Exactly 5 months ago I asked all of you to vote for your favorite deal site, and 375 of you did. The top 5 results now stand like this:
| Deal Site | Votes |
|---|---|
| fatwallet.com | 82 |
| dealtaker.com | 82 |
| dealnews.com | 59 |
| slickdeals.net | 39 |
| markdownmonkey.com | 38 |
The poll was set so that each vote is bound to a unique IP meaning if you used your IP once you can’t vote again. This is not a bullet proof way of eliminating duplicate votes but you really have to make an effort to go around this limitation (if you wanted to).
The comments on the other hand were not limited by IP. With that in mind, here are a couple of facts that make me think the poll results are rigged.
DealTaker comments are spam
I analyzed the comments and it appears 5 out of 8 comments praising DealTaker came from the same IP despite different usernames/emails in the comment signature. Here is the IP used: 71.74.145.215
I appreciate if anyone can find more about the geography of this address. This of course doesn’t prove anything but I am just curious.
Only top 5 positions gain votes
If you take a look at the complete list of votes, you will notice that only the top 5 positions have gained any significant number of votes. The #6 is shopping-bargains.com with twice less votes than #5, markdownmonkey.com
I might be wrong but this fact tells me that the voting is done by the people specifically interested to move the deal site to the top 5 to get listed on the front page.
Results go side by side
This is not as obvious unless you have been watching this for some time like I have and is especially true with positions #1 and #2. I have been seeing it happen several times: as soon as fatwallet.com becomes the front runner, some magical force comes along and dealtaker.com catches up. I have not paid as much attention to #4 and #5 but as you can see they have very close number of votes as well.
I am really distressed by these findings and consider pulling the poll results off the front page. What do you think?
I would vote to remove them. These polls can always be rigged and it is really tough to moderate it.
I can think of the following choices to use that valuable space.
1. Advertisements.. should provide some cash flow so you never stop blogging :)
2. Provide a snapshot of “Popular sites series” (I’m a fan of this series :) ).. You can just provide the best and worst performing sites in each category.
Example: It could look like this
February 2008
Coupon sites (in small font)
coupons.smartsource.com(19.04%) deallocker.com(-30.41%)
Deal sites
dealigg.com (1.56%)
gottadeal.com(-25.44%)
Price comparison sites
.. well you get the idea..
don’t know how difficult this will be . but am assuming it is doable..
thanks for a great blog.
Thanks for the tips. I am against more ads. I already have two sponsors and this is more than enough for me. I love my sponsors but I hate those “thanks to my sponsors” blog posts. ;-)
Let me sit on your other idea. Sounds almost like “market movers” among deal sites. I might do just that if I don’t find anything else to fill the space with.
By the way, you have a very cool website. ;-)
71.74.145.215 = Canton, Ohio
http://www.geobytes.com/IpLocator.htm?GetLocation
According to
http://loookup.com/71.74.145.215
City: Akron, OH.
I know the pain of combating cheaters as my site actually gives out weekly prizes to most voted deals :(
Even better, based on whois, it is an address directly allocated for use by Road Runner. E-mail them about it at abuse@rr.com and you might be able to nail this spammer.
thanks.. I did put a lot of effort into my website and finally came up with this simple design :)
anyways.. I forgot to remove the link from my name. I do intend to bring the website alive soon.
Yan, I think all polls probably have some bias and that includes the ones done by major media outlets and research firms. Your poll has actually stood up very well to attempts to game the system.
As for that space, you could keep the poll up but rotate the results by showing a different set of 5 rankings each page load. For example, 1-5, then 6-10, then 11-15, etc. I also like Navin’s idea.
Alternatively, there are some cool third-party data sources and tools at http://www.compete.com that I’m sure you could find creative uses for.
Also looks like you have a typo in your poll / article. It is slickdeals.net not slickdeals.com which is a crap ad site.
Fixed the typo, thanks
Ouch.. No hanging chads in this one?
Mike:
A lot of people mention Compete.com as a reliable source, I find them wildly inaccurate.
Maybe Yan could do some Compete.com analysis like he did with Alexa… I know he has a lot of free time on his hands ;)
mike g. refers to this analysis, which took me a day to compile, and I have to agree with him on Compete. I have also found their stats being inaccurate in many instances. This VentureBeat article is about one year but it highlights the problem pretty well.
Well, the truth finally comes out. We knew for a long time that DT would stoop to such measures. They spam other deal sites, so it was kind of obvious when one deal site took the lead, without cheating, DT soon was back on top. So sad. But her is the kicker….DT has a paid mod who lives in Canton, OH
Hey Yan! It’s Matt from MarkdownMonkey! It really makes no sense that I’m in the top 5 favorite deal sites. I only get about 3,000 uniques a day and have never even broken into your “popular deal sites” list.
When I saw that I was added to the voting list, I sent an email to my immediate family saying “vote for me!” I never expected more than 3-4 votes, but within a week I was #3 on the list. I never asked her, but I assumed my sister-in-law forwarded my email around to her friends (she’s very social).
Anyway, I don’t think I’ve gotten more than 1-2 votes since then. I’ve just been slowly moving down the list.
Before you decide to remove the poll from your site completely, maybe you would consider removing MarkdownMonkey from the list. That should leave only one questionable listing in the top 10. Who knows, maybe the owner of the other site in question will speak up too and help restore the integrity of the poll.
No integrity there. What you lack in quality, you make up for it in spam. Perfect example of how one bad apple can ruin the whole fruit basket.