Last week, at SES San Jose (a popular convention for the SEO crowd), I participated on a panel called, “Don’t Call It a Comeback: Semantic Technology and Search.” Three of us gave presentations: Nova Spivack from Twine, Lane Soelberg representing Wolfram Alpha, and I spoke about some tests in semantic SEO that we ran on HubPages. Rounding out the panel were Othar Hansson from Google, Hope Hackett from Ask, Kevin Haas from Yahoo, and Mark Johnson from Bing, who also wrote a terrific recap of the panel on his blog. Dana Todd from Newsforce and SEMPO was the moderator.

Most of the discussion revolved around the impact of semantic knowledge aggregation would have on search engine results and the balance of search traffic. During Q&A at the end, though, Dana asked about the impact of semantic technology on advertising.

I offered my opinion that semantic ad targeting has been and will continue to evolve as a refinement of contextual advertising. Semantic technology’s strength is identifying what something really means, what the real meaning behind symbols and words. For example, a simple contextual match on a page about how to clear your browser cookies might serve up ads for Mrs Field’s. Semantic technology would notice the preponderance of tangentially-related terms on the page, such as browser, Web, program and the like, and probably result in an ad for Google Chrome instead.

Semantic technology should also sense the mood of a page, and know, for example, if serving up ads for Microsoft is appropriate on a site that could be potentially bashing them (not a good idea when most savvy Web users know contextual ads are often priced by click!).

Beyond text, the big opportunity is with display ads that, without proper and extensive tagging, usually are clumsy or impossible to contextually target. Making sense of metadata, tags and even copy and design elements within the creative grants ad networks the ability to make the most of their display inventory.

What technologies exist in this space? I already wrote about Peer39 and adpepper’s iSense last year. Here are a few companies that have developed semantic matching technologies in the online advertising space:

• Semantic Match (Peer39)
• iSense (ad pepper)
• Personifi (Collective Media)
• Apricot (4AdNetworks)
• ContextIn
• myLOBster Media Corpus (CID)

The field is being rapidly littered with failures, like Clikkit’s SemSense, Aduna, and Dapper’s MashupAds, so the technology is still in development and its full potential not yet achieved.

UPDATE:  We are glad to hear that Dapper’s MashupAds is alive and kicking … and has been renamed to Dapper Ads.

YieldBuild will be exhibiting at, sponsoring, and attending a number of upcoming events. Please contact us if you’d like to set up an appointment to meet.

Affiliate Summit – New York City

August 10-11

Exhibiting: Merchant Mart Table T119

Paul Edmondson (CEO) – Fawntia Fowler (HubPages)

.

Search Engine Strategies (SES) – San Jose

August 11-12

Exhibiting: Booth 516

Jason Menayan (YieldBuild) – Ryan Hupfer (HubPages)

.

AdMonsters Publisher Forum US XXI – Portland

August 17-18

Sponsoring

Paul Edmondson (CEO)

.

BlogWorld and New Media Expo – Las Vegas

October 15-17

Exhibiting: Booth 415

Jason Menayan (YieldBuild) – Ryan Hupfer (HubPages)

Q2 estimates are in from the big four online advertisers (AOL/Platform-A, Google, Yahoo and Microsoft) and the big news is that the ad market is no longer dropping. Holding steady at about $7.9 billion among the four, the stasis in the growth trend is being heralded on TechCrunch as the new trough in the market, from which we can likely expect growth.

Is there cause for optimism? Although the data from the past decade isn’t granular at the quarter level, the last retraction saw its nadir in 2002, from which a new growth trend was reset. The IAB reported $1.46 billion in Internet ad revenue in Q2 2002, and $1.47 billion in Q3, a similarly flat Q/Q change, and then grew from that point on ($1.5 billion in Q4, $1.63 billion in Q1 2003, and so on).

The unabating focus on performance, ever-increasing publisher inventory, combined with improved efficiency, will probably continue to exert downward pressures on overall spend, but if you’re as “guardedly optimistic” as Obama is about the overall economy, then maybe this would be a good time to call Dave Winer out on his doomsaying.

Online research and advisory outfit Outsell estimates that advertisers will have moved $65 billion in 2009 to online marketing, but we’re not talking about advertising. That enormous figure, about equal to total US spend on television and cable advertising, includes spending on SEO and their own destination sites. So, in addition to buying traffic, advertisers are investing in their own mousetraps, hoping the growing online audience will beat a path to their door.

Is this bad news for the online advertising ecosystem? No. It is a reflection of advertisers’ growing sophistication and understanding of the online marketing universe, which we all know extends beyond advertising. Online traffic isn’t necessarily always bought–online properties can be groomed to receive traffic through search or through cross-channel investments (which can be surprisingly effective). However, advertisers needing scale or granular targeting will always have to reach outside their zone of control and buy. Unless they get too aggressive, budgets dedicated to online advertising will probably continue to be  well spent as long as the gap between time spent online and online advertising spend continues to yawn.

Have we turned a corner? The ability to accurately predict the nadir of a recession is not something anyone has been able to do consistently, but analysts and advertisers are nevertheless sussing out signals that a recovery is imminent. This interactive infographic on the NYT says leading economic indicators (final slide) are suggesting the worst is behind us. A report by Publicis Groupe’s ZenithOptimedia contends the bottom of the worldwide advertising market is in Q3 2009.

Does a general economic recovery portend a rebound for online advertising? According to a recent forecast by Pricewaterhouse Coopers (highlights), yes, but when you dig a bit deeper, the answer is yes-and-no. First, the bad news. Display is expected to drop: from $4.8 billion (2008) to $4.4 billion (2013) in the US, and from $5.1 billion to $4.8 billion in EMEA. The final couple of years of the forecast period will see a modest recovery for display, but not enough to offset the drop from the previous years.

Overall spend is expected to grow through the forecast period (through 2013), so where is that expected to come from? Search. Performance-based advertising will earn a growing share of online ad spend as advertisers look to maximize ROI through search’s greater accountability, as well as take advantage of the search terms they’re targeting elsewhere (read: contextual advertising, widget/microsites, apps, etc.).

Assuming PwC’s projections are accurate, the question is what it will take for display to stop falling and resume the growth other formats will continue to enjoy. Accountability and targeting will probably be the two buzzwords on most inventory sellers’ lips, as advertisers’ expectations to understand what their spend is giving them begins to be heard.

An interesting opinion piece on Minsiders, authored by Accord Media president Marta Wohrle, dovetails off a panel discussion whether behavioral targeting and bigger ads—two buzzwords that have captivated online advertising punditry ever since the market drifted into the doldrums—were the answer to lift online advertising growth. I thought this was an interesting insight:

When I think about old media, it strikes me that it has been very good at relationships with advertisers and rather poor at cultivating relationships with its consumers. The thing about digital media is that it has gone way too far the other way. Audiences are center and front of successful Web sites and, of course, the whole kit and caboodle when it comes to social media. Advertising, on the other hand, has been commoditized.

While it’s true that technology has created some failures in the way it’s been implemented, I don’t think a focus on targeting and exploiting technology is the root of the problem, although I fully agree that publishers’ focus on them at the exclusion of what makes successful advertising work presents an interesting opportunity. But publishers differ.
.

Large publishers with direct traffic

.

The large portal sites, the ESPNs, Dictionary.coms and Hulus of the online world, stitch together large, agglomerated audiences. Browsing behavior can be aggregated and analyzed to paint a picture of broad interests, but since these sites rely much more on casual browsing than active searching, granular information on what purchasing decisions lie within the realm of possibility of their viewers is generally impossible to come by.

Television’s advertising model has proven broad interests and large audiences rely on novelty, creativity and humor to make the message penetrate viewers’ skulls. With the Web you have the possibility of interactivity, although without exceptional novelty, there usually isn’t the incentive to engage. Non-interactive ads need to inject the interactivity of gaming (like Innovid does) to make it compelling. And any novelty that becomes a hit will not last that way for long–today’s hot is tomorrow’s boring. Effective advertising for large publishers will always be a moving target. In this case, IAB’s call for the unleashing of the creative spirit and the OPA’s new large-format ads, are attempts by these bodies for their large publisher members to stay a step ahead of the game.
.

Medium- and Small-Publishers with Search Traffic

.
Although not strictly exclusive of the largest publishers (think About.com and the well-SEOed New York Times) the SMEs of the online world are often less equipped at amassing large audiences as they are at delivering content to a large number of individual searchers. Instead of a huge audience of people you know share a love of football but you don’t know much else about them, search publishers identify 1,000 microaudiences with niche interests and deliver each of them a piece of content that provides advertisers with precise information about what they want to buy and maybe even when.

With the lack of scale at delivering jaw-dropping creative for people interested in hummingbird feeders, publishers have relied on contextual advertising’s ability to mate niche advertisers with niche content. What’s been missing here is not necessarily more engaging advertising (although that can’t hurt), but better targeting. When niche publishers identify visitors with an intense interest in gas masks, then serving up ads for Halloween masks represents a huge missed opportunity. Likewise, it’s a failure of behavioral targeting to not know that someone looking at an article on interest rates after spending several months reading about cars online is probably a pretty good candidate to close a deal with.

So where does that leave the agencies and ad networks? What do they have to focus on: targeting technology or creativity? Depending on the breadth of publishers they’re hoping to work with, advertisers are going to learn quickly that they shouldn’t be picking one.

Google’s AdSense is finding more inventory, now on mobile devices, tapping into the healthy proliferation of applications (apps).

AdSense for Mobile Applications launches today and supports both iPhone- and Android-based applications. IAC’s Urban Spoon, music discovery app Shazam, and radio app FlyCast are among the first using the new ads.  Google claims the service gives developers control over the placement and appearance of ads, presumably similar to standard AdSense for content.

On the advertising side, the process by which advertisers select their audiences and bid for placement is largely similar to the Web-based version of AdSense. “Placement targeting” (choosing which app to advertise on), keyword and broad geo targeting are carried over, although placement targeting is still in the experimental phase.

Google began testing mobile app ads on Urban Spoon three months ago. Its biggest competitor is AdMob, already entrenched on dozens of popular apps.

Since February, AdSense publishers have been able to choose from among three font styles (Times, Arial and Verdana) for their ad units. Now, publishers have some control over font size, as well.

Publishers can choose small, medium and large, subject to what can fit within the ad unit’s dimensions, with small being the current default (that will soon be changed to medium). A size default can be set across your entire account, or you can adjust size at the ad unit level.

It’s great that AdSense is giving its publishers greater control over the way text ads are presented. Publishers might have their own aesthetic preferences—such as wanting the font in an ad to match the font used for content on the page, in terms of face and size—or they might want to experiment with the sizes that improve ad performance. Larger fonts are likely to perform better than smaller in most scenarios, but, as we like to say, without testing, you can never be sure.

When it comes to predicting when the macroeconomic recovery is supposed to come, it’s anyone’s guess. Online advertising, though, is widely understood to be weathering but a small setback in its naturally high growth rate, and recent statistics suggest that the first-time drop in online ad growth is fairly modest, all things considered.

None other than the Beast from Redmond (or rather, a representative from one of its European operations) expects that a recovery won’t happen for another 3 years. In the meantime, John Mangelaars, Microsoft VP Consumer and Online for Europe, Middle East and Africa, suggests an industry shakeup is likely with even established players being winnowed away by an increasingly competitive market.

Why? Part of the reason is efficiency: technology has enabled marketers to better understand and target their spend. Implicit in this is the opportunity to grow when spend can be scaled without compromising ROI.

I’m current at OMMA Publish, the association’s annual conference dedicated to the issues confronting publishers, and this morning’s “great debate” was whether publishers should work with ad networks to monetize their remnant inventory, or if they should sell all of their inventory themselves. Panelists Walker Jacobs (Turner Sports) and David Koretz (Adventive) repeated that allowing advertisers a low-cost channel to purchase advertising through an ad network leads you to a “race to the bottom.”

It’s possible then that advertisers are simply becoming more sophisticated, finding the lowest-cost pathway to buy media, often through ad networks. Networks are offering greater control over ad placement (to avoid problems surrounding sensitivity around context), audience targeting and performance analytics, and their ability to scale might be giving advertisers the right combination of features that they’re looking for. Ad networks have certainly grown their share of total ad spend over the last few years. Whether the drop in unsold inventory has come with a concomitant drop in spend is certainly up for debate.

Yesterday I attended DM2’s ad networks and audience targeting mini-conferences, Digiday Networks and Digiday Target (Twitter #digiday), part of Internet Week here in New York. Both conferences consisted of panels delving into the directions ad networks and behavioral targeting are headed in this time of tightened online ad spend and the growing specter of consumer backlashes over privacy.

The biggest question tackled by panelists in Networks was the fate of the 400-600 (?) ad networks that currently serve up a fragmented market today. Cases were made for their consolidation and buyout by agencies and larger networks, as well as their continued proliferation. Mark Zagorski of the behavioral targeting data exchange Exelate argued that a shakeout has been predicted for quite some time, but still hasn’t happened–an increasingly fragmented ad market continues to serve increasingly fragmented media. Because audience behavior continues to be fine-tuned, ad networks that can deliver the specific audiences advertisers are looking to target will command higher prices, while those without granular audience information will earn less. And, increasingly, audiences are meaning specific purchasing behaviors and stages within the buying process rather than conventional demographic slices, although, according to Amanda Kanaga of Time Warner, agencies are still not yet asking for them.

When it comes to bridging the “data gap” between data providers and agencies, the onus is on the data providers to not only deliver data relevant to advertisers, but also provide the means to apply data-driven insights to media buying and campaign analytics and ROI measurement. Unilever’s Jim Keyt said he was interested in “solutions,” not just data.

There was surprisingly little discussion about privacy concerns and the threat of government regulation of data ownership and use outside Stephen Baker’s (BusinessWeek columnist and author of The Numerati) keynote. However, a speaker argued that the issue goes far beyond online advertising, noting that his supermarket loyalty card probably has more private information about his buying behavior than his ISP.

One of the most interesting startups I heard of was MetricMesh. I spoke with its CEO, Shah Ullah, who explained how the company is aggregating cross-device audiences in order to plan campaigns across media as different, but complementary, as online, mobile and television. Agencies and advertisers can target MetricMesh publishers who can serve ads to audiences that have displayed the sort of behavior across TV viewing, Web site and mobile usage patterns. Interesting stuff!