73.9 is the magic number for online video

Yep – for online video to reach revenues that equate to 10% of current FTA revenue it needs users to spend 73.9 minutes per week watching video content online.

And this number implies everyone – ie, all of Australian’s 13m internet users – are watching this amount of video online per month.

Over the course of a month this works out at around 5 and a half hours per user.

Last week I wrote a series of posts that tried to take a bit more of an in depth look into online video – the challenges it faced with content, commercialisation and ad formats.

One thing I didn’t look that closely at was, how would it actually get to the revenue target of $370m?

I read the other day (from someone misinformed no doubt) that Google’s business model was apparently ‘build first then monetise’ which struck me as a bit simplistic. I am absolutely certain (or hopeful) that in the area of online video none of our publisher chiefs are looking at the area in this way. They would want to build and monetise at the same time, there’s no room to bleed cash here.

Ok – so how did I get to these figures?

Well – FTA TV generated around $3.7b in 2008. 10% of $3.7b is $370m

I assumed that a video CPM of $60 is what the main players would be working towards. Is this feasible? I don’t know … but for the purposes of this it seemed like a strong figure and in line with current publisher wants.

Personally, I think that in the next 3-5 years there will be effectively 4 ways to make revenue off video. Short form (ie less than 5-10 mins), mid form (under 10 min – 30 min),  longer form (1 hr) and content creation (ie producing video for advertisers and placing as editorial/advertorial/content)

Now, this is a stab in the dark but most predictions are … that each area would account for the following % of total video revenue

– Short form – 40% (1 ad per video stream)
– Mid Form – 30% (3 ads per video stream)
– Long Form – 20% (5 ads per video stream)
– Content creation – 10%

As dollar values per year they work out at …

– Short form – $148m
– Mid Form – $111m
– Long Form – $74m
– Content creation – $37m

Ok … so based on the factors above (ie CPM and ad placements per format) – below are the amounts of streams that would be needed

– Short form – 2.4b
– Mid Form – 616m
– Long Form – 246m
– Content creation – N/A

So based on 13m Internet users you’re looking at per user per year

189 short form clips (3 mins)
47 mid form shows (30 mins)
18.9 long form shows (60 mins)

Breaking out these by week you get 73.9 minutes per user.

So – with this in mind – is $370m a year in online video revenue feasible?

Maybe. Maybe but a few key factors need to be explored. Namely

CPMs. These estimates are based on a $60 cpm across the board … this is definitely at the high end and still far more expensive than TV. Hypothetically if the CPMs drop to $30 then you need to double the amount of time users spend with online video from 73.9 mins a week to 146 mins a week … or just under 2.5 hours.

Sell through. These estimates are based on 100% sell through. Likely? Doubtful … but not impossible … 60-75% would be a good effort based on these volumes.

Volume. Whilst these figures seem achievable (which they are) they involve an absolutely massive increase in online video consumption. Ninemsn stated a few months back they were generating 11m streams across their 2.8m users. This works out at just over 5 streams per user per month … so based on each clip being 5 mins that is around 25 mins per month. For $370m to be a yearly revenue figure users need to be spending 12x that amount of time watching video online.

Content. 73.9m per user per month will require amazing content. Exclusive, first run and a big selection.

Mass adoption. This is the key area. Remember, the numbers above are based on 13m people watching video … which implies universal adoption. Can the publishers get all Internet users embracing video en masse? If they can’t, the figure will be almost impossible to reach and advertisers will be difficult to transition.

So what are your thoughts? Am I way off?


3 responses to “73.9 is the magic number for online video

  1. Ben, a very interesting piece of prognostication. Frankly, I have no idea whether you are right, wrong, in the ballpark etc.

    So, I thought I’d look at the issue from a completely different angle. I thought I’d look at how TV earns that $3.7b.

    Step 1. I calculated using OzTAM and RegTAM data for all of 2008 how many commercial minutes a day we watch. For 7/9/10 we watch 113 minutes a day, and for SBS we watch 8 minutes a day.

    Step 2. The commercial FTAs run 13 minutes per hour of paid ads, but SBS runs around half that. So, looking at it probabilistically, of the 113 minutes on 7/9/10 we’d expect 13/60ths to be ad minutes, so that takes us down to 24.5 ad minutes. On SBS we’d expect 6/60ths to be ad minutes, so that takes it down to 0.8 ad minutes. The total expected ad minutes per day are 25.3 minutes. (Please note I have excluded STV because it’s simply too hard to work out which channels show ads and at what frequency).

    Step 3. But an analysis I did around five years ago found that during TV ad breaks somewhere between 5% and 8% of people aren’t watching (e.g. left the room). So if we take the fact that programme minutes are watched more than ad minutes, we should reduce the 25.3 minutes by 8% which leaves us with 23.25 active TV ad minutes per person per day in Australia.

    Step 4. If we then gross this number up by the 22 million people in Australia, and by the 365 days in a year, we get 186,700 million active TV minutes annually in Australia. These are the minutes that intercept with viewers that earn those $3.7b ad dollars (this may include STV dollars – I don’t have the 2008 CEASA report, jump in and correct it if you do).

    Now let’s try and apply this to online video and Ben’s scenario.

    Step 5. Using Ben’s target of 10% of the $3.7 billion, and if we can assume that TV and online CPMs are comparable (again, jump in and shoot this down if I’m way off!), then online video would need 18.7 billion ad minutes per annum.

    Step 6. Rather than use Ben’s 13 million Internet users figure, which I suspect is the monthly active users, I’m going to make it a bit easier for online and use 16 million, which is based on penetration estimates of 72%. This means that each user across a year, would need to see 1,169 ad minutes per annum. To put this back into Ben’s weekly time-frame this converts to 22.5 online ad minutes per week.

    Step 7. Importantly, we need to recognise that these are ad minutes and not actual ads. On TV, 40% of ads are 15s, 55% are 30s, and 5% are 60s per Nielsen AdEx. So each TV ad minute contains on average 2.75 ads.

    Step 8. If we can assume that ad lengths online end up roughly in the same proportion as TV (personally, I think online will end up shorter as it is a more immediate medium), or that online can maintain CPM parity to TV with shorter lengths, then we can convert the 22.5 online ad minutes per week to 62 online ads per week.

    Step 9. Now, this is where it gets really hexy. If I take Ben’s predictions on short-form etc as feasible, we can look at the ‘ad-density’ of online versus TV. Short-form (40%) is 1 ad per 3 minutes, Mid-form (30%) is 3 ads per 30 minutes, and Long-form is 5 ads per 60 minutes. So in an hour Short Form would be 20 ads (40% of the time), Mid Form would be 6 ads (30% of the time) and Long Form would be 5 ads per hour (20% of the time). So the typical online video hour would contain 10.8 ads. But as this excludes the 10% of content creation let’s gross this up to 12 ads per online hour.

    Step 10. This means that the 62 online ads per week, would require 5.16 hours of video content to achieve, or 310 minutes per week.

    This figure is virtually quadruple Ben’s estimate of 75.9 minutes per week, a huge variance. Mind you, it would have been more like 380 minutes if I had used Ben’s 13m online universe figure! On the other hand, if the $3.7b includes STV then my 310 minutes is on the high side.

    So why the big discrepancy? First, both analyses are full of huge assumptions. If any readers have better data or assumptions, PLEASE correct us.

    However, I think the BIG difference lies in the ‘ad-density’. Commercial TV has an ‘ad-density’ of 13 minutes per hour – roughly a quarter of all TV time is ads. Using Ben’s weights, the online video ‘ad-density’ is more like 10% compared to commercial TV’s 22%. Put simply, if CPMs are comparable, then online video will need more than twice the viewing to match TVs ability to extract advertising dollars … or online video users will have to put up with much more ad content than they are currently experiencing, which given the immediacy of the Internet I don’t think is going to happen.

    Therefore, I conclude one of two things. Either the 10% horizon is a lot further off, or I cocked up my analysis badly. Frankly, I don’t know which it is, but it is all food for thought.

  2. talkingdigital

    very interesting john.

    i think the ways we’ve gotten to the figures is different – but in saying that both are entirely feasible.

    i think the way you analyse it shows how much work is needed for online video to get into commercially appealing territory revenue wise.

    My gut tells me for online video to be something that doesn’t lose money it’s very possible it will require heavier ad loads than we are currently getting on FTA. And that’s factoring in that it’ll be able to charge higher CPMs … which it won’t.

    Very interesting and thanks again John for adding something big to the debate!

  3. Pingback: Is Hulu coming to Australia? And if so, what are the challenges? | talking digital – Ben Shepherd

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com 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 )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s