Why should I care?
Does it matter whether I know to which products my search terms direct customers? Yes, if you want more insight into your customers and want to do a better job of meeting their needs (and increase profits). In e-commerce, we don't meet our potential customers face to face. They rarely fill out survey forms and we know very little about them. We typically only have a few small pieces of information available to optimize our SEO and Adwords strategies, increase revenue and decrease costs.
When you understand what your customers are looking for with each product, you can optimize your product titles which leads to significant performance gains.
How to figure out the relationships between search terms and products.
Here are 3 techniques to figure out which search terms in your Adwords report match with each particular product. There are pros and cons to each:
- Technique 1: Pro – it’s easy to set up. Con - it’s time-consuming to manage.
Google allows you to create product groups containing a single product. Reporting on search terms per group then is a trivial exercise. However, if you have a catalog of thousands or tens of thousands of products, there is a lot of hours of work in setting this up, and even more when you try to optimize your reverse keywords and PPC strategies.
- Technique 2: Pro – it’s accurate. Con – it’s very complicated
This technique involves measuring semantic similarity between each search term and every product in your catalog. Then using a stable-marriage algorithm in descending order of similarity to link search term to product until that product group’s Total Clicks matches the total in the report. A complicating factor is that Google does not report on all search impressions or clicks.
- Technique 3: Pro – Less effort than the other Techniques. Con – it’s a compromise.
This technique looks for all the products that the search term could have applied to, rather than the exact ones that were actually clicked. If the product title, description, bullet points, and other fields contain all the words in the search term, then it could have applied to that product. You would say there was a correlation. For example, if you have 3 T-shirts that vary in size but are otherwise identical (and the search term does not include size), Google could have shown the customer any of them. Yet from an optimization perspective, it really doesn't matter which they clicked on.
My Recommendation: Technique 3 is the best approach for amount of effort compared to the amount of useful information extracted.
I know which search terms match which products in my feed. Now what?
Once you found all the individual products related to a specific search term, you can calculate the minimum/maximum product price, the product price curve for the search term, and average product price.
Now you can:
- Create reverse keywords for different campaign priority/cost levels related to the product prices and margins. This allows you to minimize your advertising costs while maximizing profit per click.
- Make a list of the most frequent and best-converting search terms for each product that are not yet in that product title. These terms can be used to increase SEO performance by filling in missing content that better aligns with searcher intent and buyer intent.
- Increase Search Term Specificity. Search terms that may apply to many different products indicate less intent to buy than a search term that can only relate to one exact product. Search Term Specificity can better optimize conversion rates, even when the number of clicks is too low for a mathematical analysis of performance to be statistically relevant.