Google Shopping is now the most profitable channel we work with. The conversion stats and cost of sale figures are often quite unbelievable and is therefore now setting the standard by which other services are measured. Google Shopping isn’t by any means easy to manage or get fantastic results from. A lot of hard work needs to go into managing things, so much so that it can become an all-encompassing channel that requires a lot of monitoring and dedication to make it work to its’ optimum.
Having compiled the ‘best practice’ list below of the top issues/areas of improvement that we feel are the most important considerations in managing Google Shopping Campaigns. If you feel you’ve got all these bases covered then you can safely say you’re probably getting the most from Google Shopping:
- Daily Budget Strangulation
One of the most common best practice issues we see when taking over Google Shopping Campaigns is unrealistic daily budget setting. Google Shopping Campaigns with a fixed daily budget that is often maxed means the campaigns can be down for hours each day with no traffic coming through. Usually, this means the highly valuable evening traffic that converts well is missed or the ads during the course of the day are filtered so severely it’s pot luck when you might show. We’ve seen some merchant run out of money by lunchtime meaning they’ve lost a significant number of potential orders. This low daily budget is usually aligned with very high bids as the old client managers weren’t seeing the traffic or conversions they wanted. The most effective solution is obviously to increase the daily budget and most importantly reduce the bids to ensure the new budget isn’t maxed out again. Clearly, it’s not always going to be possible to increase the spend on a poor performing campaign so adjusting the bids is a better option.
- Mobile Bid Adjustments
Mobile bids adjustments have been the saviour of a lot of campaigns. A campaign without a mobile optimised site can destroy a very good campaign and cripple the ROI. Whilst mobile sites and optimisation is the biggest subject in the e-commerce world right now the ROI stats from these devices are incredibly diverse and often require a lot of adjusting. For many with no mobile site, the conversion stats on these devices make your eyes water. We’ve sadly seen a significant volume of wasted money which once adjusted has made a fairly poor campaign into a very profitable one. Whilst mobile does play an undoubted part in the customer journey, there’s little to be gained by getting traffic on a device which does your site no justice.
- Time of day scheduling / Bid Adjustments
In line with the point above adjusting your activity and bids to reflect your shoppers’ buying habits makes for a much better-optimised campaign. For many B2B clients, we set large bid deductions outside of business hours which make an incredible difference on ROI. For several other clients with very limited budgets, time of day scheduling also helps to keep the products in play at the right times. For many merchants, an uplift in traffic in the evenings also has an effect, so setting small bid increase adjustments for just this time. The dimensions tab in AdWords is a very useful tool in identifying opportunities for bid adjustments.
- Product Level Bidding
The vast majority of clients we speak to are still using the categorisation tree in the Shopping Campaigns to set their bids. Whilst this is the format Google has specified it’s a fairly unsophisticated way of setting your bids. The most common means of breaking down the products is by brand within a category. So you’d effectively have all Nike trainers in the Trainers category. The trouble with having all Nike trainers bid at this level means they all have the same bid. You’re not telling Google which items you want to promote and likewise which items should get less traffic. If all items have the same bid and very similar keywords, Google will choose the item with the best click through rate and show this item more frequently as it is regarded the most relevant result. Sadly this doesn’t always mean the one that converts the best gets shown.
There is, of course, a means of adding in product level bids to the Google system, however, this relies on a rather cumbersome lookup by ID in another report to get the item ID you want to promote (who honestly can say they know what product each ID relates to?!). Changing bids like this is often quite a long winded process. Within the Onefeed management system, we have the facility to set these product level bids in a quick and easy report which also reports on the ROI, clicks, impressions and sales data so you can make a very quick and informed decision on what you want to change. These bids feed through the AdWords API straight into the Google tree and are updated within seconds.
- Automated Bidding Adjustments
Another very important feature for best practice that everyone needs to be using is some level of bid automation, especially for items that are not converting. There are lots of bid management tools out there right now, but very few have a feed management solution as part of the process. Wherever possible it’s vital to ensure product level performance reports are run at least once a day and relevant remedial action are taken. When there is no automation in place it’s very easy for a product to blow a lot of money very quickly, so much so that the overall campaign performance can be damaged severely.
Whilst we’re not big advocates of automated bidding up, we are very much of the opinion that bidding down is a vital ingredient in any management solution for Google Shopping. Due to the very nature of the Google bidding platform, you can simply bid down to a level that is profitable or one that means the traffic is localised more to the long tail searches. The Onefeed system has a bid optimiser that runs daily to bid items down based on their performance. Needless to say, this takes a lot of the headache out of poor performing products.
- Filtering Shopping Campaigns (by margin/product type)
The common theme for a lot of campaigns is to break out the campaigns by product type or margin so they can be better maintained and managed through more robust budgets. For best practice and larger accounts, this makes it an easier way to manage large volumes of data which in turn allows for a more granular budget setting approach. Google now allows filtering of campaigns by feed attributes such as the custom labels, categorisation or brand. The really important benefit to this option of filtering is the facility to set negative keywords a lot more specifically to a range of products. The negative keywords can only be set at an adgroup level which is pretty useless when all your products appear in the same adgroup, hence the requirement to break down campaigns to several filtered collections.
- Keep Track of Bid Changes
Sadly Google doesn’t have any graphing functionality on products in the Shopping Campaigns, which means advertisers are pretty stuck at seeing what effect their bid or product optimisations have had on CTR / click volume or sales. What is required for best practice, is a very long winded approach of comparing date ranges to see the effects. Let’s be honest, that’s really not a feasible way of working especially on a large scale. The Onefeed platform has an option of graphically showing the effects of the bids/price changes have on the clicks/sales etc through a triple time graph. Effectively we can compare the click volume and CPC bids with the sales that were achieved to see what the greatest bid range should be. Ultimately it’s all about finding the ‘sweet spot’ for your bids, which as many of you will know is where all products need to be. The last thing any merchant wants to be doing is pushing the bids up too much so that the costs spiral and likewise lowering the bids too much so that the traffic drops off. It’s a real tightrope!
- Custom Labels
In line with the filtering options for campaigns with best practice, there is also a much more useful feature of incorporating up to 5 custom labels into the feed. These can be anything that helps you set the bids at the level you want. We’ve found the most useful way of using the labels is to provide a better classification of products where the Google taxonomy isn’t very granular. There is no limit to the data that can be used to populate the labels. Another use of these labels is to sub-divide products into more specific areas of perhaps profitability or stock level so that the bids for all these products can be better managed. The labels are driven through the feed and require 5 new columns to be added. Within Onefeed we can set rules or structures in place to create these labels automatically.
- Keyword Optimisation of Titles
The hardest thing for most merchants to get their head around with Google Shopping is the lack of keywords; that is the principle association with Google. All advertisers that have used Google for years are ingrained with the notion that Google uses keywords to drive traffic. In the case of best practice for Google Shopping, the keywords in play are those that are derived from the Product Titles that you supply in the feed and can’t be ‘specified’ in the AdWords interface. Using this fundamental reality it’s imperative that product titles that we are using are optimised as well as they can be for the searches that we want to show for. In the past year, we’ve made a big shift towards targeting specific search terms within the product titles that have yielded some excellent results.
As with many merchants there are always going to be search terms that result in the most traffic conversions. Taking this into account with best practice, there’s no better way to manage the campaigns. A recent addition to the Onefeed system is the product title A/B testing feature. This allows us to compare and contrast various different product titles for each product with a view to get the most exposure for the keywords that convert efficiently. This is an automated process that will rotate between titles giving us the most effective title to use. Ultimately the goal is to find the title with the greatest CTR which can mean we can pay less for the click as Google knows the product is highly relevant. This in turn means we can bid less and get a better ROI. A good title can improve the CTR by 3-4 times the old title. When combining the title adjustments with a coherent bidding strategy on these items, makes for the biggest possible return.
- Keep an Eye on Product Price Changes
Most merchants are actively keeping their pricing as competitive as possible. The trouble with this is that often these changes are automated and the person managing the performance, or ourselves, aren’t privy to the price changes, especially when an item has dropped in price significantly or has gone on sale. With automated bid systems, when a product gets optimised down in bids, these items won’t really get much of a chance to get traffic for a while, if at all. The impact here is that items that have changed price could well have been optimised out, or down, and stand no chance of selling. What can also happen is an item may only be getting steady levels of traffic and isn’t pushed as hard as a similar item. In the Onefeed system, we’ve recently added a price change reporting tool that shows all items that have changed price within a date range. Price change reports are vital in isolating the items that should be pushed and increased in bids. Quickly finding an item that has gone on sale can give you an advantage over the competition and steal a march on the promotions.
- Generic Search Terms / Bid Tipping Point
One of the nuances of the Google Shopping feature, especially the PLAs, are the generic search terms, and when you show them. I think everyone has at some stage questioned the result that you get for the product images on the Google search (PLAs). Occasionally the items that are shown aren’t really what was searched for, likewise, the same merchants show for a lot of the searches. This is often down to very high bids being placed. We’ve seen some huge volumes of clicks when we bid up too much and often it means some items are showing for wrong search terms or for terms we don’t want to be showing for. As has been mentioned above there is a very fine line between bidding too much or just right. You can easily go too high and beyond the tipping point where the clicks avalanche through and the ROI is trashed.
As has been seen significantly on Google Text Ads, generic search terms are great at driving traffic, but often poor at converting sales as these clicks are from browsers and not the ones likely to buy.A lot of this can be controlled through careful bid optimisation and monitoring at the product level using our graphs. A good example of where the effect is most felt is when someone is doing a brand search for an appliance, let’s say a ‘Hotpoint washing machine’. Clicks on this will be very high and less likely to convert. For best practice, a more specific search where perhaps a model number and brand is used is much more likely to convert. Keeping a very close eye on the search terms reporting can pay a massive dividend as it will often show some very high clicked search terms that yield very little return. Adding negatives on these terms can make a big difference to the ROI.
- No excluded campaigns/products
The final point we’d like to make with best practice is around the bidding options open to all clients. There are still significant volumes of clients out there that simply exclude product types / brands / products purely as they don’t see themselves as competitive or perhaps unlikely to convert. We don’t see any benefit in removing products from the feed completely or excluding them from the Google Shopping Campaign. As has been mentioned before Google allows you to set bids right down to 1p. Instead of removing the items from the feed, why not simply set the bids for these items to 1p. There are several ways of doing this which is perhaps a little misunderstood. By creating a new campaign for the products and filtering them or by using the custom labels it’s easy to get all these items grouped together and bid right down.The critical thing about these items is that if someone is really hunting these items out they will get found and more importantly convert. The most popular exclusion rule we see is for items under a certain value being removed. All that is required is the bids to be set on these items in relation to the value. You’d be very surprised at how well these types of products do especially considering their bid level.
For best practice, it’s vitally important to ignore the feed. Rejected items or disapproved accounts are the first port of call and it is to be assumed that these feed issues are already in hand.