Audience measurement for web 2.0

  • Jon Lund 

Yesterday I had the pleasure of giving a presentation for some one hundred internet professionals at the “How to capture the customers of web 2.0”- conference in Bratislava.

I called the presentation “audience measurement for web 2.0”. Here it is:

Audience measurement for web 2.0
Thank you for the introduction.

And hello everybody.

I’m very pleased to be able to be here today to share some of the experiences from the online industry of Denmark.

Somethings happening in the state of Denmark
The case is, that in just these days and hours something great is going on in Denmark…

At all the major Danish websites project managers are following up on the work of programmers. Sales departments are rehearsing how to communicate to their customers.
Contentproviders are adjusting all they thought they knew about what users really liked about their content.

Why?

Because at the end of June, only a month from now, a brand new way of measuring the audience of the Danish websites are taking effect.

A way of measuring that will bring with it a whole new vocabulary.
A whole new way of speaking about traffic.
A whole new way of thinking advertising on the internet.

A way  of measurement that for the first time will give a truly realistic picture of how many people are actually visting Danish web sites. And who these people are.

A way of measurement that will harness the powers of the online medium. The powers of  truly being able to document and measure its actual use. Who is being hit by an advertising campaign. And how many are being. While at the same time leveraging the vocabulary used by marketers.

The new audience measurement in Denmark is conducted by the internet research company Gemius. They were chosen to do so in competion with other research companies like TNS, who does the measurement for both Internet and all other mediagroups in Denmark today. In competion with Nielsen/Netratings, the worlds leading online measurement company. In competion with Comscore, the company leading panel based measurement in the US. And in competion with several others. Based on a thourough Requirement Specification Gemius won this tender. And did so in a very convincing way.

This new online measurement system is what I’d like to elaborate over for the next 20 minutes. Before that, however, I’d like to introduce myself and highlight a few things you should know about Denmark.

This is me (or at least my blog):

(Full size picture click here)

And this is who I am:


(Full size picture click here)

Denmark
This is Denmark:


(Full size click here)

The first reason
I said: something great is going to happen in Denmark: a brand new way of measuring the audience of the Danish websites takes effect at the end of this month.

There are several reasons why this is particularly great. The first one is this one:

After years of non-sense technical geek jargon, light will be spread on the Danish Internet market. This will make the internet more attractive to traditional advertisers. And it will lead to more money being spend on the internet.

Why’s that? What hasn’t been clear until now? What’s wrong with the present state of things?

I’ll tell you.

The money talk
Are any on you guys able to tell me what a “hit” is?
What logfiles are?
IP numbers?
Cookies?
“Unique Visitors”?

Maybe you are.

Or will be.

Or at least would be able to learn.

Like you some of the people with the marketing money does.

Some.

That is: some ONLINE mediaplanners.

A very few TV- or printplanners.

And nearly no advertisers at all.

What they know is consumers.

Targetgroups.

Men, women, young ones, old ones. With wife and kids. Or living alone. Well educated – or without education. In one business – or another.

That is: Real people with real money.

The advertising money is not attracted to hits. Not to IP-numbers. Not to logfiles. Not to cookies or “Unique Visitors”.

Nope.

The advertising money goes where their effect is guaranteed. To the media, that delivers the right target group.

The advertising money knows concepts like Gross Rating Points – GRP. And – particularly – Target Rating Point – TRP.

They know these concepts from the traditional advertising world. Which is where they spend nearly all of their money.

And until the day, where internet media start talking that language, they’ll never fullfill their potential. Never seriously be able to capture the advertising money, that really belongs to online media.

This is the problem. And this is why the new audience measurement in Denmark is so highly anticipated. For this is exactly what the new measurement proposes: a new vocabulary. New terms. New words: Real users. Target groups. Affinity. GRP. TRP.

Dangerously unstable
There is a second reason also, however, for looking forwards to the new measurement system to be introduced in Denmark. The case is that the current way we measure in Denmark is unstable.

It’s the same method that’s being used throughout the world. It’s the method employed by TNS Gallup. By Nielsen/Netratings. By Google Analytics. By all parties who conduct so-called sitecentric measurement today.

The method employed by all these parties is unstable. Unstable in a dangerous way. Unstable because the numbers it provides us with does not correspond to the facts of the world. Because it claims that the largest Danish websites have more users than the number of people living in Denmark. Because it thereby threathens to undermine the trust in the figures provided to the market.

“Can we trust this?” people might ask. With good reason.
 
For every day passing by, the trust in our currency is therefore being challenged. For every day passing by, more and more advertising money which should have been online, finds other places to go.

—-

Fuelling a success
When the end of this month draws to a close, the results from the new measurement in Denmark, will hit the streets. And it is – as I’ve said a few times – much anticipated. Like everything online is, these days.

But why are people so passionate about online? Why do people care so much about online and how it’s measured? I think the reason is the feeling that the internet has enormous potentials, that has not been unleashed yet. That vast amounts of money are bound to go there.

These feelings have some solid evidence to build on.

Consider these graphs:


(Full size click here) Index of the number of visits at the largest Danish websites. Source: FDIMs Danish Internet Index


(Full size click here) The amount actually received by Danish online publishers. Source FDIM/Deloitte. For more on this have a look at the FDIM Website of adspend measurement (in Danish) and (in English) Best ever national internet ad spend measurement launched in Denmark.

 


(Full size click here) Advertising revenues, 2006. Source FDIM/Deloitte. I’ve written more about this here: Danish 2006 internet advertising revenues surprisingly close to US figures.

(Full size click here) Total advertising revenues in Denmark 2006. Estimated from 2005-adspend survey (Reklameforbrugsundersøgelsen) by Dansk Oplagskontrol. Final official data expected later this summer. Check for updates here (Danish).

The takeover of the online world
These two trends – the historic growth in traffic AND in internet adspend – are part of the reason why online is so much appreciated these days: everybody are online, surfing. Lots of admoney are already spend there.

But there’s more to this end also. For where is this leading? How large a mediagroup can online grow to be?

One radical view was presented by Steve Ballmer, CEO of Microsoft, a few weeks ago at a conference in Copenhagen. In ten years, he said, all reading and all writing will be online
… no more paper newspapers
… no more radio “in the air”
… no more broadcast TV
… no more video or dvds
… no more books

All media will be online
And
All advertising will be online

Even if Steve Ballmer might be overexagerating a little, the trend is clear:

Internet advertising will continue to grow.

And this is why we see such heat in the discussion of online.

(I’ve written more on the Steve Ballmer presentation here: Steve Ballmer to me (and a few others): ALL advertising to be online in ten years)

In sum
Up until now I’ve presented you with reasons why traditional measurement of online media doesn’t work. Because the words, the terms, the vocabulary surrounding the old way of measuring is counterproductive to having more money spend online. Also I’ve explained the fragility of the old measurement. How it threathens to blow up in our hands, if we don’t find a better way of measuring; a way that corresponds better to reality.

I’ve also given you historical reasons why online is important. I’ve shown you how more and more people surf for a longer and longer time. And how more and more advertising money are being spend online.

And I’ve hinted, that online will keep growing.

Now – what is it exactly that the new Danish measurement can do, to help the success of online to keep growing in the future? What is the measurement about? How does it do it? This I’d like to dwell over for a moment.

Defining the user
The cookie, the “unique visitor” doesn’t tell you much about how many actual users you have. Not even if do everything that’s technically possible, you’ll get a correct number.

However, if you do a proper job in correcting and adjusting the cookie-figure you will get a number that can tell you something else. You get a number with which it makes sense to compare yourself to others. A number with which you can tell: this site is twice the size of that site. Or: this site has only 60 percent of the traffic of that site.

And more than this: When you know how many cookies there are on all websites in a county, like Denmark, you’re able to tell how large a percentage of all the Danish surfers you have on a given site.

The preliminary test-data from the month of March suggests that 48,17 percent of all Danes visited dr.dk, the website of the Danish public service TV-channel. While on the other hand, a very small niche site, like the one of my own organisation, FDIM, only had 0,19 percent of the online Danes stopping by.

Getting the universe right
Using the cookie-figures in this way, all that’s needed is to find out, is how many Danes are actually surfing the net. And this is – if not a trivial task – then at least a rather straight forward one. How do you do it? You ask! You simply phone up those 1000-1500 people that makes up a representative sample and ask them: Have you surfed the internet today? Yesterday? In the past week? The past month? Or haven’t you been surfing the internet at all.

So this is what Gemius does: once a month they have a regular Danish research company asking exactly these questions. In March they found, that exactly 3,546,189 Danes above 15 years of age surfed the internet.

The rest, then, is simple math:
48,17 percent of 3,546,189 Danes, equalling 1,708,046 Danes, surfed dr.dk in March
0,19 percent of 3,546,189 Danes, equalling 6730 Danes, surfed fdim.dk in March


Telling us who they are
Being able to tell how many Danes actually surf a given site is, however, not the same as being able to say who these surfers are. In order to do so, Gemius employs a third source of data: the popup panel.

All the websites taking part of the survey, asks a randomnly selected sample of their users to participate in the panel. Asking them questions like: how old are you? Are you male or female? Where do you live? What’s your education? Etc.

These panellist not only answer questions about who they are. They also consent to let Gemius register what sites they’re visiting. They say: “Yes, it’s all right: feel free to monitor what of the sites in the survey I visit”. Thereby enabling Gemius to set the exact profile the users of the different sites connected to the system.

In order to do this in a credible way, however, the panel has to be very large. And so it is! Because it’s one central survey, the panellists gathered at one site also works as panellists on all other sites: together all the sites builds up a huge panel. A panel which in the month of March reached 91.615 Danes. Nearly a hundred thousand!

Using the system
The results of all this you see here:

 
(Full size click here)

On the left you can choose what sites to analyze. I’ve selected dr.dk and fdim.dk.

Accessing the central statistics on how the population as a whole has used the sites in question is not the end of the story however. Actually this is where it starts.

Try clicking the small icon in the top that looks like a target – that opens up the target group selection window:


(Full size click here)

From this one you can define what users you’d like a closer look at. Here, I have chosen women aged 25-34:


(Full size click here)

Clicking “OK” now gives me the stats on the use of dr.dk and fdim.dk by these young women:
 


(Full size click here)

What the system tells us here is that of the 1.708.046 Danes that used dr.dk in March 190.340 were in our target group (while the same applies to 921 of the total 6730 users on fdim.dk)

Note the “affinity index” column! It says that fdim.dk actually has more young female users than the average site in the survey, whereas dr.dk is extremely close to the norm.

One could off course spend much more time going through all the features and possibilities of the system. Actually the system would deserve not just showing two sites side by side like this. On the contrary: to show the strengths of the system, I should rather have showed how a mediaplanner would use the tool if he were to run a automotive campaing targeted to men in the agespan of 23 to 45. What sites should he use to hit this target group in the most economic effective way? What is the cheapest way to reach the potential buyers of a new – lets say – VolksWagen, Audi or Volvo? This however you’ll have to ask someone else to do!

Back in Denmark project managers are still trying to follow up on the work of their programmers.

Sales departments are still rehearsing how to communicate the new Real Users-terms to their customers.

Contentproviders are still trying to get to grips with what of their content their Real Users really likes.

Now you know why.

A final note: On the airplane on my way here, I read through todays programme. At first I thought: why am I doing a presentation on Audience measurement on a conference dedicated to Web 2.0? But after a while it suddenly became clear to me.

Some argues that web 2.0 is about design – you know gradient colours, logos with dropshadow and a little Beta-mark, saying “this site is never finished, always being developed”.
And true, many modern 2.0 websites looks like this.

Others argues Web 2.0 about usability. Keep your site focused! Keep it simple. Only have three buttons which you can push! And navigate by tags-clouds and search only.
Also this is true. Simplicity is a virtue these days.

But the thing that distinguishes Web 2.0 the most, I think, is none of those. To me, Web 2.0 is about empowerment of the people. It’s about how everyone can create their own video and upload it to youtube.com. Its about how anyone can record a song, and share it with the world at myspace.com. It’s about how noone has to keep quiet if they have something to say. Just write it down, and post it to your blog!

In other words: the fundamental thing about web 2.0 is that the powers have been taken away from the webmasters, the sitemanagers, the producers and the editors. And the powers are instead transferred to ordinary people. What once were masses are now seen – and acts! – as individuals.

And in that sense what I’ve presented about measurement of Real Users, hits Web 2.0 spot on. It’s a measurement for the modern times.

Thank you for listening!

0 tanker om “Audience measurement for web 2.0”

  1. Hej Jon

    Thanks for a very interesting presentation of how you will measure web metrics in the future.

    Am I right in my understanding that the new approach is adding a qualitative approach to the traditional quantative figures?

    Do you think smaller companies will also in the future have to survey their users regularly, like your members are doing now?

    I would also be interested in knowing whether you have addressed the whole question of mash-ups in web 2.0; ie. that what users get onto their screen will increasingly be a melting pot of content from a number of different sites – Jeff Jarvis has an interesting discussion of this in his weekly column for The Guardian:
    http://media.guardian.co.uk/mediaguardian/story/0,,2089337,00.html (it’s free to register for this site)

  2. Hi Jacob,

    Yep, you’re right: this way of doing things ads a qualitative aspect (who are the users) to the pure quantitative (how many are they)

    It’s relatively cheap for small sites to join the full survey. It costs from some 1250 euro a year to be included in the full measurement within the FDIMsetup (then you also needs to be an FDIMmember which has its own fee).

    The question of how to measure sites based on AJAX etc is very interesting. It brings into play questions like: “what is a pageview?” or “can you track content depending on where it’s shown?”. For all I know there are no simple answers to these questions yet – but it’s definitively an issue we’ll have to look more into in the time to come.

    Best

    Jon

  3. Pingback: New Media Trends » Audience measurement - Giant leap forward - now we’re in the race

  4. Pingback: New Media Trends » Podcast: How the new Danish audience measurement works