Return goals in Google Shopping are different for every retailer. Some e-commerce pros tell us that they’re focused on increasing revenue and less concerned about decreasing spend. Others are more interested in cutting cost while slowly growing revenue.
Any way you slice it, hitting and maintaining your goals are the most challenging parts of managing the Google Shopping channel. Even then, there’s no rest for the weary. Many e-commerce pros tell us that they are doing a good job of hitting their goals, but aren’t sure if they’re leaving money on the table. They want to determine if they can be getting more from the channel and how to move forward.
In all cases, the solution has a lot to do with how you bid your products. Here’s why.
The Importance of the Bid
The most basic purpose of Google Shopping’s structure and management is to assign the right bid to every product in your catalog to achieve your marketing goals.
You could (literally) set a bid for each and every product by hand, but we all know the inefficiency that comes with that approach. Aside from the fact that this would be laborious and nearly impossible to manage without automation, the biggest problem is lack of data. In most cases, there is never enough data associated with one product to intelligently optimize bidding.
That’s why Google provides the ability to create product groups. Nearly all e-commerce pros group their products by Google’s pre-defined method — according to brand, category, and/or product type.
But here’s the challenge with all of these approaches. The right bid for a product doesn’t depend on any of those groupings. It depends on the performance of the product and the goals you want to reach.
In Google’s words: “There’s no one recommended bid amount that works best for everyone. The right max CPC bid for you will depend on evaluating the cost of your products, the type of campaign you’re running, and your profits.”
That said, consider your catalog. Pick a brand. Do the products within that brand range greatly in price? Do those products convert at different rates? Do customers buy some of those products in multiple units per order, and other products in singular units per order? If so, these products warrant different bids.
For Example …
Take a look this example from an e-commerce retailer. The chart below shows AdWords data for 28 products, all within the same brand, from the retailer’s catalog. The brand and retailer are redacted.
Performance metrics vary. Prices range from $19.97 to $1,149. Conversion can be a low as 0.28 percent and as high as 25 percent. AOV runs from $19.97 to $7,797.99.
Based on these metrics and a target cost/sale for each product, an estimated bid is provided. As you’d probably expect, the estimated bids vary because the performance metrics vary.
The retailer would have to set the same bid for all of these products if it grouped them together. Call it 50 cents. Now take another look at the estimated bids needed to hit the cost/sale goal. The retailer would be bidding incorrectly on every product. It would come close (within 10 cents) on only four products.
Looking further, the retailer would be under-bidding on 14 of the 28 products by 10 cents or more. It would be over-bidding on 10 of the 28 products by 10 cents or more. That leads to a pretty inefficient campaign.
Go Beyond the Typical Taxonomy
This example is not an unusual case. We see situations like this every day from retailers of all sizes and in all segments. If you’re not bidding your products based on performance — and continually updating those bids as needed — you’re probably bidding many products incorrectly. As a result, your products might not be rising to the top of the Google Shopping channel and capturing all of the available revenue opportunity or you might be wasting spend on the wrong products.
The best way to set bids is to let the performance metrics tell you what those bids should be. And that involves thinking outside the (retail) box.
- Break out individual products. Test what happens when you assign a unique bid to an individual product. Pick a popular product within a given group in your catalog, and break it out into its own group (a group of one). Use the data you have to assign it a bid, and re-evaluate the bid often. See if the product has higher revenue and lower cost/sale this way. Try this exercise with a few products.
- Create custom labels. You can create up to five custom labels in your data feed, and use those custom labels as attributes to create more specific product groups — and therefore, assign more specific bids. For instance, some retailers create custom labels for top sellers, products on sale, etc. But keep reading, because there can be pitfalls here.
- Avoid big-bucket groupings. Custom labels are a step toward achieving more unique bids. But even a custom label — such as “adolescent gift items” — can contain products that vary in performance. You’d still end up with products that are overbid (meaning you’re spending too much budget), and products that are underbid (meaning you’re missing revenue opportunity).
- Get granular. The best way to create custom labels is to base them on specific metrics. For instance, one custom label could be products with a conversion rate under 2 percent, another label could be products with a conversion rate between 2.01 and 4 percent, and a third could be products that convert above 4 percent.
- Use your most crucial success metric as a custom label. Product margin. Cost of sale. Return on ad spend. Customer lifetime value. Profitability. Identify your most crucial success metric and use it as a custom label to help focus your campaign on what matters most to you.
- Build product groups using multiple custom labels. Using multiple custom labels lets you create groupings that contain products which are more closely related. For instance, you could create a product group containing items that are, say, high heels (product type), convert above 4 percent (custom label), and have an AOV between $50 and $100 (custom label). Products that share many of the same performance metrics tend to deserve similar bids.
- Continually update bids. Performance metrics change due to seasonality and product life cycle. An accessories retailer wouldn’t bid high on cashmere scarves in July. Continually evaluate and update your product groupings to ensure you’re always assigning the right bid to every product.
Product bids are a math problem. It’s all science. Some of the art comes into play, however, with product titles, which are the second most important factor in achieving your Google Shopping goals. Optimizing your product titles has several do’s and don’ts to it as well, which we’ll explore in an upcoming post on our blog. Stay tuned.