Thursday, March 14, 2013

more texas sharpshooter fallacy

A report from Ebay this week appears to confirm what we have suspected for some time.

That for well-known brands most of the buying up of search terms is largely a waste of time and money and has little effect on sales because the clickers would have found the brand anyway and purchased anyway.

The report, a hefty 25 pager (pdf available here) summises that "Advertising may appear like it is successfully attracting these consumers, when in reality they would have found other channels to the firm's website,".

In several experiments eBay tested removing paid-search keywords using its brand name from Yahoo and Microsoft search engines while maintaining them on Google.

[NB: There is, of course, the possibility that Ebay may take some arch enjoyment from having a pop at Google]

The authors of the report; Thomas Blake, Chris Nosko and Steven Tadelis offer three key take-outs.

'First, “brand” keyword advertising (where firms purchase advertisements on searches for their own brand name), a practice used by many companies, is ineffective because, absent paid search links,consumers simply substitute to (unpaid) organic search links. This implies that brand keyword advertising expenses have neither persuasive nor informative value to well known corporations, and arguably, for other companies as well.

Second, the effectiveness of SEM for non-branded keywords is small for a large and well-known brand like eBay and the channel has been ineffective on average.

Finally, we show that there is a small yet detectable positive impact of SEM on new user acquisition and on informing infrequent users about the value of using eBay.

This last finding supports the informative view of advertising and implies that targeting uninformed users is a critical factor for successful advertising.'


Upon understanding this last point then the shlock-headline baiting of articles such as this one from Business Insider from late last year (yet somehow on a second lease of life) is further testament to the grip that the Texas Sharpshooter Fallacy has on marketers and the ROI question.

ComScore 'research' apparently determined that fans of Target on Facebook were 97 percent more likely to spend at Target, and friends of fans were 51 percent more likely than the average population to spend at the retailer (the brand has over 18 million “likes”).

Thus 'proving' 'A Perfect Example Of Why Retailers Should Be Investing In Social Media.'

Whom the 'average population' that are not on Facebook is unclear. As is how on earth those numbers can be quantified.

In short, they would have bought there anyway.

Similarly this week in AdNews comes the revelation from 'Facebook-centric' [natch] marketer Mike Watkins of Mudo Media who calims that every dollar spent on Facebook marketing by his client - teen fashion brand Supre - had delivered $192 in return, again 'proving' that 'Facebook will outperform TV advertising campaigns every time'.

Tellingly Watkins reveals that he is 'sick of Facebook being a marketing bolt-on and wants to be first to the table when it comes to budgets and campaigns.'

Of course he is ;)

There is no doubt that, correctly applied, paid search and Facebook will contribute to the overall effect.

But again the social media experts and prone to inferring the general from the particular and blind to the fact that marketing to existing customers and ergo measuring sales from who would have bought in any case doth not make any case for the replacemenet of this-with-that.

ROI is indirect and fuzzy. In all likelyhood we're no closer to figuring out which 50% of advertising is wasted than we've ever been, and the chances of ever fully figuring it out won't happen any time soon.

And I'm cool with that.

HT Sam

*UPDATE Friday March 15.

From the Wordstream blog, an alternative view.

Dear eBay, Your Ads Don't Work Because They Suck

blog comments powered by Disqus