Shopping campaigns can be pretty tedious to set up, to say the least. Up until recently, there was very little functionality for shopping campaigns in the AdWords editor and there are still features to be desired. It isn’t unusual to come across campaigns that have been structured with broad groups of products (and it can be tempting to structure campaigns this way to try to save time!) but a hyper granular structure pays off in dividends. Not convinced? Read on!
The Background
Our client is a premium retailer for a multitude of brands. The site is e-commerce but a high percentage of sales are made by phone. When we took over the account, there was one PLA campaign and it was structured in such a way that each brand had one ad group. In this structure, all of the like-branded products were housed within said ad group.
The PLA campaign was one of the top performing campaigns but we believed there was still opportunity for growth with an improved structure.
The Hypothesis
- Separating each brand into separate campaigns allows us to have more control over transaction revenue and ROAS in instances where budgets need to be capped because some brands have higher margins than others. It also helps us to maximize visibility based upon inventory.
- Building out each product as its own ad group would make analysis and optimizations quicker and easier, which would positively affect performance.
The Action
We built new campaigns for each brand. Each campaign was structured so that every product was contained in its own ad group – excluding all other product targets. This allowed us to write ad copy specific to each of the products. Beyond that, it allowed us to see so much more data at a glance, without downloading and analyzing tables from the dimensions report so we could make quick and easy changes.
*Extra Bonus – Using Acquisio, we’re able to pull in call data as well as transaction/revenue data so that we can see the most import KPI’s in one place and make adjustments. The set up is not a necessary component of this structure but it was undeniably useful since calls are key to campaign success.
Because our original PLA campaign generated so many calls and transactions, we were very eager to mitigate risks. We phased in the new campaigns by activating a campaign and pausing the corresponding ad group in the original campaign. Once that campaign’s performance stabilized, we launched the next brand and paused the corresponding ad group.
Analyzing the Data
We took the data from two weeks after the restructures and compared it to the two weeks prior.
Here’s what we found:
- Traffic increased 102% after opening up budgets. Conversions (calls plus transactions) increased disproportionately at a combined rate of 578% .
- The cost per combined call + transaction decreased 15%
- The conversion rate increased 17%
- Our click-through-rate decreased, which was a surprise since our ad copy was more tailored than before. Among all of the metrics, though, CTR is one that we believe can be easily remedied in the future.
Conclusion
All of our primary KPIs showed improvement after the restructure. Furthermore, we expect the campaign performance to continue to improve as we make optimizations. We haven’t yet been in a situation that required us to limit one brand’s budget in favor of another but we feel confident that the budget control will be beneficial in the future.
How do you structure your PLA campaigns? We’d love to hear about your learnings and success stories in the comments!
Amy Bishop
Latest posts by Amy Bishop (see all)
- DSA Series Part 2: 6 Tips for Making the Most of Dynamic Search Ads - March 30, 2015
- Are Dynamic Search Ads Right For You? Part 1 DSA Series - March 23, 2015
- Case Study: Restructuring Shopping Campaigns To Improve Performance - February 10, 2015
- Restructuring PPC To Improve Performance by Geography: A Case Study - December 18, 2014
MamboMan says
Two weeks isn’t enough, especially with such low volume.
Amy Bishop says
Thanks for reading, MamboMan. These are early results, you are right. I’m happy to report that the results have remained consistent since this case study was compiled on January 27th. Although the campaign is still new (now 4 weeks in), we continue to see the same efficiency with CPA, which I would expect to be stable by now.