This is a post from awhile back that I wanted to re-post in order to spark some new discussion about the topic, particularly in light of the recent announcement about the launch of Twitter ads – Twitter’s new contextual advertising products and services.
The value and utility of contextual advertising has suddenly resurfaced as a hot topic in the search community as affiliates and search marketers ponder new contextual search distribution channels, like Twitter and Facebook.
Will Twitter ads work or are they simply contextual advertising’s shiny new toy?
Ultimately, ROI for marketers will hinge upon how the ads are priced (CPM, CPC, CPA) and how relevant the ad is to the conversation. Marketing viewpoint aside, in my opinion, these Twitter ads (in their current form) will probably have some significant user backlash that will be problematic.
I am not a fan of contextual “search” marketing. It’s my opinion that contextual advertising inventory is typically overvalued and its usefulness is overblown. In fact, I believe that most of the real value that contextual advertising provides is for the publishers who are selling it, not the advertisers buying it. (Note the Adsense ads to your right.) While I’m a seller of Google Adsense ads, I’m no longer a buyer. I could never make the ROI numbers work consistently for me.
The idea of contextual marketing sounds reasonable and, at face value, it makes sense.
According to Wikipedia:
Advertising on a Web site that is targeted to the specific individual who is visiting the Web site. A contextual ad system scans the text of a Web site for keywords and returns ads to the Web page based on what the user is viewing … [for] example, if the user is viewing a site about sports, and the site uses contextual advertising, the user might see ads for sports-related companies, such as memorabilia dealers or ticket sellers.
If you’re selling red sports cars, you put your ad in an article about cars (hopefully it’s an article about red sports cars) … sounds perfectly logical, right?
First of all, should this really be considered “search” marketing? Absolutely not. Contextual advertising is just a different form of banner advertising. A relevant text ad? Hopefully. A better form of banner advertising? Yes. But it’s still, at its core, banner advertising. Contextual advertising is very clearly NOT search. So pretty please with sugar on top, STOP CALLING IT “SEARCH.”
Setting proper classification aside, search marketing should be divided into 2 significant channels of distribution: search engine results pages (SERPS) and contextual networks. The difference between SERPS and contextual networks is a critical distinction … it’s push vs. pull marketing. Contextually targeted advertising is (radio, TV, banner ads) whereby the advertiser is pushing out his messaging to his target audience versus the “pull” effect of keyword advertising-based (SERPS). A consumer that goes to a search engine and does a search query on the term “Buy Sony KE-50XBR950 Plasma Television” is in buy-mode and is most likely predisposed to buying in the near term. Consumers doing a search and finding a “pull” advertisement on a SERP are much further down the road in the buy cycle and are typically much closer to a buy decision than consumers responding (see, clicking) to a “push” marketing message. When comparing contextual to straight search, you get a consumer in sometimes wildly divergent locations within the buying cycle.
There’s really no arguing that the value of contextual inventory and SERP inventory is divergent. SERPS are undoubtedly considered THE premium inventory for advertisers and particularly for performance marketers with respect to ROI. It’s not hard to see that this is an active pool of buyers actively looking for products and services. But sometimes you have to spell out the fundamental differences between the two methods. And let’s not assume that all searchers are actively in buy mode … some people utilizing search engines are in research mode and some searchers are in buy mode. It’s our job as search marketers to separate the two. The reality is that SERPS are a finite source of inventory for marketers so, naturally, additional sources of inventory are in high demand. Search marketers are willing to do just about anything to get their hands on more sources of qualified traffic so contextual inventory has been a welcomed source of additional inventory and an obvious source of distribution for advertisers looking to extend their reach beyond the SERPS.
The largest source of contextual “search” inventory is Google’s AdSense Network with Yahoo’s Publisher Network a distant second. There are other players in the marketplace that provide similar ad inventory including:
• Industry Brains
• Pulse 360 (formerly Kanoodle)
• Context Web
• Vibrant Media
As a search marketer myself, I’ve tried most of the contextual networks and have been, to say the least, less than impressed. But contextual networks can work and should work for advertisers and marketers. From my experience, it simply takes more time, effort and resources to make it work (Ugh!). It is, in a nutshell, much harder to utilize the contextual networks effectively from an advertiser perspective. But it can be done.
Here are some of my major gripes with contextual advertising:
- Lack of transparency in reporting and tracking within each network.
- Weak targeting capabilities or inability to drill down sufficiently to specific sections and specific pages on sites within the network.
- Pricing disparity (still) on contextual inventory. Publishers want more money for the inventory than it is really worth, unfortunately IMHO.
- Bastardization of inventory.
The lack of transparency on the content networks has historically been a vexing problem for advertisers. In the beginning of Google’s AdWords program, you couldn’t separate Google SERP’s from Google’s AdSense network inventory so there was no way to separate out the bad apples. Google (being Google) quickly figured out that this was untenable for advertisers and was bastardizing the value of its more valuable SERP inventory.
Google has done a fair job of segmenting the primary components of their ad network. But they still don’t provide enough transparency in their reporting on their “Search Network”, essentially SERP’s of their major content partners such as AOL and Netscape. Google’s Search Network is very clearly SERP inventory, but it would be helpful to see within Google’s reporting interface how many clicks, impressions and conversions came from their network partners, AOL and Netscape, etc., simply for purposes of identifying the performance of network partners specifically. So while Google’s transparency on their Content network is adequate, it’s still not good enough. Google might have access to some folks in-house that could figure that out … they’ve figured out most of what advertisers need, but this is very clearly a piece that is missing.
I just spoke with my rep at Google and he confirmed that you are not permitted to buy a Search Network only campaign on Google. You can opt-in to buying both a Google Search AND a Google Search Network partner campaign, but you cannot opt-in to just for a Search Network campaign. So that’s a verifiable weak spot in the transparency of the network.
Targeting capabilities on the content networks could also be much better. Content campaigns in Google are essentially “site targeted” campaigns, which is admittedly a big step forward from only being able to run across the entire network. While being able to site target is wonderful progress in the Content Network world (relatively speaking, of course), advertisers who run site targeted campaigns are still running essentially “Run of Site” (ROS) campaigns, which is much better than running ROS across the entire network … but is still NOT GOOD ENOUGH. Google’s Content Publishers ultimately control the advertisers ability to drill down and target sections on the site. Some publishers within Google’s Network will allow drilling down at specific sections and categories. But, in the end, it’s up to the publisher how much targeting they will provide for their advertisers.
Back in mid-year 2007, I tested a few very targeted sites in Google’s Content Network and was able to target specific pages on each of those sites. I thought it was the greatest thing ever. Impressions were obviously very low, but the conversion rate was ridiculous and the ROI was obscene. Naturally, only a few months into this super-targeted campaign, Google shut it down. They had changed their policy on targeting within the network. My rep explained to me that I would now only be able to target a site’s domain “up to the first slash,” meaning the top-level root directory of the site. So if I was a Paid Search consultant looking to advertise my services on www.searchengineforums.com to a topically focused search engine audience, I would only be able to target the top level domain up to the first slash, I could not drill down to more targeted content area that more specifically targets my audience such as http://www.searchengineforums.com/apps/searchengine.forums/action::topiclist/forum::ppc/
Most recently, Google again changed it’s policy to allow targeting up to the 2nd slash in the domain. Still not good enough. At the end of the day, the targeting capabilities are still relatively weak, making it tougher for advertisers to reconcile their spend on these networks. The proof is in the pudding. As advertisers, we can substantially improve our ROI by drilling down into the content with our messaging. The lift in ROI will give us more money to spend on the inventory, allowing market forces to bid up the price of the more valuable inventory for the publishers. It would also push down the price of less targeted, less valuable ad inventory. But more effective targeting capabilities will simply provide advertisers more ammunition to spend more money on the inventory. This is a call to action for both Adsense publishers and Google … give us deeper targeting capabilities on your site(s).
Presently, many Google AdSense publishers have been spoiled by overpriced inventory.
Yahoo Search Inventory — Mystery Meat?
Yahoo Search is a prime example of the bastardization of contextual ad inventory. When you advertise on Yahoo Search you are getting Yahoo SERP’s but Yahoo combines their Search network with other distribution sources, including parked domain traffic, resulting in a nasty mystery-meat hot dog. How can anyone in their right mind possibly consider parked domain traffic to be SERP traffic? At it’s very, very best, parked domain traffic can’t even be considered contextual advertising. I’m sorry, but a bunch of ads on a page with no content can’t be considered contextual advertising. It’s not even a bastardized version of contextual. And there’s no way to separate this traffic that Yahoo calls “Search” or an opportunity to track it separately. So in this respect, Yahoo fails miserably at the transparency test. I’ve complained to Yahoo repeatedly about their pot-luck inventory, but to no avail. At the very least, put your parked domain traffic in your Contextual bucket, but for God sakes please don’t call it Search inventory. You’re embarrassing yourself. Worse yet, it has destroyed the performance of my campaigns so I was forced to revised my bids down to levels that cuts my Yahoo spend to less than 5% of my total Google spend.
Nice work Yahoo executives.
Social Site Ad Inventory: Sizzle or the Steak?
And finally, I just don’t understand how these social sites think their going to puke Adsense all over their sites and cash in. News flash – wrong audience for that. It will never work in that manner and not nearly to the scale that was first envisioned. Facebook and MySpace’s inability to effectively monetize their contextual advertising inventory is a primary example of how devalued this ad space can be with insufficient transparency and a lack of targeting, tracking and reporting capabilities. Google has reportedly been struggling mightily to monetize the MySpace inventory that it brokered from it’s much bally-hooed $900 million ad deal with the social network stalwart so much so that Fox is considering backing out of the deal entirely.
I was always suspicious of this advertising inventory, but never more so than when the ad rep
that shall go unnamed at Google told me that I should exclude any of the available social site inventory from the campaign. Now granted, the campaign was in the personal finance category so it was completely logically to exclude this, but she very specifically told me not to include MySpace in the campaign. Hmmmm.
Monetizing social content will take work. Hard work. While these big social sites might prefer sticking an older, easier, lazier less-complicated ad model on a new beast, there is a better way. It’s called grunt work. Lazy monetization has ended folks … like gasoline for $1/gallon and stamps for a dime. Sayonara.
- Effect of User Mindset on Relevant Search Results - TweetSeveral years back, Yahoo! released a product called Yahoo! Mindset which dealt with the challenge of differing mindsets that users might have for the same...
- Big Mistakes to Avoid with Your First PPC Campaign - TweetWhile you’re learning on the fly with your e-commerce business, if you’re anything like me, you’re going to make lots of mistakes. Your goal should...
- Automating Your Twitter Account The Right Way - TweetAutomation of your Twitter account is one of the most controversial subjects within the Twitter community. It’s important to understand that any type of automation,...
- Setting Up Google Alerts for Twitter Automation - TweetThe first place to start the Twitter automation process is by setting up several Google Alerts, which will monitor Google results (news, blogs, Web, video)...
- Safe Audit Banner Transaction System – Product Review - TweetBy Paul Lang, Editor, Sell It! May 23rd, 1998 As part of the promotion strategy for your Web store you may well decide to use...