July 17, 2013

eBay: Paid Search Is Worthless

A study done by eBay on the effectiveness of paid search for established brands has found it to be worthless.

The eBay study called Consumer Heterogeneity and Paid Search Effectiveness: A Large Scale Field Experiment (no, I'm not kidding) concluded...
  • There is no short-term value in brand keyword advertising.
  • Stopping brand keyword advertising resulted in no detectable drop in traffic and sales. 
  • The non-brand keyword experiments show that SEM had a very small and statistically insignificant effect on sales.
According to the study...
"eBay historically managed over 170 million keywords and keyword combinations using algorithms that are updated daily and automatically feed into Google’s, Microsoft’s and Yahoo!’s search platforms."
Presumably, eBay commissioned the study to figure out how much of their search money was being wasting. Their conclusion seems to be: all of it.

Now before you read the entire study, just a warning. It is written in that academic style that makes your brains fall out, and includes sentences like this... 
“The instruments isolate the exogenous experimental variation in spending to estimate the causal impact of spending on changes in revenue.” 
Which I think means something like this...
"We wanted to see whether paid search is worth anything or is just another online whack-off."
It also has formulas like this...

I don't have room for the rest of the formula, but I'm sure you can work it out for yourself.

Also, the author seems dangerously fond of the word "endogeneity," which I have always felt is inappropriate in polite company.

After wading through several pages of this stuff, I think the most useful piece of information is this -- eBay concludes that people who click on a paid search listing are just looking for a shortcut to a website which they would find anyway. So why pay for it?

I have no idea whether this study is worth anything or not, but I'm hoping it starts a great big shit fight between eBay and Google, which would be tremendous fun to watch.

Near the end of the paper the authors pose this rhetorical question about paid search:
"...why do well-known branded companies spend such large amounts of money on what seems to be a rather ineffective marketing channel?"
And they answer thusly:
"...Google offer analytical advice that is not consistent with true causal estimates of ad effectiveness."
Which I think means something like this...
"Google is full of shit."

Big thanks to Dave Trott for sending me this study.


dave trott said...

Thanks for putting that into plain English for me Bob.
I once read a bumper sticker that said "Eschew Obfuscation" and I've stuck to that advice since.

bees1844 said...

As a former search professional, I'd vouch that a more realistic conclusion is 'ebay's search strategy was [largely] worthless'.

i.e. they were advertising on some of my client's names for products they didn't sell and never had sold. No surprises those keywords didn't convert. We saw that across numerous other markets.

We have built businesses almost solely off PPC and it can be done. Ironically it is big money, lazy advertisers such as eBay who make it harder for niche businesses to compete.

eBay struggled in keeping focus in their advertising. No surprise when they automated their keyword selection by algorithms.

The big elephant in any room with Google is that they want to make money off you, so yes, their advice can be questionable. But if you remember that, Adwords isn't the problem - its what you do with it that is.

Jeff Chausse said...

The study describes how BRAND keyword advertising (paying for ads based on your own brand name) is pointless (not exactly shocking)— not that all keyword advertising is useless (though they say it's less effective than commonly believed). But that's a much less sensational headline, no?

Sean Peake said...

It's worth mentioning that, even though branded keyword and non-branded keyword search ads do not appear to have any benefit pay for big, well-established brands, they may for companies with little brand recognition or are trying to break into new markets. (But that's not where the real money is for Google, Bing or Yahoo.) The bottom line is that the big brands need to see if their paid ad strategies are worthwhile—Google provides the tools to determine that for free.


Cecil B DeMille said...

My George Parker branded bullshit-o-meter is going off, not because of the result but how it's presented.

This did not require a study – which is apparently done to make it more preten...er, credible. All you need to do is look at results. All that shit is trackable. Did people come to the website from that paid search ad? Did they buy something? Study over. Conclusion reached.

Riddle me this: Why the fuck does this require academia and the abuse of the word "endogeneity?" This is a ploy. And why would eBay publish it? If your competitors are wasting their money on paid search, why alert them to the fact? Why not shut your mouth and let them, while you pursue better avenues? This reeks of an agenda – moreso than usual.

Timothy Ng said...

I think the comment "eBay concludes that people who click on a paid search listing are just looking for a shortcut to a website which they would find anyway" is not entirely correct, or shows a lack of online marketing knowledge.

How would they find the shortcut to the website they were looking for anyway? By scrolling through pages and pages of search engine results? Of course the pages that are using SEM wouldn't rank that highly organically on SEO.

david said...

eBay lacks online marketing knowledge?

A Numbers Guy said...

I work in marketing analytics for the ecommerce division of a Fortune 500 company, and our paid search experiments have shown similar results to eBay's. In fact, we ceased the vast majority of our paid search advertising in May of 2012 and have seen no decrease in revenue and an increase in profits (since we are no longer paying search engines for the orders we would get anyway). Scoff all you like commenters, but the numbers don't lie. My company has one full year of unaffected revenues to prove it.

Cecil B DeMille said...

That may also have been the cause of my bullshit-o-meter going off, but I wasn't smart enough to think of it at the time.

Gerald Hensel said...

Oh yeah. So half of the marketing world is investing their money for nothing? And all the analysts have no clue of what they are doing? That sounds reasonable.