Skip to content

Rugby World Cup web traffic

October 29, 2011

It was the largest sporting event ever in New Zealand. How did the country’s media sites do? Let’s take a meander through the metrics.Silver fern coursty of

The short answer is “not as well as you’d think”. While the sites’ sport sections did quite well, sites overall didn’t really budge much from normal.

You can see that in these category report numbers showing average daily UBs  (all metrics from Nielsen NetRatings, NZ traffic only – and the usual disclaimer, I work for Fairfax NZ, publisher of the website).

Newspapers & Magazines Publisher Home Pages Sport
Average RWC week traffic 545,942 1,053,571 353,585
Change from previous average 24,654 7,989 106,403
% change from previous average 4.7% 0.8% 43.0%

Wow. The Sport category went bloomers, nearly doubling in traffic. But homepage traffic? About the same as before (Stuff and NZHerald tended to perform the best here, recording low but significant percent increases). And traffic to news sites overall only budged a bit.

So it looks like news sites as a whole didn’t actually win many more regular readers as a result of the RWC, unlike in other major news events. But their regular readers were reading a ton more sport stories.

Average Daily UBs have their sins, like any metric. But in general they’re a good indicator of regular audience. And if you run the same numbers looking instead at UBs, PIs, Sessions, UB Freq and Average Session Duration, you see exactly the same trend – negligible change on the category as a whole, but a big lift for Sport.

Let’s look at the major media sites now, still average daily UBs.
Average RWC week traffic 12,967 24,321 2,577 4,109 336,027 269,507 97,117
Change from previous average -351 2,010 799 -694 15,275 18,008 1,782
% change from previous average -2.6% 9.0% 44.9% -14.4% 4.8% 7.2% 1.9%
Average RWC week traffic 458,623 248,658
Change from previous average -7,549 4,913
% change from previous average -1.6% 2.0%

Now things start to get even more fascinatingly interesting.

First off, it looks like RadioSport was cannabilising traffic to its sister site, NewstalkZB. Both are pretty tiny though, so could be swung easily by one-off spikes and dips.

Now, ODT, WTF? The country holds a massive sporting event, and your regular audience  goes down? It actually gets worse than that for the ODT, as we’ll see when we look at only sport traffic in a sec. Yahoo is in a similar boat.

As for TVNZ and MSN – going up only a smidge during an event like the RWC isn’t much to be happy about either. If they were my companies, I’d expect more.

I have to admit I’m a wee bit disappointed in 3News’ traffic. They’ve seen such big spikes in past events. For the RWC they had a nice increase, but off a small base (compare the raw change across the row). Mind you, TV3 still had a bigger raw increase than TVNZ…

Interesting too that the NZHerald beat out Stuff in raw increase in ADUBs. That’s the first time in a while they’ve succeeded at doing that. If NZHerald keeps up the new higher level of ADUBs, they would have won a lot of new regular readers. And that is what this digital media thing is all about.

Enough teasing, let’s look now at traffic just to Sport sections, again at ADUBs.
Average RWC week traffic 1,595 7,294 76,432 20,967 142,600 99,635
Change from previous average -792 3,977 21,017 9,843 54,208 41,843
% change from previous average -33.2% 119.9% 37.9% 88.5% 61.3% 72.4%

Now things start to look a bit better, but ouch, ODT, WTF all over again. It’s one thing to hold a major sporting event and see a site’s traffic drop. But to see your sport section drop? That’s a whole new bag of sucks-to-be-you. The drop is consistent across the RWC weeks and starts suddenly in the first week. They’re doing something wrong.

They could explain this if the RWC traffic wasn’t tracking in their sport section, but if that’s true, it seems like a pretty big boo-boo. Advertisers are going to look at their sport section traffic without knowing those kinds of details.

Moving on, 3News is looking its old spikey self again, recording a massive 120% increase for the RWC weeks. Still, that’s off a very small base, and starts to look measley even when compared to arch-rival TVNZ’s modest increase of 10,000.

Yahoo recorded a sizeable increase, but if I were them I couldn’t help but be disappointed with it. With overall site traffic down, this seems to suggest people were finding RWC elsewhere, and those that stuck around weren’t that interested in sport anyway – or at least,  Yahoo’s RWC coverage was a bit lacking.

MSN doesn’t report a separate sport section, since they’ve basically outsourced their news to the NZHerald now.

Stuff is the hands down winner here. A big increase off a big base, and if I were them I wouldn’t be too worried that the NZHerald had a higher percent increase. It’s the 40,000 ADUB gap with the NZHerald I’d crow about.

Taking the sitewide and sport section traffic numbers together now, we see the trend we spotted at the start coming back again for some sites. Yahoo and MSN didn’t win much traffic from the RWC, though Yahoo’s sport section performed about what you’d expect if you weren’t expecting too much. 3News and TVNZ didn’t win many new regular readers, but their RWC coverage did lure a lot of the existing readers into their sport sections.

Stuff and the NZHerald both recorded nice increases sitewide in regular traffic, but most of the growth in their regular sport audience was actually from their existing readers.

So all in all, the RWC doesn’t seem to have affected NZers digital media habits much at all, and it certainly wasn’t a lure for people to flock to the country’s news website. That’ll be especially true if Stuff and the NZHerald lose that regular sitewide audience they gained from the event.

Ok, on the home stretch now. Let’s quickly look at a common engagement, average session duration (measured here in seconds). This is sitewide.
Average RWC week traffic 178 221 102 104 325 347 264
Change from previous average -1 2 -2 -2 16 21 -7
% change from previous average -0.7% 0.8% -2.3% -1.8% 5.2% 6.3% -2.7%
Average RWC week traffic 287 78
Change from previous average 11 1
% change from previous average 3.8% 1.7%

Hmm, ok. Yahoo actually does ok here, recording a 4% increase. That’s actually quite good for ASD, which usually doesn’t change much at all.

ASD going down is often not a bad thing. It may just mean more people coming to your site to read one hot story. But when ASD goes down or doesn’t change, and site traffic goes down (I’m looking at you, ODT), that’s a cause for concern.

But it is interesting that Stuff and NZHerald, the two sites who did best from the RWC in terms of raw traffic, also did the best here. A 5% increase in ASD with a jump in traffic is a great sign that your content is engaging and readers like it. Some of this will likely be more people watching videos, but a lot will be people reading more articles.

UB Frequency (roughly speaking, how many times people come, in this case weekly) shows the same trends, though Yahoo strangely does quite well here (8% increase). This is related to RWC traffic but seems to be fed from elsewhere. It may just be people checking their email more often or something like that.

The NZHerald and Stuff swap places in UB Frequency.
Average RWC week traffic 4.08 3.60
Change from previous average 0.13 0.10
% change from previous average 3.2% 2.7%

So people are coming less often to the NZHerald, but reading more things. That may be more casual readers coming since the NZHerald tends to do better here, but the fact they’re consuming lots of content is good news for APN.

Finally, let’s bring things to a close by looking at daily traffic around the final big game. This is UBs on the Sunday and Monday, vs a typical Sunday and Monday in August.
Sun, Oct 23 8281 17830 229660 202018 77209 456266 248237
Raw change, Aug 28-Oct 24 (Sunday) 4,804 14,818 5,877 -6,177 22,496 -14,730 -2,071
% change, Aug 28-Oct 24 (Sunday) 63.7% 76.2% 2.3% -2.7% 24.7% -3.1% -0.8%
Mon, Oct 24 16982 39361 311759 272996 116468 452464 246177
Raw change, Aug 29-Oct 24 (Monday) 2,068 10,840 -45,921 -8,436 20,696 -16,040 2,368
% change, Aug 29-Oct 24 (Monday) 13.9% 38.0% -12.8% -3.0% 21.6% -3.4% 1.0%

Goobers! What’s going on? The finale of the country’s biggest sporting event, and web traffic to news sites is lacklustre. Though the ODT, for once, has something to be happy about.

3News does particular well on the day of the final, possibly from people searching for clips. I was otherwise engaged at the time, so can’t say for sure.

It’s the NZHerald and Stuff thought that I’m interested in. They don’t see much change at all for the event itself, and the day after saw a traffic drop. That drop is because Monday was of course a public holiday, which means no one was at work. So that would suggest Stuff gets a higher proportion of its traffic from people at work, vs the Herald, and is therefore more dependent on workers.

Still, traffic on Monday was higher than it typically is on a day off. So more people were coming in to read the post-analysis than usually appear on public holidays.

Plus, we’re seeing something similar to what I pointed out during the Royal Wedding – if something’s on TV, people will watch it there instead of getting their news online. They’ll go online before and after the event, but during it, forget about it. We can’t see hourly traffic in NetRatings, but I’d bet dollars-to-doritos you’d see a big empty blotch during the game. It’s even worse in for the NZHerald in Auckland, of course, because lots of people were at the game, at parties, at Fan Central, or in my case at home alone trying to learn SQL.

3 Comments leave one →
  1. October 30, 2011 1:01 am

    Hi Reuben,

    Nice post. I’m assuming your data came from NetRatings?

    How accurate is that compared to say, hypothetically if the information were publicly available, comparing Google Analytics statistics for both sites?


    • October 30, 2011 5:06 pm

      Yeah , all data from net ratings market intelligence.

      Daily UBs generally match pretty well between the different measurement systems I’ve used (GA, nielsen and omniture).Visits/sessions too.

      In a perfect world we’d all just share all our private data and everyone’s site would be the better for it. But where’s the fun that?


  1. Major news events’ effect on web traffic « MediaMaths

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: