TALKING DIGITAL: The online video test

This year one major topic discussion in the digital world has been the broad area of online video. Personally, the area that I think will drive this is professional content – that is, content made by the same people that are making the professional content we already get on TV.

And I think in the medium term that content we generally refer to as 'TV shows' will be the driver.

We hear a lot about 'web only' soaps and comedy shows but 99.9 times out of 100 there's a reason why they're web only. For every Beached Az' there's 1,000 shows like Random Place or PS Trixi (no, you probably don't remember them, nor should you) that fail badly.

So, the challenge it is to utilise the web but not do it in a way that eats into existing broadcast revenue. The problem is, it's bloody hard to pull off.

Quincy Smith, the outgoing head of CBS Interactive Worldwide, said when asked about free online video site Hulu and CBS's decision not to join: "I have to reinforce the big screen because there are 120 billion reasons to.”

"Hulu got extremely successful extremely early, and that might ultimately hurt them … In the end it is not about destinations. It is more about your core content, that it gets seen, that it gets counted, and that we get paid for,” he added.

That's an important thing to note – content that gets paid for. And this is where the workings around professional video online need a bit more attention.

A show on broadcast TV that rates around 2 million average audience over an hour can pull in around $1 million in revenue, maybe more. And that's based on 12 minutes of ad content.

Let's say you move that same show online. And the same amount of people watch it – 2 million.

The first problem is, there is no way you could serve 24 30-second ads online. People just won't stomach it. So you serve five 30s ads, and the 'all people' cost per minute (CPM) is $60 (around three times what advertisers would pay on TV).

With those numbers, the revenue would be $600,000 – half of what you're getting on TV. And a side issue would be, would advertisers pay those sorts of CPMs? Even if they would, you've just lost half the revenue you used to make, but the audience is the same. My gut feeling would be this 50 per cent dip in yield would remove all profit and probably mean the program would be broadcast at a loss if online.

Contrary to popular belief and the general trend of online so far, things that don't make a profit struggle to sustain themselves.

Let's look at a similar example for a first run 30 minute show that is pulling around 1 million average audience metro. A show like this should generate around $400,000 in revenue for the network nationally. Probably more, but let's be conservative.

Let's throw the same show online, and assume the entire audience follows and watches the show in its entirety.

For a 30 minute show my feeling is three 30s is the most ads you could get away with. At a $60 CPM and an audience of 1 million, the revenue for the show broadcast online would be $180,000.

That's a 55 per cent revenue drop from TV to online – with the same audience figures.

Now, if I was a TV executive I wouldn't be seeing much benefit here and I'd be looking to protect my broadcast interests at any cost.

So, let's take a different tack. Let's look at digital as 'complementary' to broadcast. Not something that will lower yield but will add revenue and offer a better, ahem, user experience.

There's no way shows online are going to pull audience in the seven figure range. I'd say right now, the absolute best you could hope for in the next three to five years are shows online generating 200,000 full plays from online. I'm stressing 'full' as often figures that are presented to market are stream 'starts' and not shows watched in their entirety.

Let's look at a show like Neighbours. From memory it rates around 800,000 or thereabouts each night in a metro sense.

If 10 per cent of this broadcast audience number watched each episode online it would equate to 80,000 people. For this to be truly valuable to Ten it would have to be 80,000 people who weren't able to watch Neighbours at 6.30pm for whatever reason. It wouldn't be of much value if these 80,000 were people who used to watch it at 6.30pm but now watch it at 11pm online just because it's available.

If Ten managed to generate 80,000 full plays of each Neighbours episode over a year, that would mean 80,000 views of 225 episodes (based on Neighbours running first-run episodes 45 weeks of the year).

If each episode ran three ads at a $60 CPM that would generate $14,400 per episode in revenue. Over a year this is $3.24 million.

If this is incremental, it's a significant gain. If it not incremental, it's a loss effectively. Why? Because if these 80,000 people were watching the broadcast on TV and drop off, the opportunity cost is significant.

These 80,000 people would have been generating around $32,000 in revenue per episode on broadcast TV, whereas online they only generate $14,400.

So while $3.24 million looks like an impressive figure revenue-wise, that revenue could ultimately cost the broadcast revenue stream $7.2 million. So the ultimate group gain is negative $7.2 million.

That is why it's so important the online viewers are new eyeballs – not existing fans who change their viewing appointment time. And this conundrum faces Yahoo!7, ninemsn, and Ten Digital. It's not as big an issue for Fairfax, News or Telstra, although News and Telstra have interests in Pay TV.