This is a guest post by James Kelly, Senior SEM Analyst of National Positions.
Product Listing Ads have grown immensely in popularity over the last year and many articles have been written detailing the importance of getting them live. At this point the real question is how to maximize exposure and profitability for your ecommerce marketing campaign.
Feed and bid optimization is the first place that many people look. Both of these are of critical importance, but I find that advertisers spend so much time worrying about these issues that the actual campaign structure is often ignored. The decision on how to structure and segment your PLA campaigns will dictate your ability to analyze and optimize the campaigns over time.
Google provides us with the ability to segment PLA’s by product type, condition, brand, ID or custom fields within your feed (adwords_grouping & adwords_labels). The first pitfall that many advertisers encounter is choosing conflicting segments, so that any single product could have as many as five or six bids coming from different targets. While campaigns built in this way may appear incredibly detailed, they are unwieldy and are difficult to manage effectively.
With this in mind, a good first step is choosing a primary targeting method. Each product will inevitably have multiple targets, but there should be one targeting method that you intend to use for the majority of your bid optimization that will have higher bids relative to your secondary targeting methods. The idea here is to try to remove as much of the conflicting bids as possible so that you can actually dial things up when they are converting, or down when they are not. If a single product has six bids and no rhyme or reason as to which one is highest, it will be very difficult to do something as simple as lower the bid on that product by 50% without also lowering the bid on all related products with similar targets.
My primary targeting method of choice is the adwords_labels field filled with a unique ID for each product. In this case, I have the ability to control my bid for individual products to avoid missed opportunities and wasteful spending. Targeting each product individually can allow greater levels of control, but it can be difficult to account for new products or changes in the feed, which is why additional targeting is required.
Secondary targeting methods can be thought of as catch-alls, whose main purpose is to pick up any products that have not been picked up by your primary targets. I generally use a product type or brand as the secondary targeting method depending on what makes more sense for the particular client. The secondary target must have artificially low bids so that they do not overlap any of your primary targeting. I have heard clients express concerns that the bids are too low, but with proper execution, the secondary targets should be more of a safety net than an active target. If you notice that a secondary target is receiving significant amounts of traffic, this should be a sign to take a look at your primary targets to see what is leaking through. To err on the safe side, I also use an all products target with a much lower bid (sometimes a penny bid) to ensure that I have basic coverage for new products that may not have other targets.
Once the various levels of targeting have been sorted out, the last remaining question is how to structure the targets into campaigns, ad groups and targets. With a smaller product count, I typically create a unique ad group for each product containing my primary targeting method in a single campaign. Secondary targets will also have unique ad groups so that all bids can be managed from the ad groups tab.
Accounts that have tens of thousands of products will more than likely need additional structure in place, if nothing else to avoid the 20k ad group per campaign limit. There are two approaches for breaking down these larger catalogs. The first is to create multiple campaigns (possibly broken down by product type or brand) so that you can keep your bids at the ad group level. The second option is to create ad groups for each product type/ brand with targets for each individual product within that ad group. Product level bid optimization would need to be done on the target level in this second option.
No matter how you decide to segment your PLA campaigns, it is important to have a primary targeting method in mind to avoid overlapping bids. Structure provides control over the bids, which will give you the ability to affect your positioning and profitability in the long run.
James Kelly, author of this blog post, is an Ecommerce Channel Manager and Senior SEM Analyst for National Positions, an industry leading internet marketing company with over 1,000 clients around the globe including Wal-Mart, Land Rover, Club Med and Samsung. The National Positions SEM department in particular has been experiencing rapid growth by providing significant revenue increases for many small and medium size businesses.