Misreporting Paid Political Search – Is McCain Really Ahead?

It bothers me when I read an article about Internet marketing that makes declarative statements based on flimsy evidence. I’ve become used to it in the mainstream press. But when a specialty publication with presumed expertise in their industry publishes such an article, it’s unforgivable.

Such is the case with last week’s Advertising Age  article proclaiming "McCain Winning Search-Ad Race" (registration required). The article uses data from the third party web site analytics company, Compete.com, as the basis for its argument:

According to data from Compete, 22% of Mr. McCain’s traffic in August came from paid search; only 14% of Mr. Obama’s traffic came from paid search.

Fist of all, ad spend is no indicator of success. From my observations, the vast majority of advertising on Google goes to waste because the advertisers do not know how to implement their ads. Advertisers will send visitors to their home page, rather than to a page within their site that matches the content the visitor was searching for, for example. A lot of advertisers do not create unique landing pages for each search ad they create that matches the content the searcher was looking for, let alone the exact phrase they were using to search.

Second, while Compete.com is a fine service, it, and others like it, including Google AdPlanner, Alexa, and Quantcast often report wildly varying numbers. Search analytics are often best understood in relative terms rather than absolute terms.

Lastly, Advertising Age did not examine the entire marketing environment, which would give you a far better sense of who is winning and who is not. That, of course, would take a lot more effort. They could have saved themselves from running an incomplete, if not inaccurate story, by simply doing a side-by-side comparison of the total web traffic to the two sites using their chosen tool, Compete.com:

This Compete.com chart shows that the McCain campaign clearly has a greater need to advertise, given the volume of visitors to barackobama.com compared to visitors to johnmccain.com.

The article also cites an iCrossing study that demonstrates neither campaign has a strong presence within the search engines when people do issue-related searches.

Digital-marketing company iCrossing has measured candidates’ natural search presence for 224 issue keywords (judged by seeing if any number of their web destinations show up in the top 30 search results) and found that they had no presence for 83% of those issues. At the same time, 87% of the people using the internet are searching for issues.

This is entirely unsurprising because neither campaign’s Web site has been deliberately optimized for natural (as opposed to paid) search engine rankings. Even so, the Obama campaign has a slight edge in that regard because on the interior pages of barackobama.com, issue keywords are included in the TITLE of the pages, which helps search engines rank sites.

Finally, Advertising Age contends that traffic to johnmccain.com is of higher quality because those visitors stayed on the site longer than visitors to barackobama.com:

According to Compete data, 11% of visitors to johnmccain.com who arrived via paid search in August left immediately, compared with 17% of Mr. Obama’s paid-search traffic.

Absent direct access to the campaign Web sites’ own search analytics data, it’s impossible to understand why people linger at a site or leave immediately. The reasons could be so varied and numerous that such broad assertions are essentially meaningless.

The reason could be as simple as the design of the site. Perhaps the McCain site is difficult to navigate, so people spend more time there looking for content.

Let’s assume those bounce rates are accurate. First time visitors to the Obama site’s home page get a splash page that asks for an email address and a zip code in order to continue (there is a skip signup option at the bottom of the page, below the scroll).  That registration page could easily account for a higher bounce rate for the Obama campaign but their conversion rate is likely significantly higher than the McCain campaign’s, so the quality of traffic is likely better.

Let’s take a look at the same search for each campaign.

"John McCain Economic Plan" – The McCain campaign is running a search ad at the top of the results that links to the front page of johnmccain.com, which is a waste of money because it does not match the content the searcher was seeking and many of those searchers are likely to leave and look elsewhere.

In the natural portion of the search results, the link goes directly to the economy section of johnmccain.com but because the link in the search results didn’t contain any of the keywords, the searcher may very well scan past that link.

"Barack Obama Economic Plan" – The Obama campaign is running a search ad at the top of the results with a hyperlink reading "Obama’s Economic Plan" which sends the visitor to a page with a signup form asking for an email address and a zip code and no other navigation. Whether a visitor leaves immediately or fills out the quick form, it’s likely the time on site would be below thirty second period that is generally considered a bounce.

The natural search results include two links to barackobama.com that go to the economic section and to the actual economic plan, respectively.

Search ads to the right include an RNC ad that leads to an attack site, an AARP site that compares the candidates’ economic plans, and a McCain ad that goes to that site’s economy section.

My point here is  to shed a little more light on all the variables involved when analyzing an online marketing and/or advertising campaign and to point out the massive holes in Advertising Age‘s argument. McCain may very well be ahead but the Advertising Age article offers no solid evidence to support the headline.

The problem is not merely that a publication of advertising experts got the story so breathtakingly wrong, but that mainstream news organizations pick up the story as if it were fact.