redesign here was to add a "speed-up" button to the belt. When pressed, the belt moved forward quickly, making more objects appear in view. We tried variations with and without the sensor in place, and although we thought the button would replace the sensor, it seemed to work best when both features were in the game. Most playtesters used the button, but the sensor was there as an additional, more "automated," way to get new objects on-screen during very fast gameplay.

These two changes, while they might seem trivial, had a deep impact on the game experience. Even though most of the core mechanic worked very well, our testing had identified a couple of minor moments where it was breaking down. Since the core mechanic of a game is repeated thousands of times in a play session, the cumulative effect of a small hiccup can be substantial. But for the exact same reason, we were very hesitant to make any changes to the core mechanic at such a late date. Our concern was that adjustments to such a fundamental aspect of the game might have repercussions for the game's difficulty. After trying many of the game levels with the new mechanics, however, we decided that these changes didn't break other aspects of the game, and we kept them in. The game is much better as a result.

Design Takeaways:

• The core mechanic of your game should be as smooth and easy as possible at all times. Look for moments when it becomes awkward or unintuitive.
• Only redesign aspects of your core mechanic with great care, because it can have many unintended repercussions.

Meaningful Play Making the core mechanic of a game smooth and intuitive is important. But beyond crafting the basic interaction, it is crucial that players know not just what to do, but also how their activity becomes meaningful in the context of a game. In other words, the visual and audio feedback that players get for their actions, the "power moments" that occur when players successfully achieve something in the game, are just as important as the mechanics of the interactivity itself. The audiovisual effects that surround interactivity are what signify to a player the meaning of their actions. And any good game had to get this kind of signification right.

During the production of Shopmania, we had designed and redesigned the visual and audio effects many times. Shopmania has multiple, overlapping economies (dollars, "thumbs up" approval points, a line of shoppers, individual shopper impatience, goals and expert goals, etc). Whenever possible, the game tries to highlight the ways that player actions impact these resources and values. For the redesign, we wanted to heighten the feeling of player reward as the result of successful action, and we implemented a number of adjustments to this effect. The focus of short-term player accomplishment in Shopmania is ringing up shopping carts, and we decided to make that moment a stronger experience. For example, in the originally launched version of the game, a cheerleader yell of "shop-MANIA!!!" happened every time a player rung up a shopping cart that was completely full. Players loved this sound, and it was even mentioned in Shopmania's online review at Gamezebo.com. So we inserted more audio events of a similar flavor at other ring-up moments, such as when a customer's cart was full and in addition full with all the same color. These additions not only enriched the sonic texture of the game but also helped to signify differences in levels of achievement (full cart versus full cart same color).

Similarly, we implemented several small changes to the visual elements that accompany the moments before and after ringing up carts. A cart full of objects received a color-specific sparkle animation, the numbers that display the player's score for a cart were enlarged, and the dollars that flew from a cart into the player's score meter were also tweaked to increase visual impact. We also put additional information on the level win graphic that let players know how many full carts they had completed during the level, further patting them on the back for going that extra mile. It is doubtful that any player would notice all of these small changes. However, the cumulative effect is to make the achievement action of the game – ringing up carts – satisfying and delicious. Anything that can make the player savor the positive moments in a game can potentially bring the player back to play again, and possibly purchase the title as well.

Design Takeaways:

• Beyond a fun interactive core mechanic, a game must signify the results of player action in a way that is clear to the player.
• Identify the "power moments" in your game where players are achieving success through action and make these moments as satisfying as possible.

A Player's Progress One thing that was clear from the low conversion rates of the original game release was that players were not getting "hooked" into long-term play. Players were not becoming motivated to stick around beyond that first hour and see what the rest of the game had in store. So during our redesign, we decided to address this issue by giving players more of a reason to stay in the game for the long haul. The challenge in trying to give players a large game to explore (whether that means several game modes, lots of game levels, an expansive game world, hidden characters to unlock, etc) is letting the players know that these features exist in the first place. It is important to create game elements that will extend the life of the game, but it is just as important that players know about them.

In a casual game, with only an hour of free play, players need to quickly be catching glimpses of the wonderful places in store for them in hours two, three, four, and beyond. We addressed these issues in the Shopmania redesign in two ways: by keeping players in touch with their progress through the levels, and by inventing a whole new system of long-term rewards. In level-based games, a common design dilemma is keeping players aware of their progress. It is essential that players understand how they are marching through the game's larger level structure as they play – this is, after all, one of the ways that players achieve satisfaction. At the same time, it's also important not to bog down players by having them look at "level landscapes" that track their progress between every level (in Shopmania's case, this screen is a map-like view of the superstore's departments). Movement on these level progression screens is usually very subtle, and seeing them over and over again can become monotonous. The initial launch of Shopmania erred on the side of moving the player quickly from one level to the next, without moving back to the level landscape screen in-between levels. However, the feedback from players that felt "stuck in the tutorial" for the first several levels motivated us to change this scheme. In the redesign, the player sees the progress on the level landscape screen after every level, and to spice it up a bit we added an animation of Lewis moving through the screen as he reaches each new level.

A more radical change to address players' limited sense of progress was the addition of $pendmore "merit badges" which are awarded to the player gradually over time. We wanted to provide players an additional level of "rewards" that gave the larger game some kind of shape. Yet because of our time limitations, we couldn't invent something wholly new. Instead, we decided to turn the existing game elements – shoppers that players encounter over time, powerups that appear in levels – into "rewards," in the form of merit badges. The narrative conceit was that once the player had played a level with one of these new elements, the player was then officially qualified to serve that type of shopper or use the new powerup. In essence, we had reframed "normal" game items into special rewards. While this may seem like a simple slight of hand, once the merit badge system and screens were implemented, they felt very natural in the game's world and in the flow of the gameplay. More importantly, they provided players with a strong incentive to continue playing. Silhouettes of merit badges yet-to-be-earned tantalize players with the possibilities of what the rest of the game has in store. Even though players can look through the help section and see all of the customers and powerups that will eventually appear as merit badges, the experience of earning and unlocking each badge over time creates a strong experience of reward and achievement.

As with other aspects of the redesign, we never expected that a player would purchase Shopmania solely because of the merit badge system – let alone the simple change of seeing Lewis move on the level landscape regularly – but we hoped that these changes, taken cumulatively with everything else, would make the difference. It's not enough that the moment-to-moment play of a game is fun. Players need to be made aware of how their long-term involvement in the game is also reaping rewards.

Design Takeaways:

• Long-term progression through a game must be made meaningful and significant for players.
• It is important to have many game elements and levels for players to discover over time: it is equally important that players are aware of the existence of those elements and levels before they are unlocked.
• Returning players to the "level landscape" is a good way of keeping them in touch with their progression through levels.
• A "rewards" system that gives players periodic medals or badges is useful to provide players with a sense of gradual and steady achievement.

Closing Thoughts The experience of revising Shopmania after its launch was exasperating, exhilarating, and very educational. It was disappointing that we didn't get the game right the first time. It was frustrating to have to devote extra resources and staff time to a project after we thought we had already launched the game. But it was also a very inspirational experience, in terms of publisher-developer collaboration, our enlarged understanding of the casual game market, and Gamelab's overall development and design process. We surprised ourselves with how fast we were able to complete significant changes and new content creation. Since then, when people in the Gamelab office feel anxious about our ability to pull off deadlines, they are occasionally reminded of the volume of work we did in such a short time on Shopmania.

iWin's respect for Gamelab during the redesign phase was crucial in maintaining our enthusiasm. Without their supportive approach, Gamelab would not have had the appropriate mindspace to brainstorm and think openly about the game – to perform emergency reconstructive surgery on the "baby" that we had all grown to love. But iWin never told us exactly what to do. They helped us research and analyze what wasn't working, they lent their brainpower as we were thinking through the redesign problems, but in the end iWin trusted the Gamelab team to select and implement the features necessary to improve the game's market performance. Their confidence in our abilities actually pushed us even farther, by making us feel like we had to live up to their high expectations for our team. And we did. Along the way, we came up with ideas and approaches that have since positively informed our later games. So many online games are produced each year, and most of them are just a flash in the pan, appearing briefly on "new games" listings before being relegated to a portal's back catalog. To produce great games takes time, patience, and the ability to improvise. We're grateful that our publisher had the vision to take a chance with a new kind of game, and the good sense to halt distribution and give us time to redesign it.

Moving the game industry forward by reaching new audiences, exploring new kinds of content, and inventing new gameplay will take this kind of courage from all of us. In the case of Shopmania, we managed to rescue a would-be castaway. We produced a "second edition" of the game before many people had even heard of the "first edition." Will our redesign improve the game's performance in the marketplace? Only time will tell. However, as of this writing, the revised version of Shopmania is currently enjoying many weeks on Real Arcade's Top Ten list, and is occupying top slots in other game portals, both large and small. It just might be possible that the Shopmania saga has a happy ending after all.

The Shopmania Team

Gamelab: Tim Conkling, Heath Goldman, Catherine Herdlick, Peter Lee, Carolina Moya, Seok Min Oh, Yunho Seo, Greg Trefry, Catherine Yoo, Jacqueline Yue, Stephanie Yue, Eric Zimmerman

iWin: Jared Brinkley, Greg Harper, Guy Hill, Jason Katsanis, Lee Lakas, Rafael Meza, Kevin Richardson, Nick Rush, Jim Stern, Andrew Viloria, CJ Wolf

Audiobrain: Audrey Arbeeny, Michael Sweet

Essay by Catherine Herdlick and Eric Zimmerman