The fact that Shopping seems to randomly pick PLA's to show based on keywords is frustrating
This is probably just venting, but I craft a data feed where I structure the entire thing similarly with similar looking landing pages, but I have no idea why one PLA appears over another. I could set certain PLA ads to some ridiculous cpc but it will never pull even when the information looks good, the part number is in the title, and nothing is different over a PLA that is exactly the same structure and will appear 1st in the ranking.
For me it is a giant problem because I'm constantly having to talk to my boss about why X or Y sku is not appearing on a search.
To the people who may have these issues, do you just constantly tweak the feed/data around? Should I change the picture, change the name around, change the description, put the name/MPN in the description. It's so frustrating that I honestly want to can the shopping feed and just focus on the search ads exclusively.
The fact that Shopping seems to randomly pick PLA's to show based on keywords is frustrating[ Edited ]
June 2017 - last edited June 2017
shopping-ads show mainly based on both the bid and quality related factors.
be certain that there are no targeting related issues that may
be blocking specific items from participating in the auctions --
e.g. negative-words, items not matching a product-group,
location or other targeting settings, campaign-priorities, etc.
inspect search-query related reports to help
determine which terms typically trigger ads.
note that as some items improve their performance and gain quality,
by more users searching for specific items, outbidding competitors,
a corresponding increase in click-through-rates, etc., those items
may be favored in the shopping-ad auctions over time.
depending on the details, variants may sometimes be an issue --
variants must be submitted per all variant rules and policies,
including proper item_group_id and proper variant attributes,
and be relevant for a search-query, or such items may not show.
there are ways to better target shopping-ads with respect to specific items --
based on a business's goals or objectives, such as return-on-investment,
brand-marketing decisions, traffic-building, margins, profit flow, market-share,
time-sensitive or seasonal pushes, customer-retention, etc. -- using methods
such as improving the overall quality of specific poor-performing items or
weak-performing product-groups, custom-label attributes, product-feed-rules,
negative-word-lists, location targeting, day-parting, inventory-filters,
however, one of the main ideas behind shopping-ads is to allow google's
machine-learning-algorithms to decide which users/searches are the most
relevant with respect to submitted inventory/product data -- especially for
frequently changing or very large inventory.
overriding google's automated, machine-learning, systems typically should be
considered if the systems seem not to be working well for a particular set of
inventory, or if specific targeting or particular subsets of inventory are wanted
to be advertised, beyond what the automated-systems are selecting.
a best-practice is to carefully measure related statistics, before, during,
and after any change to any shopping-ads detail -- including the website,
landing-page, submitted (feed) data, merchant-center settings, or campaign.
that said, this is mainly a peer-to-peer forum -- forum-members can offer
mainly suggestions based on the information posted here in the public forum;
posting specific details such as specific searches and landing-pages without
impressions, specific searches and landing-pages with impressions, specific
submitted data, specific shopping-campaign related details, etc., may allow
other forum-members to offer more specific suggestions.
otherwise, yes, there are less automated, more controlled methods, of
creating finer-targeted ads, using other campaign-types and ad-formats.