Thoughts on running a competition online and on Twitter

So now that the last of the Free Travelfish Guides have been credited out and leaving just some Twitter freebies to be tied up, here is a bit of a reflection on what worked and what didn’t — hopefully will be help to others thinking of running competitions on Twitter — and in general for that matter.

On Sunday, July 12 we ran a series of competitions commencing at 8am (GMT+8) and completing at midnight the same day. There were three major prizes and a bunch of smaller prizes (t-shirts, Flickr upgrades, Travelfish Guide giveaways).

The small prizes were given away via regular (roughly hourly) “Quick Quizzes” through the day — on both Twitter and the Travelfish Forum.

The big prizes were given away via a 30-words or less competition on the Travelfish Forum. For each prize a contestant had to say in 30 words or less why they should win the prize. These opened at 8am and closed at midnight.

Results at a glance
Including duplicates and double answers, we saw the following number of entries (and views) on the Travelfish forum:

$1000 travel voucher 48 (933)
Four day live aboard 23 (514)
PADI course 25 (511)
Free Travelfish Guides 56 (762)

Separately we saw the largest number of new members join the site ever, site traffic was up roughly 20% for a Sunday along with a significant spike in Adsense earnings (the easiest bit of advertising on the site to quantify). Travelfish Guide sales spiked on the celebration day (odd as we were giving them away) and the Monday was our largest day ever for sales (by value).

Pre publicity
Aside from a couple of tweets through last week, the competition was mentioned in the Travelfish newsletter and the Facebook group. As the prizes are being paid for by us (as opposed to being donated by a company looking for exposure) there was no pressure on us to overly publicise what we were doing. In our view the prime reasons behind the competition were about giving something back to the community (yeah I know a lot of PR people are probably gagging reading that — sorry) — rather than getting out there beating the drum.

Thoughts on the competition in general
It was a lot of work
It was far more work to manage than I expected. While the twitter and forum quizzes and the major prize quizzes were pre-written, it was still time consuming to pick winners, close old threads and add new questions. This was complicated by a spammer who repeatedly hit the forum midway through the competition. Tracking winners in Twitter was particularly problematic (see Thoughts on Twitter below).

It was too complicated
Some people complained they didn’t realise the big prizes were being given away all day; or that there were too many different quiz threads. I forgot to close some threads (when we were being spammed) which led to people adding their answers despite my already awarding a winner on the thread. These answers then bounced the forum thread back to the top of the forum, and more people answered. It was messy.

A lot of window shoppers, not many buyers
Almost one thousand people read the thread about winning a $1000 travel voucher yet less than 50 entered – a 5% entry rate. That’s pretty dire. It’s not like they had to undergo a chromosome test to enter. Particularly as the east coast of the US woke up we say a spike in reads, but not more entries. Admittedly we’re a travel website for independent travellers, but you can see similar hit rates with the diving prizes. Overall we were under whelmed by the number of answers. What did do better was free stuff.

People love free stuff
To get a free Travelfish Guide, all people had to do was add an entry to the message board saying which guide they wanted. It doesn’t get any easier right? We saw a better response rate (albeit still pretty poor) for this part of the competition. Interestingly, the Vietnam pack was by far the most popular request — something not borne out by site usage. So either there’s something wonky with how people use the site, or when people are offered a range of goods for free they take the most expensive — regardless of if they need it or not! (The Vietnam pack, retailing for US$19.95 is our most expensive good for sale — most guides are around the $4.95 mark.

It was difficult to pick the right question
Ask an easy question it goes in five seconds, a hard one (or not even all that hard actually) and nobody knows the answer. Some of our readership are totally new to travel and Asia, others are more expert. It was more difficult to find a balance. For example an easy question, “Name Singapore’s international airport” was answered immediately, while a vaguely difficult one “Name two islands in Laos” was never answered. Finding the right balance was difficult.

Why on Sunday?
That’s just when the Travelfish birthday is. Sunday actually isn’t a bad day at all for traffic for us.

We’re based in Bali and some of our top readerships are in Australia, US, UK, Thailand and Singapore. As we didn’t want to let the competition run unwatched (a view vindicated by the spam attack) and I wasn’t up to staying up all night, we decided to start earlyish and finish late. This let people in North America and the UK/EU access the comp at the start and/or end of their day, while people in Asia and OZ/NZ got the most convenient viewing slots. There was never going to be a way to do this that worked for everybody and short of me deciding to stay up all night, I don’t see a better way to do it.

Some people struggle with simple instructions
The big prizes had a 30 word limit on the answer. Many entries went over 30 words. Our preferred entry for the $1000 coupon was for an entry that was over 30 words. Moral of the story? Enter a competition and follow the instructions and you’re already in with a good chance!

Thoughts on Twitter
The majority of the new site members came from Twitter. So, assuming we’re able to convert these new members into participating members longer term, the Twitter effort will be worthwhile.

While I use Twitterfox day to day, I used Tweetdeck to manage the competitions so that I could set up different columns to track the hashtags and mentions. There were a few problems with this.

a) When answering a quick quiz, people didn’t always use the hashtag. As a fictional example, a contestant may enter “Phnom Penh is the capital of Cambodia” as:

Phnom Penh is the capital of Cambodia
Phnom Penh is the capital of Cambodia #travelfish5
Phnom Penh is the capital of Cambodia #travelfish

@travelfish Phnom Penh is the capital of Cambodia
@travelfish Phnom Penh is the capital of Cambodia #travelfish5
@travelfish Phnom Penh is the capital of Cambodia #travelfish

d travelfish Phnom Penh is the capital of Cambodia
d travelfish Phnom Penh is the capital of Cambodia #travelfish5
d travelfish Phnom Penh is the capital of Cambodia #travelfish

Or they’d Direct Message me and one Twitterer (who I know personally) insisted on emailing me his answers each time!

b) There seems to be a lag between the live feed (the left hand default column in Tweetdeck) and the search results (whatever columns you set up to the right). People who had a standing search on a word — for example “Cambodia” would also answer, thinking it was just a question, not realising it was a competition.

c) This was particularly complicated by people who would direct message me answers as initially I wasn’t even tracking that!

The easiest way to get around the above it to either (i) specify exactly how an answer should be formatted (good luck in under 140 characters!) or (ii) steer all answers to the TF board for answering. While (ii) would have seen some drop off as people who wanted to answer wouldn’t want to register on TF to do so, from an answer management POV, it would be how I’d do it in the future.

d) For the Twitter questions, we wanted to give away two prizes on each question, so sticking with the above example, we would say something along the lines of “2nd and 5th correct answer get a prize”. In practise this didn’t really work all that well. Generally the first prize would go, but the second — especially if it was above the fifth reply — would drift unanswered. I put this down primarily to people following large numbers of people and not tracking the #travelfish5 tag. If I was to do this again, I’d run more individual quizzes but one prize apiece.

Would I do it again?
Yes, absolutely — but not for a year 😉

Any advice?
If you’re using Twitter for quiz questions and answers, set yourself up with some kind of a spreadsheet already loaded up with the questions. As the answers come in, add them in with the timestamp for each answer, along with the users TwitterID. If they’re a winner, follow them and direct message them telling them they’ve won and you’ll be in touch at the end of the day/next day with more details. Going back at the end of the day and dissecting all this information, which was my approach — was a real pain.

Choose a short hashtag — we should have used #tf5

If you’re posting URLs to promote the quiz midway through it, have one landing page and plug that and that alone. If you post different urls through the day, people retweeting may not use the URL you want — for example using a URL for a question that has already lapsed — and this is problematic. We switched landing page URLs midway through the competition and it was a big mistake.

In closing
It was fun and I’d do it again. Already some of the new members have been active since on the site and the winners were chuffed with their prizes — makes it all worthwhile!

Other experiences
Travelblogexchange has a dedicated discussion on the topic of running travel competitions.