Thoughts about media in Canada

1000 GRPs = 3.1416 Oompa Loompas


I’m Marc and I’m following up on my previous post, Baby Stepping into Digital Number Crunching where I was saying I would be covering the subject of “not all GRPs being born equal”.

According to a survey (Sample Size: A few hundred people in the media industry in Canada, 2006-2011, margin of error: extreme), half of the industry cannot explain properly what a GRP is. In a nutshell, GRP, meaning Gross Rating Point, is the % of a given population that was exposed to an advertising campaign, multiplied by the average ad exposure frequency.

Now… not all GRPs are born equal. If it was the case, 100% of advertisers would just buy indoor OOH or CPC based Facebook self serve with no call to action, but this is not the reality we know. Newspapers have one of the worst Cost per GRP aka Cost Per Point (CPP) but still manage to get a big chunk of the media budget. Why? Because newspapers work. It is not a matter of GRPs, Impressions, Reach… it is a matter of: If I invest XYZ dollars in daily newspapers, I typically get a positive ROI. End of story.

GRPs are not born equal because some formats are bigger, some are animated, some have sound, some are more present in the day to day routine of the consumer, some are there at very specific & important moments, some are closer to point of purchase, some are meant to educate, others to make you look cool. Depending on your goal, the value of a GRP is different from a media to another.

GRPs are not born equal, even if they look that way. I’ll use an example I was giving back when I was in the OOH industry. Imagine 2 buys, both are Transit Shelters (TSA), both give you 80% Reach and 20 frequency for a total of 1600 GRPs. Same format, same reach%, same frequency, same amount of faces, same efficiency right? Wrong.

Let’s look at buy “A”. I bought them all on the biggest streets in the city. Very condensed buy. Half of my reach for this campaign (40%) passes by every day. They have a frequency of 39. They are fed up of seeing the same creative. They are starting to dislike the brand. The other half saw the campaign once as they were going to see their mom in law (and their relationship with her is “ehh…”) They are not in a mood to appreciate the ad and anyway didn’t notice it. And you can even add to this, if you want, the fact that the street is crowded and the advertising is hard to see.

Buy “B” is spread out all across the city, small and big streets. Pretty much all the reach of the campaign does see the ad 20 times, sometimes going to work, sometimes going to the grocery store, sometimes coming back from mom in law.

Campaign B > Campaign A

GRPs are not born equal because the measurement of each media is different. If I run a newspaper, My circulation is audited and, also, consumers are surveyed and asked if they read me. I have experience, I know when audits happen, give or take a few days. I increase the number of people giving newspapers to people on the street. I do a promo where you get a free Ford Mustang Cobra 1968 if you buy the journal. I artificially boost the circulation numbers.

I know when the surveys run. I know how the sampling works. I can research and make sure to be top of mind against the demo that will be surveyed. And to be honest, it’s not exactly rocket science; there ARE ways to take advantage of a “random” survey sampling.

I’m not saying they are doing it; don’t get me wrong, I’m saying it is very feasible to modify the end result, like getting an inflated reach, if you have questionable ethics.

TV/Radio use Portable People Meter (PPM). PPM are devices that you attach to your belt and that “hears” what station, what show you are listening/watching and gives you audience figure per minute. Now are you buying commercial breaks GRPs or shows GRPs? If you are in Canada, probably show…

Also, as far as I can remember from a presentation I was given by someone that worked on implementing PPM in the Montreal market (aka… no official source… I may be wrong), you need to watch a minimal amount of TV per week to be eligible to have a PPM. Granted, data is then weighted to be representative, but… What if your targets are Very Light TV watchers & therefore are not wearing PPMs. You assume that they watch the same shows/stations than a “similar” population and you buy your 1000 GRPs. Maybe… just maybe… your specific niche target actually watches and engages massively with that station that has 0.1 ratings according to the PPMs. Maybe this station is actually one of the best station overall. You will never know. The sampling is flawed, right off the bat. It probably is not going to change because that would mean giving more PPMs to people that adds very little to the database but will cost a fortune. Better to weight results and hope no one cares about that bias. And let’s be realistic, as an agency or an advertiser, are you willing to pay more to get PPMs data that will be slightly less flawed? Meh… no.

Web is no different, let’s be honest. Cookie deletion & multiple devices per person make it so that your reach is probably very inflated on the behalf of frequency. Ads served below the fold (meaning they are not visible as you open a page, you have to scroll down) inflates the impressions level. Et cetera.

What does this all sums up to? What is your point, Marc? Very good questions!

I’m getting there with this last piece of history, about radio.

When PPMs were introduced in the Montreal radio market (radio was previously using surveys), the results were clear: People listened to much less radio than they thought (about 35% less). Now did the radio industry start giving huge bonuses to compensate for the past 500 years of selling unfortunately inflated numbers? Of course not! What they said was (and they are right), GRPs are just an indicator to help you understand media, it could be called Oompa Loompas, what would it change? Nothing. Except people would look at you weird when you present your media plan.

The price of GRPs fluctuated over years to eventually stabilize at XYZ because this is what advertisers were willing to pay to generate results. Economy. Offer & Demand model. 2+2=4

If I used to buy 1000 GRPs that was usually generating $1M sales and now I buy a more accurate figure, 35% less, 650 PPM GRPs. How much sales will I get?


1000 survey GRPs = 650 PPM GRPs = 3.1416 Oompa Loompas = $1M sales…

Same difference.

So my point is: GRPs, as it stands, are useless. We are comparing inflated oranges with genetically modified apples. If we want to say web should be measured using GRPs… sure… but can we please start using them properly? 1 web GRP =/= 1 TV GRP =/= 1 TV commercial break GRP =/= 1 Oompa Loompa.

Instead, let’s start saying, for a DR campaign focusing on Upper Funnel Conversion against A18-34 with 10 fingers living in Chilliwack, XY media1 GRPs = YZ media2 GRPs.

And then we’ll start talking… about how this is better but still wrong and that we should look at the media mix and how to generate more results. 2+2=5


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 )

Google+ photo

You are commenting using your Google+ 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 )


Connecting to %s

%d bloggers like this: