Friday, April 18, 2008

Ad:Tech SF - Day 3 - Update 1

So it's the last day of Ad:Tech ... with a stack of smaller sessions to round out the event covering some pretty interesting subject matter. This morning I went to 2 sessions, both of which were great ... especially the session on optimising across media, but more on that later.

10:45: Marketing with Downloadable Media: Podcasts and Vidcasts

Good panel - we had panelists from Revision3, National Podcasting System, Podtrac and Roxanne Darling who hosts her own podcast 'Beach Walks with Rox'.

First thing that grabbed me was that there are standards for podcasts and vidcasts in the US - yes, standards across the board that all podcasts use ... I thought this made complete sense and has no doubt helped the use of these channels as it removes any potential logistical/material problems that can be faced when using the medium.

Some statistics were shown - 9% of the US pop. over 12 had listened to an audio podcast in the past month, and 8% had watched a vidcast. Demo wise the split was pretty even but what struck me as interesting was the audience was a little older - this isn't your 16-35 heartland ... 62% were aged 25-54.

Another interesting stats - 71% of podcasts are listened to on a computer - be it a desktop or laptop. AND 78% of vidcasts are watched on a PC or laptop ... so less than 30% of each are actually watched or listened to on a portable device.

Revision3 showed their use of Google AdSense for Video on an episode of Internet Superstar - nice overlay but the contextual match was way off. Here is the official word on Google's video ads from Google -

Jim Louderback from Revision3 also showed a great clip from diggnation - - where they incorporated in programming product placement but did it in an entertaining way. Louderback said the key to ensuring people don't fast forward your ads is to make them entertaining enough that people want to watch them. Makes sense, sure, but what is also important is that the delivery of said marcomms makes sense - authenticity. The diggnation example was a great one - albeit very blokey (ribs, beer and lingerie)

Revision3 claim 100% unaided recall on at least 1 of their advertiser, and 93% on at least 2 ... he also commented on emerging ad serving technology such as freewheel - - that would make reporting on podcast/vidcasts more robust.

Kim Nobles from the National Podcasting Service also spoke on the 5 key principals for making a great podcast - he called these the 'Rules of Engagement'

1. Pre-pre production. Make sure you are committed. There is no such thing as 'A Podcast' ... it is 'A SERIES of podcasts' ... make sure you are doing this over a period of time. Another key was engagement - don't just try and fit existing content into this new medium - you need to engage people to make them WANT and NEED the content. You also need to promote - especially internally ... make sure all consumer touchpoints know about the podcast and can spread the word.

2. Pre-production. Map the series before you start - what do you want to do, what do you want to cover? How do you want to communicate? When? Where? What will the elements be ... this will create efficiencies from the start. Do you invest in pro-talent or use a subject expert. Nobles said he would almost lead to a subject expert as they are saying things people WANT to hear ... and you can fix up any amateur delivery in editing.

3. Production. Ensure the delivery/output is quality - audiences will not forgive poor production. Invest in what you can to create the best possible output for your budget.

4. Post Production. Use of stingers, graphics and editing to create something sharp and exciting

5. Delivery. He touched on the issues you can face with hosting (as you need a fast, reliable connection given the bandwidth you will be transferring). Formatting - what do you use? Flash, WMV, Quicktime? And lastly, compression. He referred to compression as a 'black art' ... you need to find someone who understands the complexity of reducing the size of data without compromising quality.

Next up was Mark McCrery from Podtrac, they are an ad network giving clients access to thousands of podcasts and really making sense of the audience, they also assist in research and reporting. He touched on the 3 key distribution areas (iTunes, YouTube/video aggregation sites/podcast website) and as well went through some case studies for GoDaddy, Netflix, Audible and Honda. The increases across the key metroics of unaided recall, awareness, intent and favourability were impressive ... my only question (which I didn't get to ask) was whether the medium has such strong cut through now due to minimal advertising ... and whether when/if it becomes more clutter the effectiveness will drop. I grabbed his card so will email him later.

Lastly we had Roxanne Darling - her website is she touched on authenticity being critical and discussed the way advertising utilsing podcasting can be extented - across blogs, comments, contact and even live activation.

She also discussed quickly Ford's Bold Moves initiative - which they spent 1m USD on.

This was supposed to be a candid look into a company seeking to transform itself, the reality was it was a scripted piece of corporate communications. She mentioned that for the first 6 weeks the rss feed on the site was broken (she offered to fix it) and that it really missed the mark in terms of the medium. In this case you have to wonder whether the 5 rules of engagement above were actually followed/considered.

12 Exchange Series: Optimising Across Media

The timing for this couldn't be better as a lot has been said about cross media measurement at Ad:Tech and I think a common consensus is developing that the amount of data online creates and often celebrates is causing the industry more harm than good. The real want now is to make sense of the data and be able to apply it to total media planning across ALL comms channels both above and below the line. This is a monumental challenge for the media industry and one that I think will be bloody.

Hosting this discussion (it was actually a discussion with a lot of audience feedback throughout) were Bill Bean from Colgate Palmolive, Peter Storck from Points North Group and Damon Ragusa from ThinkVine.

Initially the gauaged from the crowd what field they worked in - most were agency side with some client side marketers. They then asked whether many of us had used Media Mix methodology to determine spend and results. A few of us put up our hands ... they then asked whether anyone in the audience felt they were doing a good job of optimising across media.

1 person put their hand up out of a room of 200.

Funnily enough, he worked for Microsoft and only in digital analytics ... so wasn't really relevant to the question. The rest of the audience just sat around, moving around somewhat uncomfortably ... as I think a lot of us in media often feel the measurement and science behind what we do is archaic and based on outdated consumption principals.

Bill posed the question of whether digital actually has ROI nailed? It was an apt one ... I think sometimes digital takes credit for all the activity that revolves around the sales funnel pre the digital action. Case in point, search and some ROI focussed clients obsession around it.

The discussion revolved around HOW you determine a proper media mix - and how can you explain and quantify real rationale around decisions that is based on the here and now ... not 3 months, not 3 years ago and not 3 decades ago.

The concept of data inconsistency across mediums was discussed. In OZ TV measurement and currency is diffetent to Magazines and online. Outdoor doesn't even have a real currency ... how can we be sure that we are making the right decisions. Or do we just consistently trial and over time find learnings ... and if so how do we really know what is working or HOW when consumers change over time and our decisions may be based on something that worked 2 years ago in a different climate.

Media mix models were deemed to be limited as most cannot measure effectively
- Entire impact of campaign (from ATL to BTL, wom etc)
- Unprecedented behaviour
- Change in market conditions
- Granular factors
- It's not reusable/flexible/cost effective or fast

Think about it. You commission a piece of solid research, often it takes 3-6 months for you to get the data back ... by that time the market and your consumer could have moved. This makes you wonder - what is the value, moving forward, of this commissioned research? Is research about patting yourself on the back for what you have ALREADY done ... or arming yourself with information for what you WILL DO in the future?

Damon Ragusa said that more data IS NOT needed. We have enough ... we just need to shift how it is used and the science behind it.

They cited Agent Based Modelling as an answer. It doesn't just look at the who (they are) and what (they consume) it looks at the how and why and when ... it is essentially the same science used behind a lot of war strategy.

Forbes article:

Definitely worth a read ... unfortunately this session was only an hour and they barely scratched the surface ... but it really was a great thought starter ... and hey, that is exactly what these things are all about.

No comments: