Dynamic Insertion of Podcast Ads | Todd Cochrane is Not Seeing the Whole Picture
I just read a post by Todd Cochrane (author of the first “podcast how to” book and CEO of RawVoice) condemning the dynamic insertion of ads in podcasts as DOA. I disagree. I think it’s only getting started.
First, a quick definition of dynamic insertion. This simply means that the podcast audio and the audio for the ad are put together on the fly rather than the ad being “hard” recorded into the original audio file. This allows the ad to change over time for the same episode. Now back to Todd’s opinion.
He states:
The second you drop in some pre-produced, auto inserted advertisment, this immediately disconnects the podcaster from the audience, and the advertiser message will not have the impact it should.
OK. I 100% agree with that statement. But why does that invalidate the dynamic ad insertion technology or methodology. Just give a listen to Paul Colligan’s latest Podcast Tools podcasts. At the time of writing this post he’s including seamless, host-spoken ads for Go To Webinar. And guess what…they were dynamically inserted!
Paul created the ad himself, in his own words and then inserted at the point of download. He’s using the Gigavox Audio Lite service (not yet available to the public) for his dynamic insertion. It’s similar to the system that has run the IT Conversations Network for a long time. This system allows Paul to change the ad at any time meaning:
- He can have a different sponsor for the same episode at any point in the future. He’s already planned a different ad for March.
- He’s not stuck with the sponsor for the life of the audio file.
- He can sell his episode archives over an over again as ad inventory (e.g. he can have a different sponsor each month and put them in every one of his episodes)
Imagine if you decide you no longer want to endorse a sponsor who put an ad in ten of your episodes last year. Do you want to have to go back and take that audio out of every one of the files? If you use dynamic insertion an ad can be turned off at any time. Granted, this is subject to the agreement with the advertiser. The dynamic nature and finite term of the ad need to be agreed upon with the advertiser up front.
Dynamic insertion is a pretty handy way to do your ads. Is it the right solution every time? No. Is it what I strongly suggest in most cases? Yes.
Dynamic insertion does not mean pre-fabbed, “radio-guy” voiced ads that break the flow of your show. It means flexibility in your ad offerings over time. Dynamic insertion is far from DOA. This is why companies such as RadioTail invested in it early on.
On another note, Todd also state in the same post that:
We have a very good data that proves ad penetration is around 90% of the total listening audience we don’t get that many conversions but 90% of the ads are getting to the intended audience.
I hope that’s true, but I’m skeptical. I’d like to see how they arrived at that 90% figure. Todd is usually one of the first to demand proof for numbers like this that get touted about podcasting.
Todd is on the forefront of podcasting and I have agreed with many things he says, but he has me puzzled with these recent statements.
February 22nd, 2007 17:31
I couldn’t agree more with you that the technology and ad content are separate functions. Publishers have the opportunity to look at the library of content as valuable over time and not just the time of publishing. Some content is only valuable at the time of publishing. Others are evergreen and can maintain value for an extended period or throughout it’s lifetime. I believe in separating content and technology. At Castfire, we have been doing dynamic episode segments for quite a while (we just celebrated our 2nd birthday!) and it is used for more than just advertising segments. It is also used for promotional segments, re-branding, etc.
February 23rd, 2007 01:24
One of the concerns we have at Innertoob and CrowdAbout is the fact that the dynamic insertion of ads can skew the total time of a podcast quite a bit. If one ad being inserted in March is 20 seconds long, and another being inserted in April is 35 seconds long, you wouldn’t think that would be any big deal. But new services that depend on the reported time of a media file to add meta data along the timeline (contextual discussions, timed show notes, slide shows, etc.) are handicapped (and thus listeners and podcasters are handicapped) by this practice. One solution might be to make the ad identify itself somehow, so that a time-based metadata service could adjust on the fly, but then if people knew where the ads were going to appear, it wouldn’t be long before they just started skipping them altogether.
February 23rd, 2007 02:01
Carter — we are actually working with Pluggd on that issue and integrating via API’s to ensure that their indexes match the current episode. This will become a greater challenge in the future and I expect that companies who build out dynamic content will have to start offering additional services. For instance, we provide slideshow functionality for podcasting, by utilizing the flickr API, and are able to immediately adjust the timing of images based on different segments or advertisements.
February 23rd, 2007 02:51
Carter,
Dynamic insertion definitely has some challenges like the one you brought up. It can cause a lot of problems if the length (and hence size) of the mp3 file is always changing.
There’s some responsibility that falls on the publisher to mitigate these problems. One way to do this is to always make sure that they ads are the same length.
February 24th, 2007 14:14
Jason, I think you’re right - we’re just getting started. There’s a lot of innovation to come in this space.
That said, dynamic ad insertion is probably only going to be used by a small number of podcasters.
February 24th, 2007 20:50
I don’t think that Todd was addressing his reservations on dynamically inserted ads at the level of the individual podcaster. Paul’s approach using Gigavox Audio Lite is more friendly to the user than other dynamic insertion methods.
The issue is with mass dynamic insertions where the ads are not spoken by the podcaster and the tone of the ads might not match the show. Anyone who tries to apply a Google Adsense model for podcast advertising will not get nearly the same results as those who let podcasters produce ads that they can seamlessly integrate with their shows.
Some podcasters may just want to eliminate the ‘hassle’ of having to produce an ad for each show. I have a feeling that those folks will also be making fewer dollars and maintaining weaker relationships with their sponsors than their counterparts who naturally convey their association with a product or service.
People need to remember that the greatest modern innovation with audio advertising did not result from some new technology. It came about when Howard Stern decided to take a more casual, conversational approach to promoting products. It would be a shame to see people ignore the wisdom of that approach in favor of one that values convenience over passion and connection.
February 24th, 2007 23:32
The article I wrote was in response to the Business Week article talking about Google possible getting into the podcast ad business along with their discussion on ad insertion in podcast. See the full article at http://www.geeknewscentral.com/archives/006871.html
Over the past 24 months I have accumulated a wide variety of survey data across multiple show categories to back up all of my claims in the above article.
One of the facts that has emerged is from listener sampling of shows that use auto insertion and those that don’t. Those survey results across the board show that Listeners hate auto-inserted advertising. They can smell them a mile away.
This survey data is conclusive in that listeners automatically discount the context of the ads. No longer is the podcaster listener relationship existent.
The same goes for ad penetration, no real mechanism expect for listener surveys and sampling of listeners can measure the penetration but being both podtrac and us have used post campaign surveys to determine ad penetration we know the message is reaching a very high percentage of the listening audience.
While at the same time I have nearly 500 podcasters earning money through advertising in because of my and my companies efforts to build long term relationships with sponsors.
Little history lesson is in order my personal show brought GoDaddy into the space quickly followed by other companies in which I was the first to sign a group of podcasters to a single contract in 2005 when everyone else was still talking about it.
It should be noted that the majority of the advertisers I have under contract have been under contract a very long time and they continue to renew every quarter. Their are very few companies that have advertisers renewing contracts for more than 18 months.
Is it because we do dynamic insertion aka the paul harvey model which is a host read ad that is more story and relationship than an actual in your face ad.
My article Ground Truth in Podcast Advertising is based upon 24 months of experience in putting dollars in podcasters pockets, delivering outstanding ROI and giving value most importantly to the listener.
I encourage you to poll your own listeners and ask them what they think about auto insertion of ads you may be surprised what you find out.
For those podcasters that want to maintain editorial control of their podcast come over and join the blubrry community we will help you monetize your show and not annoy your listnership.
February 25th, 2007 12:42
Rob & Todd:
Todd doesn’t make it clear if that is the case. I may have misunderstood him. But to me it sounds like he makes a blanket statement saying that dynamic/auto insertion is worthless.
I 100% agree with that. I’ve pointed out the shortcomings with Google Adsense for audio here on this blog. This would be a poor use of dynamic insertion. However, if dynamic insertion is used by or in conjunction with the host, I still maintain that it can be very effective.
In that context, the article makes more sense. I understand your position better now. Maybe this is just a matter of semantics, because the truth of the matter is that I agree with the majority of the points you make in the article.
But I still think that your point should not be against “auto insertion”.
Auto insertion does not equal prefabbed annoying ads that disrupt the flow of the show and the relationship with the podcaster.
Whether or not the ad is auto inserted is not the point. It’s the method of production and presentation of that matters, not the tech used to put the ad in the audio.
I have no doubt that you know what you are doing and talking about. It’s just in your blanket use of the word “auto insertion” that I disagree.
That’s the wrong question to ask. I doubt most of Paul’s audience realizes he is auto inserting ads let alone having an opinion on it because he reads them himself and inserts them in an unobtrusive manner. Sure, if I asked my audience what they thought of prefabbed ads that were not geared towards them, then they would say they hate it. But auto insertion is not the enemy here.
Again, auto insertion does not equal annoying your audience. Poorly produced ads, not letting the host read the ads, etc….these are the things that annoy the listener. That is the point that I wanted to make clear.