Q4: impressions versus GRPs

Steven | marketing, advertising & campaigns | Monday, 09 April 2007 - 19:21

Question 4 in Philippe’s new marketing FAQ: “How can impressions be compared to television GRPs?”

First the terms.

Impressions is a term used in online media, describing the number of times an ad is loaded. Loaded in a user’s webbrowser that is. Some people will describe it as showed, defining impressions as the number of times users viewed the ad, but this is false. It’s not because an ad is loaded that the users actually viewed it. He can overlook it, the ad may be invisible on his screen (below the scroll area for example) or he might not view the just loaded page at all (something happening more often as tabbed browsing becomes well spread).
However, the term impressions echoes the wish that users at least notice, maybe unconscious, the ads.

GRP is the abbreviation of Gross Rating Point, a term describing the reach of a television ad. It’s the combination of the target group reached at a certain timeframe and the frequency the ad is displayed. So if you reach 10% of your target group the first time is showed, and 15% the second time, your GRP will be 25%.

How do they rely to each other? As I pointed out describing impressions we have a problem defining actual reach. When does a loaded ad become a real impression? But at least we have some numbers. In televisions GRP you don’t know the actual effect either. Maybe the viewer is sleeping, or getting a drink. And also notice we don’t have the correct number anyway, they are based on survey (at least in Belgium).
But they do rely to each other because both are attempts to help a marketeer plan his/her campaigns. They are both used to buy a certain reach. In television you will buy up to a certain GRP, online you’ll buy a number of impressions. And yes, they are both guesses.

How can these terms help you, as a marketeer, to book successful campaigns? First know it’s all about reach, reach of your target group. On television it’s much harder to reach these specific targets, and only them. Example: if you book an expensive ad (say superbowl) and you reach a pretty good percentage of your target group, you’ll have a great GRP. But still you are paying a lot of money reaching people you don’t need to reach. So you’ll have a significant reach overhead.
On digital media, in principle, you can avoid this overhead. In principle, because in reality it’s very hard to do this flawless. There are however some factors helping you getting better results:

- The environment. This is the website you’re advertising on. You won’t advertise for McDonalds on a website for vegetarians. Why? Because you know most of the people you reach don’t eat meat. So although all other variables may be great (age, sex, ..), it’s not an interesting audience.
Of course, on television the environment plays as well. You have specific type of programs. If you want to advertise for food, a cooking program is a good environment. You have specific channels as well: music, youth, sports, cars, .. These are all great for advertisers. But remember, in general television is a broad a medium, reaching a broad audience.
On the internet, you have all kind of websites, tackling all kind of subjects. You are able to reach niche groups. This is important! It’s not because you can reach your full target group. It might be possible, and if it is it would be great, but don’t count on it too much. However it is in my mind important because you hardly have any reach overhead. In other word: you know your message is well communicated to most of the recipients. This makes your campaign more efficient.

- The user history. On television you don’t know much. You have an idea about how many people are watching, and hopefully know some interesting key figures about them, but that’s it. On the internet, every user has his/her personal history and profile. Not only can you see what site they come from, you also learn a lot from previous campaign. This is were sales houses can do their advantage. If they have implementations on multiple sites, they know all about you over these websites: how much do you visit them, have you seen (loaded) certain ads, have you clicked them. Based on these information they can maximize affect.
The most advanced you can go concerning user history is behavioral targeting. In this model you literally keep track of what users are doing (on your site for example) and use this information to profile and segment them. If you do this proper you can maximize effect even more, and segment on a semi-individual level.

- The learning campaign. In online media you can measure results. This information is very interesting as such (an other advantage of online compared to offline), but can be used in a realtime set-up to auto adapt the campaign. Let’s simplify things and say you want the highest possible click trough. If you have 3 banners, and 1 is over-performing, while an other is underperforming the ad serving system can automatically “correct” the campaign by serving more banners with the higher CTR.

- The interaction. You probably know I’m a big believer of interactivity. It helps the message to stick. In online advertising you can create interactive ads. In offline adverting, this is still a rare thing. Again Belgium speaking. We aren’t as far as we want in interactive television advertising, but the first results are looking good.

To conclude: as far as I’m concerned GRPs can in a way be compared to impressions. But GRP is a measure invented to claim a certain reach of target. Online you should less try to guess your success, and more rely on measurement. If you can’t measure it, it doesn’t exist.
Second, online you are in the position to track down your target group. Where television will probably go broad, online can go niche. In the future we most likely want to pay more for valuable leads. Keeping your niche group healthy might be the smartest thing you can do. Television needs to find an answer for this, because me might get to a point where almost nobody is willing to pay the reach overhead cost anymore.

5 Comments

Sorry, the comment form is closed at this time.

RSS feed for comments on this post.