Heath Ledger spikes my website, Day 3

Website traffic is starting to decline again although, as the graph shows, yesterday’s traffic was still the second-highest this month and the decline was relatively small.

Two things this morning. One, a look back at how much time this has taken so far, in response to a less-than-polite comment from Todd and a sensible question from Crispin. Two, some speculation about how the traffic might go from here.

The Heath Ledger Experiment is generating tons of interesting data for very little investment in time or money.

Once I’d heard via Twitter that Ledger was dead, the idea congealed in my mind while running an errand to the shops. After that, Day One went something like this:

task	time
Write web page	5 mins
Source & prep photo	5 mins
Figure out Google AdWords	30 mins
Run initial AdWords campaign	30 mins
Deal with 1st batch of comments	15 mins
Run second AdWords campaign	30 mins
Deal with 2nd batch of comments	15 mins

The set-up was incredibly fast because, well, it’s just a matter of writing words and pressing “Publish”. Google’s indexing robots hit my RSS feed hourly or so, so it’s not long before I’m indexed.

The most time-consuming part was setting up a Google AdWords account and figuring out how it all worked. Now that I know, next time will be quicker.

That table says it took 2+ hours, but that time was scattered across the working day. I was concentrating on other things, but came back to poke The Experiment — 5 minutes here, 5 minutes there. With a “real” campaign I’d set it up, let it run for a day, and then come back and analyse it.

You can probably add in another 2 hours’ worth of me talking at people and saying “Wow! This is so cool!” about 374 times.

I’ve probably spent 2 hours a day since then poring over the numbers and writing up the analysis. Again, in a “real” campaign I’d have a much more focussed process for evaluating the campaign’s effectiveness — and of course acting on the results. As I’ve said before, my sole aim here was to see how it works, not to specifically generate traffic or produce sales leads.

The potential research papers here, if that’s not too grand a description, are:

1. Lessons for small businesses wanting to try this for themselves.
2. How the rise in traffic to the jokes page unfolded, compared with when the story was covered in the media.
3. How the price of the search terms changed over time.
4. A linguistic analysis of the jokes and comments themselves.

“Research notes”, really. Any preferences for what you’d like to see first?

Now, the future.

I suspect that Friday’s drop in traffic was a weekly cycle thing, and that traffic will increase again across the weekend.

• Traffic was down on Friday generally across some other high-traffic sites we host.
• The younger people who are more likely to look for these jokes will have more time on their hands — both for word-of-mouth referrals and looking for amusements. This is especially true in Australia, where it’s a long weekend for Australia Day.

Any predictions from the floor?