The scramble for eyeballs

What is the most important metric for a digital publisher?

One has told me that “the currency we trade in” (which I think, in saying ‘We’ he meant both his company and mine) is page impressions.

The other figure tossed around is unique users/browser – which is a pretty cringeworthy term we all use but really shouldn’t.

It makes sense I guess. TV networks main currency is viewers and magazines/print’s main currency is readers.

I know that TV audiences are based on an average figure – so there is an element of quality control in the numbers. With print and magazines I’m not so sure but I don’t think there’s a quality component (ie you have to be a regular reader, or purchase or read once every x days/weeks).

With digital maybe it’s time we started trading in deeper data than just page impressions and users.


Well – both are reasonably easy to game and don’t give anyone a particularly good idea of the quality of users and the quality of the media brand.

Page Impressions can, and are, gamed every day. Slideshows are one way to do this. Articles over multiple pages is another. Clunky UI is another. There is almost an incentive to make the user jump through a few hoops – it increases the amount of inventory one can sell.

Case in point – someone can spend 1 minute on a news site and watch a 40 slide slide show. This user will register 40 page impressions … which seems like they’re incredibly engaged … the reality is they’re not. They spent a minute on the site. The problem with the data is right now you can’t distinguish.

Unique Users is also a funny one. There’s no distinguishing between a user who views one page by accident and never returns to the property … and a user who returns daily and views 100’s of pages. One could argue that if a user stays for less than a minute and views 1-2 pages and never returns that they’re not really a ‘user’ … they are just someone who has stumbled in by accident, not found what they’re after and left never to return.

So when you hear spin like ‘3.8m Australian’s turn to us every month for the news they need’ the reality is often a lot different.

How often are they turning to you? How many never return? How long are they on your site for?

It’s a lot different than 1.5m people watching a new bulletin nightly for 30 minutes. Or the Herald Sun paper daily without fail and reading it over breakfast for 30 mins. Let’s not kid ourselves.

Personally I think it’s weird that there is such dedication to boosting page impressions when most publishers can’t sell the inventory they have at the price they’d like to – and large chunks (30-50%) are sold via low cost remnant arrangements.

Tony Faure on this blog almost a year ago was asking for “less quant, more qual” measurement and I agree with him.

Qualitative measures are a way for suppliers to illustrate their point of difference, because everyone has big numbers around PIs and UBs so ultimately the huge amounts of supply will ensure the CPMs continue to decrease as demand isn’t keeping up.

Anyway – when Page Impressions and Unique Browsers are your key currency it surely impacts the end product as the incentive is to boost traffic and page views regardless of the quality of these. If, at the end of the month, all you’re looking at is PVs and UB’s, you’re missing out on the quality metrics which distinguish what a media brand can offer.

If you look at the top 3 news providers in AU they’re all effectively the same.

They have a similar audience at the end of the day. They have similar loyalty. They have comparable content. They all have hundreds of millions of pageviews.

Would you pay any more for one than the other – probably not. Why? Because they’re pretty much the same thing with a different header. They don’t communicate enough about the qualitative elements that make their audience any different to that of their competitors. Either they don’t know … or they do and they’re not sharing it … or they can’t communicate it simply.

So … can we move beyond these metrics? What do they really mean? I can’t see the story they’re trying to sell unless we add some further data around quality.

And without these quality metrics I can’t see yield increasing any time soon – for that to happen advertisers need way more clarity around what they’re getting. Not smoke and mirrors.


8 responses to “The scramble for eyeballs

  1. Very interesting topic Ben.

    I think brand metrics lies at the heart of this debate i.e. how website publishers go about measuring their brand equity.

    While basic traffic figures indicate popularity, they don’t tell you much about the strength of a website’s brand (e.g. awareness, associations).

    So how about looking at other metrics, such as the Net Promoter Score i.e. how likely a visitor would be to recommend the website to others.

    That would give you an advocacy metric …

    And then there’s “engagement”. How about creating an “engagement index” for websites, based on more than average session duration.

    That would help qualify strength of connection …

    There’s no shortage of metrics out there – the trick is to find the most meaningful ones.

  2. Page impressions figures also need to be ignored on sites like SMH / Sydney Morning Herald which auto-refreshes every 5 minutes.

    If a user leaves their site open in a window but doesnt look at it whilst working, say 9am to 5pm, it will count 96 page impressions which no one actually looked at.

  3. Ben,

    100% agree with your stance on the need for a rethink on success metrics. Im routinely amazed at how much time and effort goes into driving up traffic on a site so that a publisher can claim to be #1 in its category, but how little though and effort goes into engaging the target audience and ensuring that they come back again following a good user experience with said publisher.

    I think that mainstream publishers getting caught up in a race to attract and retain the ‘largest’ audience (based on UBs and PIs ) is a no-win game, especially when for the most part news audiences are generally brand agnostic (or simply disloyal), but it’ll most likely be a teir 3 or 3 publisher that will come up with a successful Total Time on Site formula to attract, engage and retain the right (not just the biggest) target audience that will cause a shift that may well lead the pack in a new direction.

  4. Good stuff Ben.
    Communciation and marketing is all about people, not pages. So the number of people, how often they visit site, and how long they stay are critical (certainly more than how many pages they view). And otehr emasures like engagement and word of motth as Adam mentioned are also valuable metrics. the challenge is for publishers to effectively communciate the value of their community a s part of their media sales process. Unfortunately many don’t do this, and resort to boasting the highest number thay can claim (usually PIs), which then becomes a commodity against competitors.
    And that ultimately reduces prices for all.
    But we must keep working on a wider range of metrics that can paint a more complete and richer understanding of a community,and their relative value to advertisers.

  5. Good post Ben. I am constantly asking the same questions and hope that other metrics that already exist can move up the decision hierarchy.

    Nielsen provide heaps of data that tells the engagement story like page and session duration and frequency and also tools like Duplication to see where sites are getting there traffic from. But from a quick canvass of our sales team, we are hardly ever asked to provide anything more than pages and UB’s. We provide a lot more than that but I’m just not sure enough buyers really care.

    Internally we look at other metrics to assess engagement on our sites. A key metric for us is registered members and posts in our forum. These stats are usually displayed in forums for all to view and we look at other competitor sites to get a feel for the engagement they have with their audience.

    inthemix for example has 155,000 registered members and has amassed 6.8M posts in its forum. Our next biggest competitor has 768 members and 1400 posts its forum. The numbers still tell a story, a compelling story about the two brands and how each is cutting through to its audience.

    The problem is that on the surface the competitor has good stats. Good SEO, SEM and passing traffic around a network results in high UB’s and pages. As you rightly said, unfortunately that’s the only currency being traded.

  6. What I’m hearing mainly from agencies and starting to hear form a few advertisers is that they want simple quant metrics of reach & frequency – they know this metric, it is a currency that can be used for trading and it is comparable with TV and other media.

    Supplementing this simple set of people-based data are an array of (mostly) site analytics that sites can use to build their case for greater share of the (growing) advertising pie – qual (with some quant also!)

    I’m not sure if the industry is interested in standardising site analytics metrics but would be open to views on this – if yes, we can move into this once we have sorted out the basics!

  7. talkingdigital

    hey neil – the point around forums is very good.

    i see many sites claim to have a massively ‘engaged’ community in terms of numbers and repeat traffic – ie selling a community of people … when in reality they have a lot of in/out users … about 1% of whom have signed up for the forums and the forum has virtually no activity a month.

    if you want to see if a community is engaged – go onto the forums and see the volume of posts in the last 24 hours. if there’s a handful they’re not really engaged because if you see a site like itm or whirlpool there’s literally thousands of posts being updated every day if not more.

    astounds me itm forums have managed to keep momentum for almost 10 years now. I first used the site in 2000 and it was buzzing then. 9+ years on it still is … it’s just the originals are all old and boring heh

  8. Hey kids

    Large scale metrics have always been controlled by the big publishers – most of them are print based companies so I guess they took their lead from their experience aggressively managing circulation audit metrics.

    This metrics story has been the same since the first australian banner ad was sold.

    Good digital media planners know the numbers they need and know its nearly always been about execution metrics not publisher metrics. If your planner doesn’t know the numbers then get one who does.

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 )

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s