As CPCs rise, proper Shopping campaign setup becomes that much more important. This setup involves segmentations around various product types, brand and products. On the feed side, it means ensuring all products have clear titles, descriptions, and images with all categories being classified correctly according to Google’s specifications. But there is another component to Google Shopping that is just as important.
Often neglected in Google Shopping management is the actual optimisation side of things. Though Shopping campaigns lack traditional keyword targeting and standard text ads, they still require ongoing updates. In this post we will detail five ways in which you can effectively optimise your Shopping campaigns to improve performance.
Though Shopping campaigns don’t include keywords for which you bid, PLAs are triggered by search queries. You can view these search queries in the “Dimensions” tab under “View: Search terms” or in the “Keywords” tab by clicking “Details” and then “Search Term > All.” We prefer to review search queries directly within the “Keywords” tab, as we’re able to exclude directly from this interface.
Once we have the queries we’ll set up a filter to view poor performing terms. The filter can be set to your desired preference. We first look at queries that haven’t converted but have seen at least 50 clicks over the last 60 – 90 days.
We’ll also look at queries with the highest cost per conversions and/or lowest conversion rates. For example, we’re likely to exclude a query that has seen 100 clicks and only 1 conversion with a cost per conversion of £100. We may then leave a query that has seen 300 clicks with only 5 conversions at a cost per conversion of £50. Determining when to exclude a query can be subjective so it’s important to analyse in conjunction with your account goals.
When excluding queries you’ll also want to pay attention to your Search Funnels report. This report shows how your keywords and campaigns are working together to produce conversions. For example, a click on a PLA might have been the first click, but the user left the site and then clicked a branded ad and ultimately converted. Unfortunately, you can’t see which query assisted, but you can see that a Shopping campaign helped.
Just like search queries, you can see individual product performance within the “Dimensions” tab.
You can easily see which products are producing conversions at acceptable rates and which are not. For products that are performing well you can bid higher within the Shopping campaign that they are housed. Within the appropriate product group you can subdivide by item ID and set a higher bid.
In this example, the specific item ID may get a £1 bid while the accessories product group may continue with a £0.85 bid. Recognising that the specific product performs better, you are giving it more exposure with a higher bid.
Conversely, you can lower bids on products that aren’t performing or exclude all together. We’ve also created poor performers campaigns that house these ineffective products. The idea is that we’ll exclude these products in our regular Shopping campaigns and relegate to a campaign with a much lower bid.
You can also view performance by brand utilising the “Brand” report in the “Dimensions” tab. Just like individual products, you can see which brands perform better or worse. The goal is to subdivide your product group by the specific brand. For example, if you are selling socks you may only want to bid on the Nike brand because it has the highest profit margin. You would then exclude everything else in that accessories product group. Thus, any “sock” related search would produce a PLA containing a Nike product.
We’ve found that dayparting and geotargeting are two optimisation techniques that are easy to do, but often forgotten. We don’t know if it’s a misconception that these techniques aren’t applicable to Shopping campaigns, but they can pay huge dividends.
After we accrue enough data (again, subjective based upon your goals and industry), we will set my time and geo modifiers. Unfortunately, these modifiers can’t be set at the ad group level (like in standard Search and Display campaigns). We may find that cost per conversion is too high during the hours of 1 AM – 6 AM, thus eliciting a substantial negative bid modifier. After setting the modifiers, continue to review at least once a month for additional opportunities.
Admittedly, optimising the feed can be difficult, but is well worth it in the long run. As stated earlier, Shopping campaigns don’t utilise keywords so Google pulls the most relevant product based upon the feed details. If we’re searching for an oval coffee table, Google is going to show a PLA from my feed based upon the product that best matches this query. We want to make sure that product is the right one.
Similar to SEO page titles and meta descriptions, product titles and descriptions should be written with the targeted keyword(s) in mind. The “keyword research” for potential titles and descriptions can come from:
Having shared these sources, you’ll want to make sure that you still include the product name in the title and description.
Another technique is to ensure that product images are enticing to click. Often retailers will use the standard manufacturer’s image. We suggest having a professional photographer take photos of all products for better quality. We realise that a Shopping “photo shoot” isn’t always in the budget, but we would encourage retailers to do what they can to make the product images stand out.
There is much to be done with Shopping campaigns beyond the initial setup. Just like standard Search and Display campaigns, it’s not “set it and forget it.” Make sure you are utilising optimisation tactics to continue improving performance.