Friday offered some more great sessions on consumers’ decision making that are relevant to the social games industry. For those who did not see my post on Friday, I spent last week at the Behavioral Decision Research in Management Conference (BDRM) in Boulder. My major takeways on Friday were as follows:
- People are overly optimistic about ability to reach goals. In general, people have overly optimistic expectations about reaching goals in the future and there is a sharp drop off right before the goal date. This fact has two effects on social game executives. First, revenue and DAU projections from your marketing department are likely to be too optimistic and you need to take that into effect when planning cash flow. You also should relook your own expectations and see if objectively they are overly optimistic. Moreover, as many with experience in the game space would agree, your development team is likely to be too optimistic about schedules. They will feel they can get work done quicker than expected and only revise close to the launch date.
- Collections are strong motivators but need to be handled appropriately. Collecting something (either a tangible or virtual good) usually is not planned or intended but comes from other stimuli. Moreover, if the collector (player) starts with two unpaired items (maybe given to them for free or started in the tutorial), they are much more likely to purchase a third item (and continue collecting). What is particularly interesting is this effect is statistically much stronger than if the player starts with any other number of items or if the items are paired.
- Customization works better with a starting price. Players are more likely to monetize when there is a starting price for a base or standard configuration. For example, if you are offering the player the opportunity to customize an avatar with costs for different items, by offering them a pre-made avatar at a set price, it will help their overall decision making (and make them more likely to monetize). The other key takeaway here is that a starting price is not a bait and switch but an actual configuration that players may want.
- People are more likely to feel they can complete a goal if you give them a head start. If someone needs to complete ten tasks in a game, they will be more motivated if you give them the same ten tasks as a set of twelve tasks with the first two already done.
- Consider prediction markets to make green light decisions. Prediction markets (like online exchanges where users predict different outcomes)consistently outperform experts in predicting future outcomes. Given the costs and opportunity costs in game development, setting up a prediction could be a very effective way to manage game green light choices.
- Default options affect meaning, not only the final choice. The default options you give players will help them frame the entire mechanic or scenario, not simply potentially lead them to a desirable (by you) tactical decision. Thus, you may lead them in the short term to make a decision that helps monetization but long-term has a negative effect on LTV.
- Experiential gifts are more socially connecting than material gifts. In this study, it was found that experiential gifts, where the gift giver creates the gift, creates more social connectedness. Given the importance of gifting in social games, this finding suggests the player should have more control in creating a gift than just sending a friend some artichokes seeds.
- Near miss events do not increase impact Conventional wisdom holds that in a luck event, like the spinning of a roulette wheel, if the player comes close to winning they are more likely to expect to win next time and behave accordingly (i.e. more likely to purchases additional spins or come back the next day to spin again). Given how many social games use a “random” lottery as part of their daily engagement mechanic, this finding suggests that some of the current practices are not driving the desired behavior.
- Don’t always trust your BI folks. An issue in academia is researchers manipulating the p-curve (the probability that what was found happened by chance and was not due to the null hypothesis being correct), ranging from fudging to dishonesty. This can be done as innocently as just using data that confirms the hypothesis and leaving the rest in the drawer (called file drawers). I am not sure how prevalent similar practices are in our space but given that our analytics experts often have much deeper statistical skills than anyone else at a social game company, it is something to keep on the lookout for. If you are suspicious, there are tests you can make of the p-scores to judge their validity. You can also ask our analysts if all the data they collected was used and, if not, why.
That is it for BDRM 2012. Overall, I found it much more useful than most of the industry specific shows and I recommend you look outside the box when deciding what shows and conferences to attend.