Soaring on Ridgelift

Analog Dollars, Digital Pennies

It's hard to think of a form of media that isn't being fundamentally altered by the transition from analog to digital and the associated distribution changes (physical to bits, fixed schedule/place to anywhere/anytime).

TV advertising hung on for a long time but faces enormous pressure as:

• People spend less time watching TV
• DVR penetration continues to grow (fast forward me passed the commercials)
• TV content is available online (time and place shifting)

Today's NY Times has an interesting article covering the increasing consumption of TV content via the computer – "Serving Up Television Without the TV Set". There are a couple of quotes in the article that really strike home the disconnect that faces the TV industry:

In an address in January to television executives in Las Vegas, Jeff Zucker, the chief executive of NBC Universal, noted that NBC.com had measured more than half a billion video streams in just over a year.

"Our challenge with all these ventures is to effectively monetize them so that we do not end up trading analog dollars for digital pennies," Mr. Zucker said, calling it the No. 1 challenge for the industry.

Compare this against:

One piece of good news for the networks and advertisers is that viewers are more likely to remember ads on the Internet versions of TV shows, partly because the commercials are less numerous and more demographically aimed online, according to many studies."

The advertisers already realize that those "analog dollars" are really worth pennies and the TV industry is trying to hold on to its old business model as it struggles to find ways of converting "digital pennies" into dollars.

Efforts like Hulu may help but we see yet another example of a threatened industry struggling to prop up its old business model – much as King Canute tried to stem the rising tide.

Will the established players in the TV industry be successful in weathering this change? Parallels in other industries suggest not – the winners tend to be new entrants who wholly embrace the new realities – it's too soon to tell who those winners will be but it's fascinating watching yet-another-industry transition resulting from the impact of the PC, mobile and the Internet.

SMS ads – amazing results!!!

I promise this is the last advertising related post for today!...

M:Metrics in a quote by today's NY Times reported a 12% response rate to advertising delivered to US cell phone users via SMS. This is a staggering response rate and makes SMS advertising a very attractive proposition to advertisers.

Even the European rates at half that number (6-7%) are still very attractive.

Not too hard to image that the forecasts for mobile advertising ($5-11B by 2012) could hit the upper end of that forecast if response rates stay in this range.

Voice recognition and advertising

Yet another interesting article in today's NY Times – "Listening to Phone Calls and Tailoring Ads".

The article describes the service recently launched by Pudding Media that provides free phone calls from your PC to regular phone lines in exchange for presenting ads to you during the call.

Pudding Media uses voice recognition software during the call coupled with basic demographic information about you (captured at sign up such as age, gender, native language and ZIP code) to tailor the ads that are presented on the PC client during the call.

Pudding Media intends to price ads using a cost-per-click model. Using voice recognition to tailor ads is a very interesting idea and yet another way to monetize a free service through advertising.

It remains to be seen how end users will tolerate this kind of relevance profiling but I suspect for many users this will be a non-issue in exchange for free phone service. Pudding Media faces plenty of competition as the market for low cost to free phone service is very crowded with low cost alternatives that are ad-free including SkypeOut.

RSS advertising – a flash in the pan?

Is RSS advertising as we know it dying a slow death?

Even before Google's acquisition of FeedBurner, I had noticed that the number of ad impressions delivered through my RSS feed were very lumpy as we started into the second quarter of 2007. Before that the impression numbers had been building steadily.

I started keeping an eye on the percentage of inventory sold on the Venture Capital channel of FeedBurner's Advertising Network. During the early part of Q2, the percentage sold was in the mid 20's but has since dwindled to zero as of today.

It's hard to tell whether this is a byproduct of the Google acquisition or whether it's something more endemic to RSS advertising itself. From the support forums at FeedBurner its clear that I'm not the only one seeing a significant decline in feed advertising.

As Fred Wilson said on the Union Square Ventures blog:

"For all of FeedBurner's success, monetizing the feeds has been a struggle. The inventory they manage always seems to grow so much faster than the advertising they sell. Their self service advertising system never really took off, and that made it even harder to monetize all the inventory they manage."

An obvious move would be for Google to marry AdWords with a richer graphic ad in RSS feeds and convert from a CPM model to CPC.

Internet Ad Exchange – Selling pork futures

There's an interesting article – "Selling Web Advertising Space Like Pork Bellies" in today's WSJ that provides background and commentary on the emergence of online exchanges for Internet advertising.

I wrote my first post on this topic ("On-line advertising – imperfect pricing") back in February and there have been many developments since then such as DoubleClick announcing an online exchange and now others such as Right Media getting into the same business.

The article also has an interesting snippet about the potential for this to be used by publishers to optimize the revenue they generate from their advertising space:

"With ad exchanges, member advertisers specify the price they're willing to pay for a certain type of ad spot, such as a banner ad that will be viewed by a female in Boston. When a woman in Boston pulls up a Web page of an exchange member with a banner slot available, software assesses the exchange's offer. If the price offered is better than the site's minimum rate for that page and higher than what it can get from other sources, such as ads sold by its sales staff, the site will usually accept the exchange-brokered offer."

As I mentioned in my post on optimizing revenue for publishers, I think this is an interesting space for a start-up. As with exchanges in the physical world, there will be the need for additional services built around the exchange to add value and help parties on both sides of the transaction optimize their revenue.

Online ad exchanges are a fascinating space!

 

 

Publishers and the ad-network black box

As more publishers leverage Internet advertising to generate revenue, there's a need for advertising networks to be less of a black box to the publisher and for services that help the publisher maximize their revenue.

My last post ("Pitfalls for publishers using ad networks?") was prompted by a dramatic reduction in the ad-revenue generated by my blog in March as a result of the end-of-month "true up" by FeedBurner (the ad network I use on my blog feed and web site). 

I contacted FeedBurner last week and received a prompt and detailed explanation of the cause.  Turns out that one of the advertisers uses a third party to verify the number of ad impressions delivered as an audit function – they only pay for verified ad impressions and the third party had incorrectly configured their server resulting in only 55% of the impressions being counted.  The reduction in verified impressions led to the reduction in generated revenue.

FeedBurner was very responsive and provides a lot of information for publishers to manage the advertising delivered with their content but it hard to track ad impressions to content. 

If you read my last post, you know that I inserted a tracking GIF into the post so that I could compare the number of downloads of the tracker against the post readership stats reported by Feedburner.  No surprises here – Feedburner's account of readership was very close to the number of tracker downloads with any differences (< 5%) explainable by duplicates in the logs from my web server.

It's much harder to determine which posts led to which ad impressions – a publisher's view of yield if you will. 

Advertisers look at yield in terms of the percentage of ads that delivered a desired result – for direct mail for example, it's the percentage of replies received for a given campaign; for e-commerce advertisers, it's the "Return on Ad Spend" (ROAS) which compares dollars of revenue generated per dollar of advertising.  I led an investment in ClickShift which provided a service to maximize ROAS through campaign management for the advertiser – the results were spectacular and showed just how much opportunity exists for improvement.

There's clearly an analog to this improvement opportunity for publishers – managing the placement of ads from different ad networks with content generated by the publisher.  This would enable the publisher to maximize revenue generation by steering impressions to ad networks that provided larger inventory and higher CPM ads.

Content is moving towards syndication over more and more channels (web widgets, RSS, Instant Messenger, Mobile,…) and there will be more opportunity for the publisher to maximize their revenue from advertising both within and across channels.  It's going to require additional instrumentation and smarts plus more transparency from the ad networks.  This will help publishers and doubtless place downward pressure on the revenue splits with the ad networks.

Pitfalls for publishers using ad networks?

How big a hole is there in Internet advertising for the publisher? How does the publisher audit the advertising that gets placed with their content?  There are plenty of companies focused on providing auditing services for the advertiser – detecting click fraud, verifying ad impressions etc. but seemingly few focused on the publisher.

When a publisher uses one or more ad networks to provide advertising with their content, how do they know that the payments they receive are correct?  Many may choose to trust the statistics provided to them via the ad network but do they really have a basis for trust?

A side benefit of writing a blog is that it offers a platform to try out different things and see how they work.  You can see a number of VC bloggers use this to learn more about what works (or doesn't) – for example, Fred Wilson's sidebars full of "bling" – not just to dress up a blog but to see how people interact and respond to different trends.  It's like having your own living laboratory!

I decided to use this post as an experiment to see how the stats I get from the different ad networks correlate with delivered impressions.  Since a lot of impressions get delivered via RSS feeds, there is a limit to how you can track and verify impressions.  This post has a transparent embedded 1 pixel image that gets served from a server where I have access to the logs.  Hopefully (!) the image will be downloaded by the different feed readers and browsers and I'll be able to track the impressions and compare them against the stats from the ad networks.

I'll post what I find in a few days!

DoubleClick – NASDAQ for online ads

DoubleClick announced today that the company is forming an online exchange to broker advertising between buyers and sellers. You can find good coverage in the NY Times today in this article – "DoubleClick to set up an Exchange for buying and selling Digital ads".

It was only a few weeks ago that I posted about the lack of transparency in ad pricing so it is great to see the emergence of an online market for advertising appear so quickly! DoubleClick is in an ideal position to transition into a market maker because they already have scale coupled with relationships on both sides of the advertising transaction. I am sure that we will see other ad networks adopt a similar approach.

DoubleClick could become a central piece of infrastructure around which other companies can build new products and services. This could be an interesting opportunity for startups to build analytics, programmed optimization services, audit and tracking services etc. in much the same way as there are product and service companies built around the stock markets such as the NYSE or NASDAQ.

If you have ideas for new companies in this space, drop me an email!

Pull-based advertising gets closer

Pull-based advertising where users pull relevant ads just got a step closer.

Today's NY Times has an article "A Richer Trip to the Mall, Guided by Text Messages" that describes one of the latest tests of pulled based advertising. The article describes a company here in Silicon Valley called NearbyNow that uses SMS to gives consumers access to location specific advertising. The company tested out its service at one of the local San Jose shopping malls (promoted by signage in the mall) and got over 2000 consumers using the service on the first day. The company's distribution strategy is to sign up shopping malls across the country – providing another link between the consumer researching what to buy online but opting for local fulfillment through traditional brick-and-mortar retailers.

Converting advertising from push to pull is going to be an important step in improving advertising yield and re-gaining the consumer's attention. There will be other interesting ways of distributing pull-based advertising; for example using metro-area WiFi together with WiFi enabled handsets.

On-line advertising – imperfect pricing

Does the lack of transparency in how pricing is set for an ad-network lead to imperfect pricing where the advertiser pays too much and the property owner gets too little?

I have to believe the answer is "of course it does!"

There's a great article that speaks to the growing awareness of this issue in today's New York Times with the title "An Ad Upstart Forces Google to Open Up a Little".  It describes the progress made by Quigo Technologies, a New York based ad service company funded by Steamboat Ventures and Highland Capital.  Quigo provides transparency to advertisers on how their advertising is placed – something that neither Yahoo nor Google provide.

Google and Yahoo both act as market makers for the pricing of advertising via key words on their networks.  Their market is many-to- one where advertisers bid against each other for key words but Yahoo and Google control the placement of the ads. 

This is great for Google and Yahoo but not so good for the property owners whose web sites display the ads.  Why?  Because not all web sites are equal in terms of audience appeal.  Sites with high audience appeal get the same deal as sites with less appeal – imperfect pricing for both advertiser and property owner.

Of course this raises the debate about cost-per-click (CPC) pricing versus cost-per-impression (CPM) – with CPC pricing; many would argue that the site with the better (more interested) audience will generate higher click-through rates than the site with lesser audience appeal.  Even if this is true, I'd argue that the advertiser and property owner are both being impacted by the performance of the other sites displaying the ad.

Better pricing for both sides of the ad equation would come from a market that was many-to-many.  Advertisers would bid against each other for placement (the bid) and property owners would offer space for advertising at a price that reflected their audience appeal (the ask).  The market maker would act as the clearing house to match and settle transactions.  The rating of audience appeal could be verified by an independent third party or by the market maker via tracking and analytics.

I think it's only a matter of time before this starts to appear especially as more brand-based advertising moves to the web.  Brand owners care as much about the quality of the audience as they do in generating click-through to their site.