AdWords is now Google Ads. Our new name reflects the full range of advertising options we offer across Search, Display, YouTube, and more. Learn more

Ads
2.6K members online now
2.6K members online now
For questions related to Google Shopping and Merchant Center. Learn to optimize your Shopping ads
Guide Me
star_border
Reply

click fraud or degradation of PLAs?

Visitor ✭ ✭ ✭
# 1
Visitor ✭ ✭ ✭

Last year during the months of October, November, and December, I did a PLA campaign for an ecommerce site that sells a very well known, specific brand of musical instrument. The feed supplied data for 20 specific  in stock products and the campaign performed very well will lots of conversions and a surprisingly low CPC. User behavior after arriving to the site product pages was also very good with bounce rates near just 30% and time on site well above the average for organic traffic.

 

The campaign was paused at the start of the new year. It was then enabled again on July 1, 2014 with all the same settings as before. A new feed was supplied, again with data for 20 specific in stock products, and following the same product information format as was supplied during the run last year.

 

This time the performance was completely opposite. We let it run for the entire month of July. No conversions, a bounce rate close to 90%, and time on site averaging less than 5 seconds. CPC was about the same. We checked everything multiple times and confirmed the PLAs were running correctly and appeared as intended. Destination URLs were also verified multiple times through both Adwords and traffic data in Analytics.

 

My first thought was that maybe PLAs just aren't as effective anymore due to their rampant spread across all product categories. Last Christmas they weren't as prevalent, so maybe they worked better?

 

We then turned off the PLA campaign and did a 1 week test of regular, basic text ads specific to 5 of the most searched for products with very narrow targeting of keywords that only included the model names and IDs for those products. Destination URLs for each ad were set directly to the targeted product page on the site. Again, no conversions, bounce rates close to 90%, and time on site averaging less than 5 seconds. 

 

Now I'm suspecting click fraud because to see the same performance across 2 different types of campaigns that targeted specific products using well known model names is just too much to be a coincidence.

 

We have now switched to a basic text ad campaign that only generally targets the brand of the products. No specific model designations are mentioned and no specific model name keywords are targeted. Only general information pages are set as the URL destination. Performance has improved greatly. A few conversions have occurred (although not at the same rate as last year), bounce rate is less than 30%, and time on site is averaging close to 5 minutes.

 

Obviously, more specific data would be needed to know for sure, but based on what I've just described is click fraud a strong possibility? Or are PLAs just not as effective anymore?

 

 

1 Expert replyverified_user

Re: click fraud or degradation of PLAs?

Top Contributor
# 2
Top Contributor

 

(1) product-listing-ad invalid-clicks that are detected by
google should be indicated as a billing adjustment within
the adwords account:
https://support.google.com/adwords/answer/77880?rd=3

(2) otherwise, click-fraud is a rather serious charge

and a click-investigation may be requested here --
http://www.google.com/ads/adtrafficquality/advertisers/click-investigation-request.html
be certain to emphasize that the issue relates to product-listing-ads.

Re: click fraud or degradation of PLAs?

Participant ✭ ✭ ✭
# 3
Participant ✭ ✭ ✭
Hey WhySoComplex,

Celebird has given you some handy links regarding the click fraud situation but let us just consider the situation for awhile and see if we can think of a few reasons this may be.

The first thing that jumps out at me is seasonality. With a number of accounts I have worked with I have seen this kind of difference in the same campaigns simply based on the time of year that the ads are running. It is a fair distance off but a really nice experiment would be to run those ads again during christmas time and see if they perform similar to the first time.

It may seem weird for guitars to be seasonal but it mat be a case of your product making a really nice gift, it will at least tell you something interesting about your target market.

The other thing to consider is these specific model keywords of yours, are you qualifying them with purchase intent keywords? for example instead of " brand name model x" bid on "brand name model x for sale" it may be that users are searching for specific models purely for research and education purposes and thus when arriving at your landing page not finding the information they are looking for and going elsewhere.

These are just some possible situations, and a thorough analysis is needed, as well as review of the campaigns in order to ensure nothing is off (do you for example still refer to xmas specials even though it is july?).

Click fraud is of course a possibility but google is usually very good at picking these kinds of things up on your behalf. If you are able to rule out everything but click fraud then see Celebird's posts.

Hope this helps