I suspect that downward pressure on agency’s costs and margins have led buyers to demand lower rates and only pay for what they get through performance based media buys.

Of course there will be those out there who push the “buy a truckload of impressions on high volume sites” barrow, taking this back to traditional media (and I’ve worked across both traditional and digital media over the last 10 years) this can be equated to midnight to dawn TV or run of book press buys. Traditional media combat this theory by restricting low rates to their less attractive inventory. When you can buy media at $1 per click that is effectively the same (or similar in most ways) to so-called “premium” inventory (sold at $25cpm) why the hell would anyone in their right mind pay $25cpm??

Publishers are shooting themselves in the foot by offering their audience up at these CPC/CPA rates. If the audience is valuable enough to an advertiser they will pay what they have to pay in order to reach them (within reason) in quality, targeted environments.

If publishers want to play in the CPC/CPA space and advertisers utilise this to their benefit, great, but as an industry, digital media (publishers and agencies) needs to do more to convince advertisers of the inherent branding value of digital media… or demand that advertisers take all the branding off their performance ads… and there’s an argument which has gone on since OzEmail were selling banner ads!

with ad networks, I’d love to know how much scrutiny agencies take in evaluating their ‘technology’ … ie their claims they can reach the right person at the right time etc. The world of behavioural targeting/retargeting/remessaging seems like a bit of a black art at times … maybe it’s basic blind buy networks polishing up the turd they sell with a bit of technbable?

There are huge differences between a user that falls on a Fairfax or news article via a search query or a google ad, compared to a user that intentionally seeks out a quality niche site and consumes their content with great interest (yes my company is a niche publisher). The differences are evident in CTR and engagement time and I doubt anyone can argue that. So to evaluate the true value of an impression the metrics need to go beyond engagement , far beyond CTR and look pre and post impression activity.

Media agents at the end of the day need to be demanding these extra metric’s to ensure they are buying the best quality inventory and the large ad networks and top 5 need to play fair and work with the local publishers on these metrics. Let’s face it, some of the big ad networks have it very bloody easy…they sign up a few large US or UK sites, geo target their AU ads and flog impressions at the lowest possible rate as any revenue from down under is a bonus to the US head office who have already been down this path and think $15CPM is absolute gold. These networks have no need to create a brand or any technology and only some create a little local content using the US HQ’s content systems.

So perhaps the media agents and advertisers need to think Australian Made as well and stop working with the networks who are shipping 60-75% of these ad dollars overseas.

Some other good reading here: http://www.techcrunch.com/2009.....l-the-cpm/

Online is not the silver bullet and yes TV, Radio, Print etc have been around for years for a reason; the audience is there and proven to be too(.)

Some mediums are beginning to be scrutised (Yellow Pages Print) – but I wont go there because last time War and Peace was written in 22 mins after the Mumbrella email was sent out…

Sites catering for the user is mega important. Sometimes publishers can be arrogant and ignorant, assuming that they have got it right and that they know their audience. Websites need to be tested and audiences need to be consulted regularly. What would you like from my website? How can I improve the site for you? Give the audience what they want and they will come back in droves, which creates a good platform for advertisers. It is always going to be a balance though between user and advertiser – it is a fun challenge.

The issue of the lackof reliability of AdRelevance spend estimates reflects the difficulty of an online ‘ratecard’. With Nielsen AdEx they use a fomula that attempts to replicate the various ratecards. For example, for print/press they use published rates and discount according to market conditions, agency buying power and client discounts. For television they try to replicate the half-hour TV grid (which is unpublished) and again apply market conditions, agency and client dicsount estimates. For online, we have moved from an extremely naive model based around average cost-per-thousand to a slightly less naive model that takes into account the CPM varies by category. The problem with this model is that the CPMs vary wildly from month to month online as deals are done – something that the other media have managed much better.

Also, if you look at the economic theory of Adam Smith (I’m pretty sure it was him – feel free to correct me) media like TV have finite government controlled “supply” of ad inventory – 13 minutes per hour. Radio too is finite (there are only 60 minutes in any hour). Print/press are constrained by maximum book size. Out-of-home is restricted to the sites they have built or have permission to use – again a finite supply. In the online world the growth in supply seems to show no signs in slowing. In fact, supply appears to be ‘virtually infinite’. When you couple that with the the reduction in demand because of the GFC, this only spells trouble and bargain basement prices – some would even say a race to the bottom in term of advertising rates. I can’t see supply reducing – I suspect rates will not be able to buck this trend.

Regarding the ‘engagement’ metrics, when someone is able to define what ‘engagement’ is in a humanistic form that is advertising medium agnostic, then we may have a chance of developing some. This is something I have wrestled with for some years and still can not see a solution that covers all of commincations planning. While myself and everyone else believes in engagement we simply do not know how to measure it yet. However, I’m sure that the way NOT to test online’s engagement is to stop advertising on all other media – that funnel will empty pretty darned quickly. Multivariate brand-specfic econometric modelling seems to hold the best hope at the moment.

it’s just, generally, clients decide where their budgets go and many still believe that digital isn’t the silver bullet marketing gift from the gods and continue to use channels that have performed well for decades.

blaming agencies is such a tired cliche. agencies are amongst the most committed of digital zealots.

I am surprised that there has not been more debate around the business model of agencies?

Some clients might be better off with a sophisticated online presence, however their agency merely throws out a few display ad’s online via a third party ad network. They then concentrate their efforts on a large and very expensive offline campaign, because it makes far more money for them than an online campaign would – re the charges associated with offline production costs.

So many brands, big and small, are nowehere to be seen online, yet are still appearing enormously offline.

The advertisers that have tracked their campaigns with us are generally committed repeat advertisers. But not many actually track at all.

i know they have a product, from experience i am not sure it really measures what matters.

Don’t be surprised if media-buyers start asking publishers to verify how many ad impressions are reaching real people and not just wasting dollars churning away in a hidden browser tab. This issue is definitely on the MFA’s radar …

On the supply/demand point: A quick look at inventory growth, News & Weather Category page impressions (source: Nielsen Market Intelligence – which is cut up into AdR ad impressions).

Oct 2006: 332,125,634
Oct 2007: 584,097,184 (+75%)
Oct 2008: 809,492,389 (+38.5%)
Oct 2009: 1,004,385,039 (+24%)

I wonder how much of this growth is natural OR due to the introduction of tabbed browsers, always-on homepages and the page impression inflation arms race where anything and everything gets counted?

@Larry, Nielsen has an online ad effectiveness service with brand-shift metrics so the call for these metrics isn’t that weird.

Is it just me can you rephrase this as “we need to sell it better rather than actually make it better”

Glass half empty maybe…