Christchurch earthquake – web traffic to news sites
One of the unfortunate byproducts of working in the media is that disasters are often good for business – when something bad happens, people switch on the news, log on to news websites, and buy newspapers to find out what’s going on. The media often benefit indirectly from other people’s hardships.
But the Christchurch earthquake of Feb 22 was an exception for Fairfax NZ, the company I work for.
While lots of people came to Fairfax’s Stuff.co.nz news website and bought copies of Fairfax-owned newspapers, the company was especially badly hit by the quake. It lost a building and lots of equipment. More importantly, at the time of writing it had lost at least one colleague in the tragedy. No increase in sales can make up for that. When the dust settles the quake will be a huge loss for the company, like many others with offices in Christchurch.
With that in mind, let’s see what the numbers say about how NZ’s news websites performed during and since the quake. All numbers below are public data from Nielsen Netratings. These numbers are rough but give you an idea about the general trends.
Here’s the site traffic (NZ only visitors) for the weeks around the quake. Click the image for a full-sized version.
Now, YahooXtra has might you might call an unfair advantage in this graph, since it isn’t only a news website and reports traffic to its webmail here as well. But even so, Stuff nearly eclipsed it on quake Tuesday (just a few thousand UBs short).
Also note the difference between Stuff and NZHerald traffic in the week after the quake, vs the week before. Substantially more people were going to Stuff for quake news. About 80,000 new visitors came to Stuff on Monday, Feb 28 looking for quake news, vs about 50,000 for the Herald. TV3 had about 33,000 and TVNZ about 26,000 visitors above pre-quake averages that day.
Let’s look at a table of traffic spikes on quake Tuesday:
|% increase in UBs||562||6||68||72||177||247||4|
Now that’s interesting stuff. Stuff seems to do the best overall, recording the biggest lift in domestic visitors, with a 68% increase. The Herald, as usual, follows a similar trend but with smaller numbers. TVNZ and 3News record ginormous percentage lifts and respectable total rises too. But YahooXtra and MSN.co.nz? Pretty much nothing. That shows where Kiwis go for their news when disaster strikes.
Now, let’s see how many are sticking around. Here’s traffic before the quake and after things start calming down:
|Domestic UBs on Feb 21||21,239||500,980||323,548||267,278||85,009||11,910||292,824|
|Domestic Ubs on March 3||35,349||500,966||338,943||284,021||103,022||10,076||286,570|
|UB increase above pre-quake||16,545||-9,187||19,565||11,209||15,766||-2,720||-9,829|
|UB % increase above pre-quake||88||-2||6||4||18||-21||-3|
There are some patterns here:
* Stuff and the NZHerald haven’t kept many new visitors after the traffic spike, though Stuff did significantly better.
* YahooXtra, MSN.co.nz and RadioNZ are back to normal. (RadioNZ’s looks down, but fits about what you’d expect from eyeballing a pre-quake average. Its traffic on Feb 21 was higher than normal.)
* TV3 has gone gangbusters, nearly doubling its normal pre-quake traffic.
* TVNZ has done well too, winning more new visitors than the NZHerald, a nearly 20% increase in traffic.
So what’s going on here? Hard to say for sure, these top-level numbers can only hint at causes. But I suspect it’s about content. Take a look at the earthquake sections of Stuff, NZHerald, TVNZ, 3News, and RadioNZ. Then check out YahooXtra’s and Msn.co.nz’s. Oops, you can’t, because they didn’t have separate sections for quake news. Hmm, maybe Kiwis are right not to go there for major news events then.
RadioNZ’s section is just boring list of recent stories, no multimedia or maps or Twitter widgets or extra info. No surprise really it didn’t manage to keep many new visitors either.
By the numbers, Stuff got the best result from its quake coverage. It had the highest raw increase in Domestic UBs – the key currency to attract advertisers – and cemented its increasingly dominant position in NZ online news.
But the TV stations were surprise hits, especially TV3. Lots of new visitors discovered their sites and seem to be sticking around. They’re still much smaller than Stuff or the Herald, especially TV3, but it shows that NZ’s TV websites can do a decent job of online news too. Before the quake I and many others had our doubts.
It helped of course that disasters make for great visuals – Stuff’s videos also did really well last week – but both TV stations made good Christchurch earthquake sections full of useful and interesting text as well. That’s especially true of TV3, and it looks like the site has won lots of new regular visitors as a result.
Homepage traffic is a good indicator of interest in the site. Casual traffic rarely goes to the homepage, but speicific stories instead, and regular readers are really what we’re after.
Here’s the UBs on the homepage of Stuff and the NZHerald. Click the image for a full-sized version.
Between Monday Feb 20 and Tuesday Feb 21, that’s an increase of about 60% for Stuff and 63% for the Herald. Homepage page impressions went up 73% and 80% respectively. So we’re talking about 75% more homepage loads than a normal day.
While the Herald went up more as a proportion of its pre-quake traffic, Stuff actually gained more homepage UBs (144,000 vs 116,000) and PIs (764,000 vs 606,000).
Stuff’s web traffic might have been slightly higher on Tuesday Feb 21, but at the time of the quake we were hit with technical troubles (not entirely because of the quake, but made worse by the traffic surge).
Now let’s look at traffic after the quake surge died down a bit:
|UBs on Feb 21||239412||184607|
|UBs on March 3||252544||196472|
|UB raw increase||10817||14564|
|UB % increase||4||8|
Stuff’s homepage is still significantly more popular, but the Herald won more new regular visitors to its homepage as a result of the quake. I think next week we’ll see that even out a bit more.
It’s interesting because Stuff is getting more regular visitors to its website than the Herald, but less to its homepage. This could be because quake traffic on Stuff is bypassing the homepage and going straight to its Christchurch earthquake section, while nearly all new Herald visitors start at the homepage.
Final note: The NZ Herald site actually did better than Stuff when you include international visitors – probably because the Herald site has better SEO – though international visitors are normally not counted by NZ advertisers so aren’t by anyone else either.
But the Herald also lost almost all of these new visitors by March 3, while Stuff has actually kept most of the international visitors it attracted with its quake coverage. I’d call that an international traffic short-term win for the Herald, but a long-term one for Stuff.