by gabrielg
Media Sales & Engagement: The 4 Greatest Lies I’ve Ever Seen
8:30 am in Contextual Advertising, Demographics, PPC Info, PPC Marketing Blog by gabrielg
If you’re buying display ads and don’t know about these lies, you’re wasting a good chunk of your money.
In Vanessa Fox’s recent article about the Food Network vs AllRecipes battle to be the top food portal, she quotes an executive at the Food Network promoting what has often been a “funny-money” metric in ad sales. “Advertisers can trust that when they buy ads from us, they are connecting with Allrecipes’ vibrant, active and engaged community.”
An engaged community… You mean one that spends a lot of time on the site, right? Viewing a number of pages, showing that the content really appeals to them?
Awesome! For the website itself…
Lie #1: Engaged Audiences Are Desirable
It’s no secret in web analytics that, in general, your lowest bounce rates and most engaged visitors come from the direct traffic segment. The problem is that these are the exact visitors that are most likely to ignore banners. I addressed this briefly in Advertising Lies: Engaged Audience.
“Engaged audiences don’t click.
Ask any AdSense publisher who clicks his ads most, and you’ll hear that search engine traffic is great. Heck, Chiticka has an ad product exclusively dedicated to monetizing publishers’ traffic earned from SEO.
Within blackhat SEO, it’s also known that providing a poor user experience – where the item sought for in the search isn’t present on the landing page – generates high AdSense clickthrough rates. The AdSense block is the most relevant thing on the page – it best matches the keywords searched for, and is the best next step for the visitor. So they click.”
I know folks using other traffic generation methods and generating double digit CTRs on AdSense, using this tactic, and that’s been the case for years. To put that into its proper context, Problogger estimates most bloggers get 2-5% CTR.
If you’re a direct marketer being sold display ads by a salesman who boasts of his engaged audience… take a pass.
Lie #2: Engaged Audiences Are Undesirable
The flip side of the coin is that some brand advertisers don’t mind not getting clicks. It’s wasteful since you could build a relationship, but they’re alright with only having their ads displayed, because they’re trying to send a simple message – e.g. a brand positioning statement that doesn’t require elaboration – because by sheer repetition they’ll create demand and preference for their brand… the sales will [hopefully - you're not always tracking this directly] come later.
What matters to these advertisers is having the correct demographic makeup in the audience, to ensure that they’re reaching their market. And one of the most reliable ways to do that is to buy ads targeted to an engaged audience, because it’s predictable who that audience is, demographically. That’s what services like Quantcast are for.
Lie #3: The Whole Audience Is Equally Engaged.
Unless you’re buying ads on a site that is wholly reliant on search traffic, chances are the audience is made up of a mix of traffic from different sources.
There are a few implications:
First, run-of-site ads in big media are almost always a bad idea. If you’re a direct marketer, you’re more interested in the search referred traffic, and don’t want the “engaged audience.” If you’re going for branding, you want the engaged audience and the search traffic – which is often irrelevant to a site’s core topic – is not what you’re really after.
Second, you want to be able to target by referring source. You can agree to sign an NDA and ask for a peak at their analytics, or get a third party to audit the data.
Third, measurements and boasts about the biggest portal are pointless. Yes, both quality and quantity are important. But when a company like ComScore makes a fuss because it appears The Food Network has more traffic in absolute numbers than AllRecipes, the best response is to shrug: “so what?” That doesn’t tell me, the ad buyer who’s supposedly interested in this fact, about the numbers of visitors fitting into the particular segment(s) I care about.
Besides which, there’s another lie that makes this big to do pointless:
Lie #4: Panels Don’t Lie.
Panel data is skewed. If you’ve compared your own site’s numbers to Compete.com’s numbers, you know there’s a significant difference. I’ve confirmed this with other publishers.
That isn’t to pick on Compete specifically, it’s just an example (and their products have other value, like finding what keywords competitors use, not to mention being more reliable than tools that just provide educated guesses, like Quantcast at times).
Similarly, since there’s a cost to recruit a user, you can expect that a large percentage of panels are simply made up of the users whose acquisition cost (CPA) was lowest.
That being said, some panels and comparable services like Quantcast are moving further along in the direction of directly auditing traffic data with their own snippet of code tagging all the sites’ pages.
So the moral of this last bit here is not to buy ads based on panel data unless it’s directly measured.
To conclude
Pay attention to the details of what traffic’s attention you’re buying. Engagement suits some but not others, and anyways there’s a variety of engagement with a site’s audience.
Get as much data you can on the audience, and see if it’s possible to only buy the engaged/non-engaged audiences that suit your goals, or else to negotiate the price on the audiences that are less appealing if you must buy them as a bundle with the others.
Opinions expressed in the article are those of the guest author and not necessarily Certified Knowledge. If you would like to write for Certified Knowledge, please let us know.
Google Ad Planner has recently made some 
