As many of you will be aware the Google ‘Shopping Campaign’ options have been on trial now since about October 2013 to a select few merchants. It is rumoured that this will be fully released to all customers next week (start of February 2014), so now is probably a good time to discuss these new features and the changes that are being introduced. Up until this release the only way of managing the PLAs and Google shopping was to setup ‘product targets’ through specific product listing ads campaigns and then using a selection of product IDs / AdWords Labels / Adwords Grouping / Brand etc to set bids. This wasn’t a very easy or straight forwards way of managing the campaigns as there was no product level reporting whatsoever, likewise there was no facility to view or change bids at the product level either. We at Onefeed adopted a policy of putting products into monetary bid buckets so we knew we could change bids at the products level through our system as well as report at the product level also within the Onefeed interface. In effect this was quite complicated but the best of a bad situation. From speaking to many clients who we’ve recently taken on board it appears the majority of merchants have been putting all products into a catch-all product targets and everything was set at the same bid level. The new features in Google do make for an easier to manage campaign and a facility to segment your data on actual values Google are seeing in the feed. Previously this data wasn’t visible at all in AdWords.
So what’s new then?
First of all the structure of a shopping campaign is relatively straight forwards, you can segment out the targeting within the interface right down to the product level using various aspects of your data feed. In effect you are encouraged to split out the ‘all products’ campaign into smaller chunks using brand / product type / categories or labels. All of these data points are interchangeable; so you could break down products within a category by brand or vice versa.
Basic new Google Shopping Campaign structure
Splitting out the campaign into more manageable chunks such as by brand is obviously the most logical way to start, especially as you can set different bids depending on what level of the campaign you are viewing. This can be done by clicking the plus button within each level of the campaign until you get to the product IDs themselves. Once clicking the plus button it’s possible to select how you want to split out the products that fall within the product set. Ultimately once everything haws been split out by brand / category down to the products level your campaign will look something like this:
Campain view of broken out Google Shopping Campaign
The above example is obviously infinitely better than the previous reporting and target options using the Google PLA campaign. Another useful feature of this reporting is that you can view the product types and categories as a group without having to break down into the product level reports. So immediately you can see which brands or product types are generating the most traffic and sales for you. Also it’s easy to change the bids on a number of items within a target, which again is a nice feature.
One of the main features of this new reporting is the benchmark stats, which gives you a summary of other merchant’s CTR and average bids for similar products. This of course is always going to be useful as you can see how you compare to your competition. This however is fraught with danger as we can see this being a catalyst for merchants to ignore what their own stats are telling them to do and perhaps bid up to get above the average which could be damaging to their own performance. This information is useful but shouldn’t be regarded as the benchmark for deciding the bidding strategy for your products. The bids should be set in view of your own performance.
What are the Issues / Problems with this new system?
In the interests of balance, here are the issues that we’re concerned with and mainly our catalyst for avoiding using this new Shopping Campaign format for now.
1. No API connectivity. Currently the only way to view these stats is through the AdWords interface. This means there is no way of re-ordering and aggregating the results to view in other platforms. Furthermore this means we can’t change bids from our Onefeed platform, where we’d want to make bulk changes through the API. For us we use the API for everything we do with Google so any lack of integration with their platform means we have to carefully consider how we manage accounts.
2. Inability to order the reports by product and cost. This means that you have to expand every product grouping to view the product stats. With no order by product it’s easy to miss products that are perhaps running away with costs and not converting. This might not be such an issue for small feed merchants with less than 1000 products, however for big feeds it’s a nightmare trying to find the products that might be spending large volumes.
3. No filtering of reports. Along the same lines as the point above; with a lack of filtering it’s possible for larger merchants to have a report that would be so big it would break your browser. Furthermore with large numbers of products in a filtered group as soon as you open the filter the report would be enormous and virtually impossible to navigate. With no filters you’re going to get frustrated going through lines of product IDs. Expect to see plenty of unresponsive browsers.
3. No Auto-bidding and auto-removal options. There is no means of controlling what you’re spending on the individual targets as the budget is set at the campaign level. As has been identified by several merchants using this system, products can easily run away and get large clicks and no conversions. With no means of automatically stopping these items merchants are resigned to go through these reports on a far too frequent basis. As there are no target assigned budgets the money could be spent in areas that are perhaps not as profitable as others. Essentially feed managers are going to have to spend a lot of time in these reports identifying the bad products and bidding down on them.
4. Reporting by ID. As feed managers we are obviously not going to understand what the products IDs mean for each client. In most cases clients themselves won’t understand what a number means either. Reporting by just an ID number means almost nothing in these reports. We’d need to see what the product title is to understand how we wish to proceed with bid changes.
5. No product price information at the product level. This may seem a trivial point, but when all you have is an ID number to gauge what CPC bid you wish to set it makes for a very frustrating experience. The price of the item is massively going to dictate what CPC you set. The only meaningful way of finding out what your IDs relate to is to go into the Products tab at the top of the page, or to open it in a new browser. Either way you do it you’re going to need to cross check every item to see what you are working on.
6. Lack of graphical reporting. As with all other AdWords features the graphs were a really great feature where you could see the effect your bid changes were having on performance. With the current report you’ve got no idea whether bidding down has had a positive effect or not. Often we see the effect of bidding down too much having so much of an effect that the clicks almost completely dry up. It’s a fine balancing act of keeping traffic up whilst keeping the ROI at the required level. Likewise when bidding up we have often seen a negligible effect for some clients on traffic volumes, in essence we just increase costs with no increase in traffic. All of these scenarios are just not evident in the reporting without the aid of graphs.
7. Adding new products not automated. When you add a new category or product to your feed you’re going to need to add it to the campaign otherwise it will either not go live or sit in a catch-all product target. This may not seem too much of an issue, but if you don’t have a catch-all campaign new items will get no visibility. Likewise if you treat your catch-all target as bucket for lesser products with lower bids, again you’re going to struggle to get traffic on new items.
The above points are areas that we hope Google will address and I’m sure in time these features will be available in one form or another.
