What Are You Doing With “Best” Traffic?

By — 07.24.15

Best Products Traffic

In light of Google’s announcement last week around improving the mobile experience to assist shoppers with conversational searches, we want to share the results of some preliminary analysis and recommendations from this, regarding “best traffic”. In an effort to better assist searchers when they are looking for top-rated products, such as a query for “best blenders for smoothies”, Google is starting to show newly-designed shopping ads with authoritative rankings and product ratings. In their own analysis from Q1 2015, Google found that this new ad type is resulting in click-through-rate increases of up to 11% for retailers for these queries. Here is an example of what this ad type looks like:

Ad Example

Internet traffic varies as much as the people entering queries. In order to serve the most relevant ads, PPC managers should work to understand the intent shoppers embed in their queries to appropriately value traffic and gain actionable insights about their market.

As PPC managers work to meet retailer goals, we often strive to identify keywords that provide immediate and efficient return on investment. While reviewing keywords across accounts, we begin to identify new consumer intent trends. One such trend centers on searches for “best” products. When customers begin their shopping process for a new product, they sometimes act to identify which items or models are considered the “best” in class.

This growing trend led us to analyze the intent of shoppers utilizing “best” in search queries. There are many different query-modifying opinion words seen in PPC traffic, but we felt that focusing on the word “best” is useful within a broad examination of shopper intent because “best” is qualitative and should apply similar meaning to queries across varying industries.

We made a prediction about shoppers who precede their queries with the word “best” and ran a quick analysis to see whether our hypothesis was on the right track. We hypothesized that, compared to the average consumer, shoppers who precede their query with the word “best” are more likely to be in the early stages of the shopping funnel. These shoppers may be window shopping, gathering information for a later decision, or comparison shopping.

We predicted that clicks from exact-match keywords preceded by the “best” modifier would convert at lower rates than clicks for the isolated exact-match keywords without this opinion modifier. We also predicted that clicks from exact keywords preceded by “best” would generate revenue at a lower return on ad spend (ROAS) compared to the isolated exact-match keywords which didn’t contain this opinion modifier. We figured this would hold true even when utilizing mixed-click performance metrics.

Our analysis compared the performance of 10 exact-match keywords with the performance of the same 10 exact-match keywords preceded by the “best” modifier. We controlled for average position, Quality Score, and CPC across the two groups. Here’s what we found:

Q1 2015 % Difference

Based on this preliminary analysis, our hypothesis seems to be accurate for conversion rate because the CVR difference between these two sets of keywords was statistically significant. We see non-significant differences in AOV and ROAS between these two groups of keywords, which provides intriguing fuel for further testing. We will be running a more detailed analysis across multiple retailers and attribution windows to test our original hypothesis that these shoppers are in the early stages of the shopping funnel. Please note that these results will not be the same for every retailer.

High-level recommendations for Retailers:

  1. PPC account managers should perform an analysis around searches for “best” products to gain new insights about their traffic and their market.
  2. Retailers looking to attract shoppers in the early stages of the shopping funnel may want to utilize ad text and landing page text that provides relevant information for consumers in the early stages of the shopping funnel. Some examples include: product benefits, comparison tools, and customer reviews. In our case, both the CTR and CVR were lower for “best” keywords which may indicate that both the ad and landing page were not optimal for the consumer intent.
  3. Retailers looking to attract shoppers in the early stages of the shopping funnel should message unique value propositions and promotions to incentivize return visits. This may be a good time for advertisers to leverage remarketing strategies.
  4. If an account allows for building brand awareness and capturing conversions later in the shopping funnel, make sure to add head terms modified by opinion words such as “best” as new keywords.

The devil’s in the details!

It’s important to note that this was a small scale study and that each retailer is different. For retailers with revenue attribution models which do not favor early funnel traffic due to inefficiency, these terms may not provide the best bang for your buck. Ideally, retailers will have a blend of top and bottom funnel keywords and they will view their accounts over multiple attribution windows in order to capture both the shopper and the buyer. We recommend taking stock of attribution type and window during intent analysis.

Stay tuned for our follow up analysis to this post as we will highlight methods we’ve used to improve performance around queries with preceding “opinion” modifiers, especially given the wider roll out of this new ad type.


Leave a Reply

Your email address will not be published. Required fields are marked *

Time limit is exhausted. Please reload the CAPTCHA.

Scroll to top