What is this? From this page you can use the Social Web links to save Online Audio Advertising Market Still Developing to a social bookmarking site, or the E-mail form to send a link via e-mail.

Social Web

E-mail

E-mail It
May 29, 2008

Online Audio Advertising Market Still Developing

Posted in: Announcements, Events, What's New

Given the still-nascent online audio advertising market, TalkShoe has decided to pause its TalkCash program. While we remain true believers in the concept, the primary purpose of this action is to allow us to focus our resources on continuing to enhance and grow our popular Community Calling service.

The original concept behind TalkCash was to share ad revenues with audio content creators (aka podcasters), creating an economic incentive to use TalkShoe. It turns out that most people engage in Community Calling for two other non-economic reasons: 1) to more deeply connect with people having like interests [as humans, it’s in our DNA], and 2) to share their voice [essentially the same force that drives blogging]. We estimate that on TalkShoe less that 10% of people are in it primarily for the money.

Two supporting data points: In the past 12 months, the percentage of unpaid host listens on TalkShoe has risen from 3% to 61%. And during the same period, the average time a caller remained on the phone has gone from 31 minutes to 51 minutes. Yes, this is my baby, but even I am amazed at this level of engagement with an “Internet” application.

One other relevant detail: TalkShoe has been paying a subsidized ad-share rate of $10 CPM (1 cent per episode-listen) to hosts. While we sell ads at $20 CPM and higher, we’ve actually been paying out far more than we’re generating due to limited ad inventory. At this time, there are more important things to spend our investors’ money on.

If you’re a casual reader not involved in driving Web2.0 and new advertising models, stop reading here.

For those of you actually interested in the guts of this stuff, read on. I’ll share insights gained through almost two years of testing and tweaking our online audio ad program (aka podcast advertising). I’ll cover Micro-Segmentation, Podcast Advertising Effectiveness, CPM Rates, Host Read Endorsement vs. Non-Host Recorded Ad Insertion and more. And I’ll tell you about a campaign worth $2,747 CPM.

Micro-Segmentation

TalkShoe has demonstrated that audio advertising can be very valuable when properly targeted to narrowly-segmented audiences. An example of this is my Cellar Dwellers Home Winemaking show. Brad Ring, the editor of WineMaker Magazine, emailed me after discovering the show on iTunes. He wrote something to the effect of “Dave, every one of your listeners is a perfect candidate subscriber for my magazine; let’s do a deal.” And even though I average just 500 to 800 unique episode-plays per week, we did. Anecdotally, it appears that a very high proportion of my listeners now subscribe to Brad’s magazine, although neither of us knows the exact numbers.

Now that I’ve seen it in action, the power of micro-segment targeted advertising is undeniable.

Podcast Advertising vs. Internet Advertising Effectiveness

I don’t know about you, but I’m getting pretty good at ignoring most Internet text and banner ads. I suspect that I’m not alone, as I’ve heard that Google’s click rates are dropping.

By comparison, Internet audio ads (my generic name for podcast advertising) have one big positive and one big negative. On the plus-side, our evidence indicates that people actually listen to them. When you turn on your iPod, you’re not talking so you listen. How many times have I heard an Acura ad on an NPR podcast, and yet I still notice them? Heck, as a lifetime Toyota/Lexus driver, I might even consider an Acura next time.

On the minus-side, unlike with Internet ads, an iPod provides no mechanism for clicking on something. Attention: Steve Jobs, entrepreneurs, and/or any venture capitalist; solve this problem and make a billion or two. Cell phones may be the key — texting a code or pressing a button on an Internet-enabled phone could provide for that immediate response. We need it.

Audio Advertising Rates

In addition to advertising for WineMaker Magazine, TalkShoe has run campaigns for Ambrosia Software, Audible.com, Citrix Systems (GoToMyPC and GoToMeeting products), GoDaddy, PageFlakes, and a couple dozen individuals. Rates have ranged from:

• A low of 2 cents per impression to…
• A high of 5.5 cents per impression.

We also tried some “remnant” advertising at 0.3 cents per impression, just to prove to ourselves that it is a very bad idea and should not be repeated. Ads that aren’t relevant to a listener are simply clutter and diminish the value of the good stuff.

Even higher rates can be justified where there is a good match between advertiser and audience (don’t miss the Brehm Vineyards example at the end).

A side note: As with Google AdWords, TalkShoe allows individuals to run online audio ads on a small scale; as low as 1,000 impressions for just $20. You simply can’t do that on offline channels such as radio. But many of these people didn’t understand how to craft a relevant or compelling message, nor did they execute sufficient repetitions. Even though most of us are experts in watching ads, creating them is a lot harder than it looks. Seriously, don’t try this at home!

Host Read Endorsement vs. Non-Host, Recorded Ad Insertion

We’ve experimented with two different approaches to podcast advertising working with Blubrry, a leader in host read endorsement, and Kiptronic, the torchbearer for non-host recorded (dynamic) ad insertion.

Host read endorsement has been the preferred model in radio for years. Rush Limbaugh gets probably a 30% premium for his personally voiced ads flogging the Sleep Number Bed because if you like Rush (or even if not), his endorsement is more likely to influence you. However, while this proven approach works for individual podcasters with substantial audiences, it doesn’t scale down very well to the network-of-lots-of-hosts model. The costs of matching a specific advertiser with a specific host, developing and delivering the message, and then verifying effective delivery, just don’t make sense on a small scale for a network. How would an advertiser even choose the hosts to whom they’d hitch their reputation? They’d have to find 100 hosts to reach 100,000 listeners. It’ll never happen.

To me it’s obvious that non-host recorded (dynamic) ad insertion is the path for a network! Among other reasons, over 30% of TalkShoe’s episode-plays are for recordings more than a month old (the legendary “long tail”). Only by putting a fresh ad in an old recording can you monetize it. Believe me - the folks who made “Snakes on a Plane” won’t still pay you for running their ads today.

Matching is done based on category (e.g., sports) or topic (e.g., winemaking), and listener demographics (age, gender, etc.) and location (derived through IP address). It’s far better than radio and will get better still.

By the way, ad insertion technology still needs to be enhanced in two other ways: 1) interstitial insertion (in program, not just pre-roll and post-roll), which we know our friends at Kiptronic are all over, and 2) a comprehensive “Audio Ad Words” type marketplace for buying impressions. Come on Google; unleash your dMarc technology on the Internet for podcasting. We want it; we need it!

Critical Mass (Audience)

It may be obvious, but a certain scale and a certain level of automation are required in order for micro-segmented advertising to be meaningful and cost-effective. You can’t pay a salesperson to chase deals that generate $10 - $16 per week (that’s what it would cost WineMaker Magazine to advertise on my show at 2 cents per impression).

TalkShoe is now delivering more than 12 million unique episode-plays annually and we can see that our business will become more valuable per unit as we grow to 50 million or, better yet, 100 million unique episode-plays. There’s a definite knee-in-the-curve where growth drives “extra-linear” value (I’d say “exponentially”, but that connotes hype I am trying to avoid).

Inventory Utilization (Emergence of the Online Audio Ad Market)

As all of the airlines know, having a seat on a plane is one thing; filling it is another. The third quarter of 2007 was TalkShoe’s best yet, when 23% of episode-plays carried paid audio ads. However, fast growth in episode-listens makes this number hard to keep up with.

Advertisers generally targeted “categories” rather than individual shows. The Business, Computers, Education, Sports, and Technology categories were most popular. I expect us to average 20% utilization over time, but with less interest in advertising in Politics, Arts & Entertainment, Pets, or Government categories (poor demographics or too controversial). The Religion category is a wild card.

To repeat the subject of this post and not to be forgotten is that the online audio advertising market is still in its infancy. A lot of work needs to be done to establish credible case studies and compelling ROI data. But it’s not all bad news. TalkShoe is using its unsold inventory to drive its own growth and it’s working! Last quarter, volume was up 70% over the previous quarter.

The Most Effective Advertising Ever

So here’s the story of what may be one of the most effective “ads” in the history of the world, on TalkShoe or anywhere else (a claim that I realize can’t be fully proven, but hang with me). In early 2007, a listener sent me a bottle of wine he made from Peter Brehm frozen grapes. It was fabulous, comparable to the best wines I’ve ever tasted. If you’re not familiar with them, Brehm Vineyards grows fantastic grapes and freezes them when they ripen each fall. That way, a winemaker can buy them and get started any time of year.

On January 16th, my wineaux partner “The Other Guy” and I bought Brehm grapes and broadcast the first of six episodes about “Frozen Grapes”. We involved listeners and participants-by-phone in vicarious winemaking with our own Brehm grapes. By the time of episode six on February 20th, at least a dozen listeners had written to say that they too purchased grapes from Brehm. With only 1,547 episode plays during that period, we generated at least $4,250 in sales for Brehm Vineyards. On a CPM basis, that’s $2,747. That’s the power of micro-segmented advertising.

The Bottom Line

So I’ve seen the future of audio advertising. It’s still lacking an instant click; dynamic interstitial insertion is not yet widely used; and no automated marketplace has achieved either full functionality or critical mass. But the ads are relevant to me and they work! As TalkShoe grows and the market develops, we’ll unpause our TalkCash program for the 10% of you who want to “broadcast yourself” and get paid! Until then, just enjoy actually talking with your online friends and groups.


Return to: Online Audio Advertising Market Still Developing