Single Product AdGroups: minimum volume?
Dear Google Shopping enthusiasts,
We've set up a pretty sophisticated Google Shopping structure with multiple campaigns using Single Product AdGroups (SPAG - 1 product per ProductGroup / 1 ProductGroup per AdGroup).
However, by doing so we lost about 75% of our impressions compared to the very basic structure we had before!
Question: can it be that ProductGroups need a certain minimum amount of traffic? I.e. is there a threshold or critical mass needed after which the Google Algorithm starts allocating impressions? Because of the granularity we have quite a few ProductGroups that have zero or single-digit impressions. Maybe this is the reason for us loosing impressions.
Can you think of any other reason why such an account could loose impressions? Anything else you'd like to share about the SPAG strategy?
Thanks in advance,
Re: Single Product AdGroups: minimum volume?[ Edited ]
June - last edited June
impressions are mainly determined by the bid and quality related factors --
internal changes, competitors, and external changes may effect both these
yes, as items decline in performance and lose quality -- via fewer users
searching for specific items, competitor improvements, corresponding
decreases in click-through-rates, etc., such items tend to lose impressions
within shopping-ad auctions over time; accordingly, traffic is tied slightly to
quality, but is supplanted mostly by relevance and landing-page experience
over time and google does not published any such threshold numbers.
adopting a single-product-adgroup auction-bid approach, has many reported
benefits, but these are often misplaced, misconstrued, or misinterpreted,
and results can often be far worse than more conventional product-group
based tactics -- potentially tied to overall business goals, specific strategies
and measured objectives.
for example, a few recounted benefits of single-product-ad-groups include
per-product bidding efficiencies, isolating specific well-performing products,
or effective automated/adaptive bidding -- however, such reported benefits
may be more about purchasing an automated/adaptive bidding system, or
paid bid-management solution, and may even work well when using such
paid systems, but may not reflect effective, real-world, positive, results or
time savings, for most merchants.
also, some discussions that focused on single-product-ad-groups, were urged
to compensate for the legacy product-listing-ads, that did not have effective
per-product reporting without changing a campaign's structure -- this is no
longer the case and adapting a campaign-structure simply to accommodate
a measurement, usually does not result in better auction performance.
such one-off items in isolation, have a more real-world result as described --
falling revenue, falling conversions, falling traffic, and contrary to untested
perceptions, are actually no more, or as, difficult or time consuming to manage,
tend not to improve impression-share, and likely isolate quality-scores rather
than a collective boost of clustering items into relevant themed product-groups
or motifs and themes defined by a business-goal.
one alternative approach is to begin with a small, predefined, well-considered,
set of business goals/objectives, such as, a better return-on-investment,
improving brand-marketing, traffic-building, increasing margins, increasing
market-share, time-sensitive or seasonal sales, customer-retention, etc. --
for specific sets or product categories or predefined products.
these business goals should define corresponding themes to group products.
for example if the product-sets are predefined by a business-driver, then
those items can be immediacy isolated within a separate product-group,
ad-group, or campaign, based on the needs of those predefined goals;
for example, a new brand-marketing offensive, may require a separate
marketing-budget by the business -- therefore a separate campaign
structure would naturally flow from that requirement.
otherwise, the performance within product-groups usually must be reevaluated
on a regular basis to determine which products should or should not be moved;
for example, if the goal is to build traffic, products with a large impression
share may be a better candidate to be isolated, regardless of other factors.
for larger ad-accounts, or extremely large inventory, these tasks usually
only become practical using automation, or a subset of assigned staff.
otherwise, there are ways to better target shopping-ads,
with respect to specific items or groups of items, apart
from more advanced or debatable strategies.
if the goal is to improve the overall quality (impressions) of poor-performing
product-groups, then techniques that include, improving submitted data,
using the custom-label attributes, product-feed-rules, negative-word-lists,
location-targeting, day-parting, inventory-filters, branded and non-branded
campaigns with priorities, dynamic-remarketing-for-retail, or simply reacting
to customer-behavior with respect to price and shipping costs and removing
items that are not performing over time, are potential alternatives.
a common best-practice is to carefully measure related
statistics, before, during, and after any such change.