The Guess Tracker for RetailMeNot's MoneyJar Contest

UPDATE: Three days after clue #9 was released, it was correctly guessed.  Congratulations to *@tea4two20 for her winning guess of $6019.35!  According to the tracker she had tweeted every day since we started our coverage.  Her two final guesses prior to the correct one were $6000.95 and $6543.15*.

*We tracked about 800 users (+300 since the initial post) who participated.  About 17 participated every day.  Better luck next time to the others that tweeted every day including: *@applespice159, @Doodle741, @natebridi, @sunshyne1977, @tannawings, @uberabby2, @won2x and 10 others.

Just the good stuff: The MoneyJar Guess Tracker will check your guess against all known existing guesses from other users, and takes into account the current clues released by RetailMeNot to ensure a good guess.

The whole story...

I try to stay up to date in the world of online couponing.  From running CouponFollow I've seen all sorts of interesting trends.  Often I've seen merchants put forth great efforts to try to get their coupon codes viral (via Twitter, I'll be releasing more about this topic in another blog post).  But this post is about a coupon site doing something interesting to further its own exposure.  RetailMeNot has a contest going on right now that uses Twitter as the guessing platform. You look at an image of their MoneyJar and simply tweet your guess.  Each "guess" tweet must contain the #MoneyJar hashtag. I should mention at this point that the jar contains between $6,000 and $7,000, which is a nice chunk of change!  You can read the full details here.

Why Track It?

First off, RetailMeNot is the industry leader when it comes to the online couponing.  They were the first (to my knowledge) to allow users to vote and organize coupons based on user voting.  It would be silly not to keep up with what they are doing.

A couple weeks ago I came across this blog post which started me thinking...  Why NOT track it?  Being a curious programmer I wanted to know a couple things.  How many people are actually guessing?  What kind of return loyalty are they seeing from Twitter users? How many users do they reach?  So I built a very simple aggregator to follow the #MoneyJar tag and track the tweets to see what kind of information I could gather.


Based on the sample data gathered so far in the two weeks the tracker has been running the following is apparent:
Sample Size: About 500 different Twitter users have made guesses.
Loyalty: Just over 50% of those have returned to guess more than once (return guesser).  Just under 10% have returned to guess everyday.
Reach: Based on the number of followers that the Twitter users had, their tweets _reached over 300,000 followers _spreading the word about the contest and the RetailMeNot website.

I also noticed that there were a significant amount of users submitting duplicate guesses.  About 13% of the guesses were duplicates.  What a waste!


What came from this experience is the MoneyJar Guess Tracker web app I've hacked together early this morning in just a few hours using SubSonic 3.0 and ASP.NET MVC 2.  It will check your guess against all known existing guesses from other users, and takes into account the current clues released by RetailMeNot... so you don't waste a guess!  I feel this will become increasingly important as the odds increase in favor of the guesser (eg. as more clues are released and more guesses are made). Oh, also you can even have it generate a "good" guess for you, if you're feeling especially lazy.  Note:You still have to manually tweet it as the rules from RetailMeNot clearly state no automated entries.

Enjoy and good luck!