Numbers Are For Mugs, See?

EARLIER THIS WEEK, I sat in a sunlit boardroom on the 32nd floor of the Toronto Four Seasons hotel and held my fate in my hands.  It was black, very light, just a couple of ounces, really... and you could slide it into a slot on this device that looked like a card shoe and it made a satisfying "thunk."

That "thunk" was the sound of perceptions changing.  And though it scared me a little, there was also a certain comfort in it.

Ratings are always a fraught subject for anyone who works in Television -- but even moreso in Canada, because the distorted way in which they are used.

In the U.S., ratings data serves a simple purpose -- measuring audiences in order to be able to ascertain how valuable the ad space around those shows is.  That's it.  You know who's watching, when they're watching, and for how long -- and you can turn that information into money.

In the haggling around the money we've come to some interesting accomodations & conclusions in the last little while -- and exploded some myths.  First, despite the hype, way more people are watching shows off air than you'd think.  PVR's have not swallowed viewing altogether.  Also, people watching playback off a PVR usually watch the same night -- and lo and behold -- watch a surprising number of the commercials.  "The Sky is Falling" makes a very dramatic article lede in a newspaper, but the prognostications of imminent network death are a bit premature.  Yes, the model is breaking down. Yes, something else is eventually going to replace it.  But no, we're not talking weeks, months, or even a couple years here.   This system might hang on for another decade, for all we know.

Stretch it to fifteen, and I'm out of the business, most likely, so I'm not quite as down with the abject panic as I might have been last year at this time.

In Canada, however, the ratings data takes on another dimension, because it gets wrapped up in politics.

There's the amusing side -- CTV & Global regularly fire out releases trumpeting the triumph of this show or that show they run; and inevitably they're U.S. programs that they had no hand in making.

Over the last couple years we had the hopeful sign of a couple releases going out to trumpet the premiere numbers of Canadian shows.  Global was thrilled about those premiere numbers for The Guard -- until they settled.   Which is the more relevant number? The 2 Million who tuned in for Little Mosque on the Prairie's premiere? Or the 4 to 500 000 left watching now in Season 4?

Where things get complicated in Canada is ratings data being used to prove a political point. The numbers get poured into all sorts of containers that weren't designed to hold them.

CBC tries to point to their ratings as being way up from a few years back to prove that they're serving Canadians and are more relevant than they've been for years -- and critics of the CBC point to an individual show's (or group of shows) performance against House to try and prove that nobody watches CBC.  CBC is up in viewers.  CBC is losing, and can only draw 600 000 or 700 000 against 3 million on CTV and therefore should be taken out behind the barn and Old Yeller'd or something.

Private stations take the tack of deliberately burying shows in bad timeslots with no promotion so that they can take that lack of ratings gusto and use it to prove to the CRTC that 'nobody wants to watch Canadian shows.'

Nobody's immune from this kind of gaming -- certainly not Creatives.  You have people regularly patting themselves on the back for "huge season average numbers" based on a couple of huge numbers with great lead-ins and a collapsing back half.  You can trumpet your numbers all you want, but if you lose 65% of your lead in, isn't something up?

Not if nobody points it out, there isn't.  You're a success -- and the numbers prove it!

Jim MacLeod, the President of BBM Canada, very graciously walked us through the new methodology of the PPM and how it works earlier this week.  A lot of it has been reported elsewhere before:  the new devices are individual & passive, meaning instead of having to enter in a code to record what you're watching, you just wear it.  It also operates by picking up a code embedded in the audio signal of a program every four seconds. The signal's just outside the range of human hearing.  This change of methodology has, since the rollout last year, resulted in some surprising gains in TV viewing numbers.    Co-Viewing is way up.  You have adults watching a lot more children's programming with their kids, and shows that appeal to a family audience, like Heartland, have seen their numbers go through the roof.

The nice thing about the audio code thing is that it makes it easier to track all sorts of things -- when you played back a recorded show, if you watched the commercials, and even if the show you were watching was from your local feed, or a timeshifted earlier station.  (And the implication is also a window into Bit Torrent use...I imagine if your PPM 'hears' you watching a recording of House from Denver when you're in, say, Nepean, and it knows some kind of chicanery was going down.)

It also allows them to better track and record at-work and out-of-home TV viewing.  At work? Surprisingly high.  Also, out of home viewing tracking is why the numbers for things like Idol & big live sports games are so steroid pumped.

The final numbers on the U.S.-Canada Olympic Hockey Final, by the way?  84% of Canadians watched.  Eighty-four.  That's just crazy.

BBM provides its data to subscribers on a daily basis, with final reports issued periodically. To the public, it issues a list of the Top 30 programs of the week, just like in the U.S.  You can see that data here.

The problem with that list, of course, is that because of the nature of our colonized broadcast landscape, it's mostly U.S. shows.  You're very rarely going to find any Canadian shows in the Top 30, and most of the time if you do, it's reality, news, or sports.



BBM's position is to put the data out there and not take a stand on what it means. They don't have software to mine the trends in the data -- that's up to somewhere else.  They do have guidelines about how to use their data in the media.   But those guidelines don't reflect what is the new reality of the war over ratings data in this country.

Unlike the U.S., where leaks and a robust industry press makes analysis of the data practically slutty, Canada has traditionally been in a kind of dark ages when it comes to getting the numbers out.   In the past, networks used to guard the data like it was precious rubies.  There were all sorts of reasons --  it was thought that Canadians wouldn't understand the numbers, because they were used to big U.S. numbers & that it would reflect badly on the station.  Also, when the ratings came weekly or monthly and in printed form, it was a whole lot harder to get it out there than now, where one email to a friendly journalist can expose all.

But 'all' is relative.  The fact is, outside of those Top 30 shows released by BBM weekly, anytime you read about a number in the press (usually for a Canadian show) it's because that number was 'leaked.'

This system doesn't serve the truth very well, because, of course, leaks by their very nature tend to service an agenda, and rarely comprise the full picture.

What's missing in Canada is a vigorous, informed, full analysis of the ratings data & what it really means. There is no column like Mark Berman's Programming Insider from MediaWeek, where you can look at all the trends & unpack some of the network spin about this number or that number & what it really means.

This is not restricted to ratings, by the way.  A shocking number of media stories in this country tend to be rewritten from press releases...single sourced items that paint a slanted picture.  Blame a lack of resources, media concentration, whatever you want -- but there is no doubt that if you're looking for context about media issues in this country, good luck -- it's exceedingly difficult to find.

Think about the whole Fee For Carriage fight, and how that got reduced to two laughably inaccurate straw man choices -- between a "Tv tax" and "saving local tv." 

Not to mention the fact that you could walk away from the recent coverage of the TV policy hearings before the CRTC -- the most comprehensive and important examination of the Canadian broadcast industry in the last ten years, thinking that the above was the only issue -- when, in fact, that is far from the truth.

Does it really surprise you to know that, reportedly, that at one of the major newspapers in this country the dictum was handed down to keep discussion of the FFC debate restricted to the business pages -- lest some of the messier "cultural" implications complicate the business interests of the corporate parent?  It shouldn't.

The reason why people rush to back up their arguments with numbers is that numbers sound really, really convincing.  It smells like evidence.  So the over-politicization of ratings numbers in this country and the lack of analysis around the data contributes to the foggy sense that nobody really knows what's going on.  And if nobody knows what's really going on, then it's very hard to disprove anything, and you can claim whatever you want.

But with a little bit more exposure, all sorts of things could come into sharper relief.

Here's one example.

I haven't seen anybody write this anywhere, so I'll put it down here, now:  the thing that's struck me over the last several years is that you can usually get Canadian programs on Canadian networks to premiere around a million viewers, or even more.

Dan For Mayor, Hiccups, Republic of Doyle, The Bridge, The Guard, Flashpoint -- but going back even further to The Associates, or Whistler, or The Eleventh Hour or Falcon Beach -- Canadians WILL tune in to check out a Canadian show whose premiere is properly promoted.

They show up. They will give you a chance. The Question is, how many stay over subsequent weeks?  

That is the litmus test of any show.  If they don't stay, it can't all be a lack of promotion.  You must look at the product.  And though some dropoff is inevitable, if you drop every week (or, like Trailer Park Boys a few years ago -- if you build every week, that says something.)

But it only says something if it's consistent, and you can see it week to week.  Corner Gas was an interesting fluke for so many reasons, but mainly because its numbers were high enough to get it regularly into the Top 30 Canadian Programs.  So for most of its run, its numbers were out there, a matter of public record. That made it hard for people (who tried, believe me) to claim that the show wasn't reaching people, or wasn't a success.

I don't know where the good Canadian analysis could come from. It seems like it should naturally fall to an organ like Playback, only I doubt there would be the political will there to take it on.

I do think a good first step, though, would be if BBM started releasing a list of the Top 20 Canadian programs by the week.   At least then, rather than leaks, we could all argue & spin & fulminate about the same set of data.

What do you think?  (And I say that with a caveat -- I'm not looking for the usual "birther" discussion when ratings come up.  I'm accepting the premise of the statistical data models BBM uses here. If you have issues with statistics or sampling, good for you -- but misunderstanding how the process works isn't an excuse for saying, "the system is rigged!")