[About the author: Guest post contributed by the founder of MyPriceIndia.com. a product comparison startup.]
Online advertising has long faced a fundamental issue of being over measured. The whole industry of online ads is metrics driven, which unlike the other forms of advertising media is not relative metrics or statistical figures derived from small sample sets but absolute real time data.
While this measurement brings in a lot of efficiency, there is a lot of brand visibility that publishers provide but go unaccounted. For example: If you choose to pay per click for an ad and that has been viewed by millions but clicked by only few, then the publisher is at loss even if has to do something with your ad’s creative. Or say when you choose to pay per impression but you never measured if anyone actually viewed the ad or for how long. While it did load on the page but did it really get any attention?
Intelligent platforms like Google’s Adwords have been making some justice by considering supportive metrics like CTR to decide on how much visibility an ad should be given and at what price. While this helps optimize things for the publishers, it still does not change anything fundamentally to encourage branding dollars to be spent online. I particularly mean brand marketing spends and not marketing spends at large.
New media platforms, like Twitter and Facebook, that haven’t been driving any large sized banner impressions or high CTR but have helped brands engage with users through conversation and consistent visibility in their streams have found it difficult to define a standard metrics. Stuff like ‘Like’ count or ‘Follower’ count have been a pure form of measurement in the past but given the regular adulteration of fake profiles these numbers have lost value. The marketer has nothing new to show to his boss. Likes and Follow count are nothing but deduced metrics to help account CPA (Cost Per Acquisition). Again this is a life time measurement and not for a given season. Users who ‘Like’ you once continue to be in the fan count even when they no longer like you. The marketer still needs another reason to spend online and another metrics to run after.
Facebook has adapted to reporting ‘reach’ and ‘conversation’ of every update made on brand pages which does help highlight its own cause but there is no standard web wide metrics parallel to this for other platforms. The clichéd ‘Lack of a common currency’ as one might call it. While a click on Facebook ad and on Google ad means the same, a new visitor to your site, a reach count of 10 for a given status update has little similarity to 10 impressions of 250 square pixels banner ad. Drilling even further, marketers have little analogy of this to the old media of TV and Radio etc. that was driven by GRP. We have criticized the relative nature of GRP but it is not the relativity of data representation that is wrong but the measurement technique that deals with extrapolation of a small sample set that has been at fault. And being in an offline world, it is difficult to measure absolute large data sets without incurring humongous cost. Some attempts have been made in the past to measure eyeballs for OOH media at absolute levels but for obvious reasons media owners haven’t been in favour of anything near absolute measurement. There will always be companies like Nielsen who will give you relative data to help make decisions but none will let you know your exact cost per eyeball. That’s probably a post for another day.
So where does Google’s proposed change in metrics come in picture? What do they change? Perception, that is all. By bringing in new web wide acceptable metrics Google is changing the way old school marketers, who still have a handsome kitty size to spend, perceive the web advertising to deliver. By bringing attention to a different metrics Google is giving the marketers an all new reason to embrace the web, just like Social Media did with “conversation” and “engagement”. Now it will be same for lead gen but a lot more for brand building.
The two new proposed metrics from Google are Active View and Active GRP:
1. Active View
Defined as the count of ad banners viewable atleast 50% on the screen for atleast 1 sec, this metrics helps bring some sanity to the crude CPM metrics that marketers are now shy of. Unlike an absolute impression based count that does not say whether the ad was on the header or footer of the page, this metrics filters out a lot of junk ad space that have low visibility or pages that are live for less than a second before they are refreshed or skipped. While CPM defines the true value of the ad space from the publishers’ point of view but only some niche blogs or premium publishers with strong foot on field sales team have been able to convince media planners to buy at CPM. All other advertisers who use DIY tools still prefer the more transparent CPC pricing. A click is counted as a click independent of how much valuable real-estate the publisher dedicated to it. As a result, publishers have to dependent on quality of ad banners for their earnings. Active View count will bring CPM, as a buyable unit, to the forefront.
The lesser significant change that will be seen is the CTRs, when calculated over Active View rather than banner impressions, would be higher and also help determine the quality of ad creative better, independent of whether it is placed in a more active area of the screen or not.
The numbers “50%” viewable for “1″ second are not important now, as they can be changed later. These are just bare minimum visibility benchmark that Google has proposed probably to keep the publishers on the safer side. Coming down from 5 secs to 1 secs would be difficult to justify to the buyer but depending on how publishers deliver, raising the ante should be possible. For now, the fact that these 2 numbers will be accounted is more important.
2. Active GRP
This metrics is closer to the deep pocket media buyers and Google well realises the fact. Defined as the same Gross Rating Point, that typically TV viewership is measured at, this is a relative metrics dependent on your target audience. Active GRP is also a better representation of Active View. For example, saying that 1000 people in Mumbai Actively Viewed an ad for 2000 times may not make enough sense but saying that 50% CXOs living in Mumbai Actively Viewed the ad atleast twice makes a lot more sense to the marketer. Google and the web in general already has some bit of information about almost all ad viewers, so this should be more dependable than the numbers reported by their offline counterparts. Once the metrics finds recognition there will be more investment and incentive for finding more data about the user.
So are these metrics really important and true to the nature of the web? Honestly, it is debatable. For now, it’s important to bring them (the brand marketers) to the game first and then redefine the rules. Of course, there can be no one better doing this than Google, that still owns majority of the online ad pie.
The current form of metrics, CPM/CPC/CPA, is very short-term objective driven without taking into account long-term effects on brand. Also the fact that it will be possible to take audience quality into account, during media planning for web ads, will mean that buyers and seller will have a common inclination, independent of whether the buying is done on the sales team persuasion or with the DIY tools. The publishers will now have an incentive to run after quality users rather quantity page views.
Both these metrics are driving a major chunk of marketing spend that is allotted for branding and online ads occupied very little of it until now. Steven has attempted a tear down on why very little marketing spend is made online by deep pocket marketers from the likes of P&G and Unilever etc. The reason he cites is that the web platform are not made for brand marketing. They are good for lead generation but not good enough for brand building. Google’s new proposed metrics isn’t changing the platform at large but only the perception. It also helps filter some junk publishers. Even though ‘Clicks on Ad’ is perceived to be a transparent metrics, reports suggest a ~20% fraud rate here. One can well imagine the faith of impression based spends. Hence the clean up required.
A lesser discussed and lesser scientific reason for web ads to go through so much of grilling is the dynamic-user-dependent nature of the platform. While a marketer who is watching his ad air during a cricket telecast can be sure that everyone else is also watching the same ad, the same is not true for web ads. Knowing that X million users watched IPL on TV is enough for a marketer who has purchased airtime but same is not true for X million users reading a news report online. If you look at ads on niche blogs, that book inventory on time based and position based terms rather than impression based, advertisers would be less concerned about the impression count because the ads are all common for all users. They have a guesstimate of audience size and type, like incase of TV, which is enough for them. While this might bring inefficiency due to lack of targeting, this helps marketers make decision without getting into granular level data. Same is not true for targeting driven ads that Google has been running after all this while. Going by this theory, local TV ads, the kind served by Google TV in US and Amagi in India, might be subject to such detailed metrics also.
For now, this is a welcome move from Google. A small step with metrics change will prove to be a giant leap in online advertising.
Watch this introductory video on Brand active solution proposed by Google: