Best Sports club in the world - Case Study

(Note: Name confidential)

Business motivation

Data Description

Questions solved

1. Where do customers come from? Which sites/ads are “feeding” our sales?

Considering websites, total sales come from:
• 69% organic (search engines, 64% google, 2% bing, 1% yahoo, 1% navers, 1% others)
• 26% direct
• 5% others
Regarding geography, sales come from:
• 71% Spain
• 5% UK
• 3% France
• 3% Germany and Switz.

• 2% USA
• 16% others
2. Which combinations of sources and mediums attract the more amount of new visitors?
• 54% organic (search engines, 51% google, 1% bing, 1% yahoo, 0,5% navers, 0,5% others)
• 36% direct
• 3% t.co, referral (twitter)
• 2% m.facebook.com and facebook.com, referral
• 4% others
3. Is there a correlation between the time a page takes to load and sales? (Are we loosing clients because the site doesn’t load?)
• In a first analysis it seems that there is a 0.67 correlation between Transactions and Loading time
• However, when one builds a linear regression on the most significant variables (significance level=5%)
• Linear regression shows 0 coefficient for load time and a relatively small coefficient for ln(Load time), but 1 order of magnitude lower than the others
4. Are people who initially came from banners more likely to buy or not?
• Users that landed on the website coming from banners have zero purchases
5. Can we build a model to predict average purchase?
• No, a regression for this would yield R2=8% but the regressions for quantity or total revenues yield R2=99%

Questions to solve

6. Adding to the question 1 of where do customer come from: do users in (club's home city) buy online? Is a significant share of the buyers from (club's home city)? Otherwise we can raise the question if the (locals) know that they can buy online…

7. How do new users find the website? How could the club further spread the brand and increase the number of new users to the website?
8. Are ticket buyers already regular visitors to the site? Or are they mostly coming for one time purchase?
9. What is the average time spent on the website? What do users do on it?
10. Are users accessing from certain brands of computers/smartphones more likely to buy? Or to have a higher Order Average Value?

Analysis of transactions vs. revenue/transaction (operating system)

Analysis of transactions vs. revenue/transaction (operating system)

11. Could we find a way to better segment the market of buyers online?
12. What are the pages with the highest average time on page? Should we suggest to have more of this content? And do these views lead to sales?
13. How could the club further spread the brand? Which dimensions/metrics could we look into to address this? Eg. What is the bounce rate vs some benchmark?

## Warning in wordcloud(gadata9$words, gadata9$count, scale = c(4, 0.5),
## max.words = 200, : champions could not be fit on page. It will not be
## plotted.
## Warning in wordcloud(gadata9$words, gadata9$count, scale = c(4, 0.5),
## max.words = 200, : benzema could not be fit on page. It will not be
## plotted.
## Warning in wordcloud(gadata9$words, gadata9$count, scale = c(4, 0.5),
## max.words = 200, : session could not be fit on page. It will not be
## plotted.
## Warning in wordcloud(gadata9$words, gadata9$count, scale = c(4, 0.5),
## max.words = 200, : benitez could not be fit on page. It will not be
## plotted.
## Warning in wordcloud(gadata9$words, gadata9$count, scale = c(4, 0.5),
## max.words = 200, : espanyol could not be fit on page. It will not be
## plotted.
## Warning in wordcloud(gadata9$words, gadata9$count, scale = c(4, 0.5),
## max.words = 200, : santiago could not be fit on page. It will not be
## plotted.
## Warning in wordcloud(gadata9$words, gadata9$count, scale = c(4, 0.5),
## max.words = 200, : perez could not be fit on page. It will not be plotted.
## Warning in wordcloud(gadata9$words, gadata9$count, scale = c(4, 0.5),
## max.words = 200, : goal could not be fit on page. It will not be plotted.
## Warning in wordcloud(gadata9$words, gadata9$count, scale = c(4, 0.5),
## max.words = 200, : zidane could not be fit on page. It will not be plotted.
## Warning in wordcloud(gadata9$words, gadata9$count, scale = c(4, 0.5),
## max.words = 200, : stage could not be fit on page. It will not be plotted.
## Warning in wordcloud(gadata9$words, gadata9$count, scale = c(4, 0.5),
## max.words = 200, : win could not be fit on page. It will not be plotted.
## Warning in wordcloud(gadata9$words, gadata9$count, scale = c(4, 0.5),
## max.words = 200, : games could not be fit on page. It will not be plotted.
## Warning in wordcloud(gadata9$words, gadata9$count, scale = c(4, 0.5),
## max.words = 200, : ancelotti could not be fit on page. It will not be
## plotted.
## Warning in wordcloud(gadata9$words, gadata9$count, scale = c(4, 0.5),
## max.words = 200, : starting could not be fit on page. It will not be
## plotted.
## Warning in wordcloud(gadata9$words, gadata9$count, scale = c(4, 0.5),
## max.words = 200, : clasico could not be fit on page. It will not be
## plotted.
## Warning in wordcloud(gadata9$words, gadata9$count, scale = c(4, 0.5),
## max.words = 200, : castilla could not be fit on page. It will not be
## plotted.