How to triple the ROI and the Google Shopping traffic in 6 weeks [case study]
In this case study I will teach you how an online store managed to increase the Return On AdSpend (ROAS) by 300% in just 6 weeks time.
Note: ROAS is calculated as: revenue from adwords / cost of adwords.
This ratio will tell you how much revenue you make for each dollar spent
in AdWords.
The shop was already running AdWords for a while, but the performance, in terms of ROAS, was not good enough. They reached out to us because they needed help from an ecommerce specialized agency.
The entire process consisted of three stages:
Improve the feed without the need to change product descriptions;
Get more traffic on well performing product segments;
Cut costs on bad performing product segments, location, devices and keywords.
In this case study I’ll walk you through the exact step-by-step process used for each step above and how we achieved a 300% ROAS boost in just 6 weeks.
1. Getting started
Before we started working on the account we had to understand the account’s status and what would the target be (also taking the industry into consideration).
What we got was:
An account which had a ROAS of 1
A target ROAS of somewhere between 3 and 4
Fairly good traffic on the website (paid and organic)
Note: Whenever you are fixing a target, it should be realistic
and reachable. You also have to keep in mind that a high ROAS
means a lower overall revenue.
In the start we had made sure that all the data we need is visible. We did this by managing the columns and this is what we used:
With this set of columns you can have a view at how much you are willing to pay for a click (Max CPC) the number of impressions and clicks, how much are you actually paying for each click (Avg CPC), the total cost of the clicks, how are you doing compared to your competitors (Benchmark CTR), how often are your ads shown (Search impression share), how many clicks have converted (conversion rate), and if your product group/ad group is profitable or not (ROAS).
2. Improvement to the shopping feed
There are multiple feed management tools out there to help you optimize the feed. We used Shopify Simple Google Shopping Feed App to improve the feed for Google Shopping to improve the data quality.
We have done three things which resulted in higher impression share and more targeted traffic.
1) Include all variants and not only the main products
We optimized the product feed and added all product variants instead of keeping only the main products. This also included all colors and sizes.
This resulted feed item number grew from 700 to 7000.
Not only there were more products available, but also searching for "red lipstick" and "orange lipstick" would display different images in Google Shopping.
2) Improve findability of the product with titles
The products consisted of the brand, base title and options. Options are the color and the size.
We modified the title sent to Google Shopping for better findability. We included the following template: brand + title + options (color and size).
This resulted in much better findability across the product catalog.
Before: product title
After: brand, product title, color and size
Next we wanted to test if the product type (like blush or lipstick) would make a difference. There is a limit in number of characters that are allowed in the title (soft limit is 75 and hard limit is 150), so we needed to be careful before adding too many keywords.
3) Add GTIN numbers for better approval rates
Google requires GTIN (EAN, UPC codes to identify each product) for cosmetics category. These numbers were not always filled in for each product.
Fortunately the Shopping feed plugin generates unique GTIN numbers for each product if they are missing. This way we were able to approve 20% more of the products.
3. Optimization process
1) Subdividing the product groups
In order to have a better understanding of the account we have subdivided the product groups to a more granular level. Because each group has more individual products that we could comfortably handle, we chose to divide them by product type.
2) 80/20 sorting and calculating target
At this point we took into account the 80-20 rule of thumb (Pareto principle): 80% of the budget will go to 20% of product groups. This means that if you sort the product groups by cost, the first few groups will get most of the budget that you have allocated for the campaign.
After having a better view of the costs, we calculated the target CPA. In order to calculate it use this equation:
Now, that we had the target CPA value and the rest of the info we needed, it was time for bid management.
3) Grow the winners & spend less on the losers
After sorting the (already subdivided) product groups by cost, we used filters to assess their performance.
TIP: Use Max CPC > 0 as an extra filter. This will only leave you with products in the list and exclude the product categories, and you won’t apply the same changes twice for some products.
Note: when increasing the Max CPC, go with small increments as 10-15% and see how it goes. Also, you shouldn’t worry about the daily budget when doing this. In case the budget will burn too fast, Google will let you know that the campaign is limited by budget. If your budget is too small for advertising every product, just exclude the poorly performing ones and use your budget for the so called ‘money makers’ (best sellers) and use the profit later on for the other ones.
(A) Money makers (Best performers) - Max CPC +10%
Time frame: last 30 days
ROAS > target ROAS (regardless of the cost)
This filter will include only products that are intensively searched for, and are reaching or exceeding your targeted ROAS, which means you earn more money than you spend on advertising them. For this category, you can increase the max CPC by 10% or even 15%.
(B1) Question marks - Max CPC +10%
Time frame: last 30 days
ROAS < target ROAS and Cost < target CPA
If some products are not performing well (don’t have too many impressions, clicks, or don’t have conversions), but you didn't spend too much budget on them, you can just increase the Max CPC by 10% just to see if there are any ‘unicorns’ amongst them.
(B2) Potential money losers - Max CPC -10%
Time frame: last 30 days
ROAS < target ROAS and Cost > target CPA
This category will include only the products that, in the long term, cost you more than the average CPA and not performing as well as you would like. In this case you should reduce the Max CPC for these products by 10% or 15%, because they have the potential to lose you money.
(B3) Money losers (worst performers) - Max CPC -30%
Time frame: last 90 days
Low ROAS and Cost > 2 x target CPA
This category includes the products that do not bring revenue and have a high cost over a longer period of time. For these products you should decrease the Max CPC by 30%. It may seem a lot, but if you think about it, these products have spent at least twice as much budget as the other products, have no conversions (or just an abnormally high CPA), and they will only continue to increase the loss.
4) Take control of locations and devices
To find this tab, just go to the campaign’s "Settings" (1), and click on the “Locations” (2) button. Here, you can find the locations you are targeting with your campaign and can adjust their bids.
For the location bid adjustment we have increased the Max CPC for the best performing locations, but not with much. 10% should do it. Similarly, for the locations that were losing money, we had their bids reduced by 10%. Remember: small changes work better than big ones.
For the optimizing the devices bid adjustment (3) you can look at the cost/conversion. If it’s much higher than the other two, you can reduce it by 10%. Keep in mind that your goal is to have a nice and stable ROAS, so do not overdo it.
In this case we reduced the mobile bids by 30% because the conversion rate was low, and increased the bids for computers and tablets, because they performed better.
Moreover, you can have a look at the dimensions tab (4) and look at ‘day of the week’ table and see in which days convert better and increase their bid adjustments.
5) Cut the cost by identifying bad brands and products
For making sure you are spending your budget wisely, you shouldn’t only focus your attention on what you want to spend your budget on, but also on diminishing the spend for things that are eating up your budget. You can find a very good overview in the ‘Dimensions’ tab.
The first overview we used was the ‘Product type dimension’ to have a better understanding which categories of products perform better and which perform poorly.
By having a look at this dimension, we could identify the best performing and worst performing product types and we could use this information when optimizing the campaigns.
Another good overview is the ‘Brand dimension’ and with this, we didn’t only have the categories performance overview, but we can also observe them for each brand.
As we did for the ‘Product type dimension’, we have used the information that we have got from the ‘brand dimension’ to optimize the Shopping campaign.
6) Cut the costs by excluding money draining keywords
Another practice for reducing the budget waste is keywords management. In order to start, we had a look at the ‘Search Terms’ tab. Here, we saw which keywords are performing well and which are losing money and had to be added to the ‘Negative keywords’ list.
Note: the negative keywords are totally dependant on your field of business and there is no recipe for a good list. Each one is unique and tailored for a specific campaign however, this massive list with 1000+ negative keywords can help you find the most frequent negative keywords used for different industries.
Again, you should keep in mind the 80/20 rule of thumb, and arrange the search terms by cost and see which ones had more budget and how did they perform. If you have search terms that have cost you a significant part of your budget and they don’t have any/enough conversion, well...I guess, by now, you know what you’ll have to do.
Another way of assessing how the keywords are carrying out, is by using our AdWords Negative Keywords Tool. With this tool, we visualized the worst performing keywords, we used that info in the account and knew exactly which keywords to exclude (we don’t need that kind of negativity in the account).
With this tool we have easily identified the budget draining keywords and added them to the ‘Negative keywords’ list.
That’s all there is to it. This is the process we used to increase the ROAS by 300% within 6 weeks. The best part is that you can also get the same results by following the same step-by-step procedure.
Note: We have reached an ROAS of 10 by combining these optimization strategies with performace-splitted Shopping campaigns (best performers, in betweeners, low performers) with different priority levels. To find out more about this you can check this article: Google Shopping: Focus on 10% of your products which make 75% of your revenue
We’d also like to hear from you: have you used any of these ideas to optimize your account? Let us know in the comments section down below!
