Dognition Business Proposal
Skills: Tableau, Dashboarding, Data Visualization, Data Cleaning
This is a business process change proposal that I developed using a real-world business dataset from the company, Dognition.
The objective of this proposal was to find the best recommendations for the company to increase their user test completion rates. I looked at several factors to explore their possible effects on user participation: number of completed tests per game category, days and times that users were most active, geographic location of users, and the effectiveness of the “Free Start” promotion.
After determining which factors to consider for my analysis, I cleaned and prepared the data by removing test accounts and outliers. Then, I created a variety of visualizations in Tableau and evaluated their significance. Finally, I put together a series of dashboards to present a targeted data story.
Objective
Dognition is a company that allows owners to connect with and learn more about their dogs through different types of games and tests.
The games are broken down into 5 dimensions (Empathy, Communication, Cunning, Memory, and Reasoning) for a total of 20 games. Based on the scores of these games, dogs are placed into one of 9 dog profile types:
Ace
Charmer
Socialite
Expert
Renaissance Dog
Protodog
Einstein
Maverick
Stargazer
The games are designed to be played in a specific order and users receive their dog’s profile type after completing the first 20 games. Users can then continue to pay a monthly or yearly subscription to get 12 additional games, one for each month.
About the Data + Company
Analysis
To find out which Dognition games users tend to drop out of, I grouped the number of users based on game subcategory. I further divided the game subcategories by grouping the number of users for each individual game.
Looking at each of the game subcategories, users tend to have higher drop-off rates after completing all the games in a subcategory, rather than in the middle of the games within a subcategory. The highest drop-off rate happens after the first 20 tests are completed. This is when users are able to receive information about their dog’s profile. Any tests completed after that are additional, so there doesn't seem to be any motivation to complete any more games for users after receiving the profile information for their dog.
The next factor I took into consideration is what time and day users tend to play Dognition games. I grouped the number of users based on when they were active for each day of the week. I also grouped users based on what times they were active in each day.
I found that users typically play the most games on Sundays and there's a decrease throughout the week until it increases again on Saturday. In addition, when users do choose to play Dognition games, they typically do so at around 6-8 pm, with the most games being played at 7 pm. This makes sense considering that people tend to have more time on weekends in general to play with their dogs. On weekdays, people have work or other responsibilities and don't have free time until after work hours, which is typically after 5 pm.
Dognition sometimes chooses to do a Free Start promotion, where some users can sign up and get the first 4 games for free. I took this into consideration to see if the promotion contributed to the number of tests completed by users. I compared the test completion rate of users who received the Free Start promotion with users who did not receive any promotions.
Surprisingly, Free Start users actually have a much lower completion rate than users who did not get the Free Start promotion. About 29% of regular users go on to complete the 20th test, while only 5.72% of Free Start users go on to complete the 20th test. The Free Start promotion doesn't seem to have the intended effect of encouraging users to continue to complete Dognition games. Instead, users seem to mostly complete only the tests they received for free and lose interest in continuing to complete the tests.
Finally, to get an idea of where most of Dognition’s users are from, I grouped the number of users based on country. I then further grouped the US users based on state. More than 91% of users are from the United States and the other top 5 countries are all English-speaking countries. Within the United States, California seems to be the state with the most users across the whole country.
Based on this data, there are certain regions with significantly higher number of users than other regions. Dognition seems to have the most influence in English speaking countries and in very specific states in the United States.
Conclusion
There are several factors that determine whether Dognition users will complete tests and games for their dogs.
Geographic location is a significant factor, since more than 91% of Dognition users are from the United States. Users also tend to play the most games on Saturdays and Sundays at around 7 pm. In addition, there are certain categories of games that users tend to be more motivated to complete.
Based on this data, I have three recommendations that would help Dognition increase their user test completion rates:
Improve the overall game experience by creating incentives for both owners and their dogs to complete each game subcategory.
Encourage users to play Dognition games by sending out promotional materials and reminders right before users are typically most active, which would be during the weekend from 6-8 pm.
Focus on the most efficient marketing methods by getting rid of the Free Start promotion and pivoting towards marketing promotions based on geographic location (primarily in the US and English-speaking countries).