What the hell is eBay’s search strategy anyway?

eBay along with researchers from the Universities of Berkeley and Chicago, have just released the final version of a study claiming that paid search ads have “no measurable benefit.” This is not really new news; we heard about the preliminary results last year, and pretty much everyone in search marketing had a good hearty laugh – because eBay has long been known for its embarrassingly crappy ad strategy.

The study is done now, and while it’s not completely terrible, it’s definitely misleading, as well as confusing in terms of what this says about eBay’s marketing strategy. In this post I’d like to go over a few points worth rebutting, especially when it comes to trying to apply these findings to the average Google advertiser out there.

It’s not the channel, it’s how you use it

First and foremost, it’s not so much about the channel (paid search) but rather, how you use it. The strategy of carpet bombing Google search results pages by bidding on 100 million keywords with ads using lazy dynamic keyword insertion results in low click-through rates, which results in low Quality Scores, which results in poor ad positioning and very high click costs (as much as 400% more expensive).

eBay would be better off being a bit more picky on where it shows its ads. Overuse of DKI and underuse of negative keywords gives you ads like this:

Organic vs. paid search

A big piece of the findings had to do with the question of whether paid search ads cannibalize traffic and sales from organic search. According to The Guardian, the report claims:

It stopped advertising entirely on non-branded search terms to 30% of the US for a period of 60 days, and found that it had “a very small and statistically insignificant effect on sales… on average, US consumers do not shop more on eBay when they are exposed to paid search ads.”

This idea that an advertiser can just capture those customers via organic search was never really applicable to your average advertiser, and now is probably no longer applicable to eBay either – since it appears to have been hit with a manual penalty and/or Panda 4.0 due to thin content issues.

Image: Moz

As a reminder, Google dinged eBay for aggressive use of doorway pages and internal footer linking to get rankings on a wide array of long-tail terms with spammy, low-quality pages. After the update rolled out, I estimate that eBay had lost 80% of its non-branded organic rankings. So it’s unclear how it’ll maintain its business with organic search alone.

New visitors vs. existing visitors

eBay’s report talked a lot about the value of showing ads for new visitors vs. existing users. eBay found that ads shown to existing customers were less valuable than ads shown to new visitors.

@larrykim You can set up a "New Users" condition in 1 list as an Exclude option, combined with other lists. #ppcchat pic.twitter.com/kz6zSPC6Fp

— James Svoboda (@Realicity) June 5, 2014

I view this as more of a campaign strategy issue (a lack of strategy, in this case) since there are ways to exclude certain audiences in your paid search campaigns:

Alternatively, you could show existing visitors your ads but set the bids much lower if they produce less ROI. This seems like an obvious solution to this finding, rather than concluding that everything in paid search is broken. eBay should look into things like RLSA, custom audiences, and other bidding and targeting options that refine the audience and improve ROI.

The value of bidding on brand terms

eBay’s report also claims that branded keyword search terms don’t provide incremental sales/traffic. Now, I think there is definitely some truth to this. But what it fails to mention is that because branded searches get such high average click-through rates, they get super high Quality Scores at a very low cost per click. Often these ads cost just pennies per click, and easily cost 50-80% less than the non-branded terms in the account.

So the cost for the bidding on brand terms is very low (relatively speaking), and there are good reasons to do so:

• It lets you customise the appearance of your search listings, including sitelinks to drive users to specific offers you’d like to promote.
• It fends off competing ads on your trademark terms (i.e. defensive tactics). It’s a small incremental uplift in visits, as the study shows, but at least those clicks aren’t going to your competitors.
• Having branded terms in your account usually lifts the average click through rate of your overall account which gives your whole account a boost in terms of Quality Score.

So what the heck is going on at eBay?

Overall I’m a little confused about eBay’s search strategy. On one hand it’s saying paid search doesn’t work. But earlier this year, it was cited as being a top spender on PLAs. It’s like it’s not really communicating internally, and it doesn’t appear to be knowledgeable about the latest AdWords targeting features (which is a real shame with budgets this big). Frankly, I also question eBay’s motives in putting out such a study — Google and eBay are kind of known to be frenemies. And large companies don’t usually share their internal SEM strategies, for obvious reasons.

At the very least, I’d be super skeptical of trying to apply these study results to other companies that don’t have as big a brand as eBay. Ironically, eBay’s business was built to a large extent via search marketing over the last decade – but using outdated tactics that aren’t helping it anymore.

This article by Larry Kim originally appeared on the Wordstream blog and is republished with permission.