Four tactics used by advertisers to help accelerate sales growth
By: Raghvendra Mani, Analytics and Media Manager
In a recent study in the European Union Five (United Kingdom, Germany, France, Italy and Spain), we found that there are four key areas that top performers from 2013 to 2020 used to help drive glance views and accelerate sales growth.
Story highlights:
We analysed around 3,000 brands that were already advertising in Amazon’s store and expanded into new regions, specifically to the European Union Five (United Kingdom, Germany, France, Italy and Spain) between 2013 and 2020. For each advertiser, we took their launch of Sponsored Products as the base date and then measured cumulative sales and glance views from months 0 through 6 and months 7 through to 12. Using this data, we created five clusters and then compared Cluster 1 (advertisers with the highest number of glance views and percentage sales increase) to Cluster 5 (advertisers with the lowest number of glance views and percentage sales increase).
We looked into which actions led to Cluster 1 seeing, on average, 16.4X higher year-over-year sales growth and 3.8X higher glance views.
Indexed success metrics performance
(Baseline = Cluster 5)
Growth in sales
Growth in glance views
For more on how we collected our data, see the Methodology section at the end of this article.
1. Top-performing advertisers use Sponsored Products and Sponsored Brands, and have more customer reviews
Insights
Over 80% of advertisers belonging to cluster 1 saw their number of customer reviews increase in the time frame analysed. Further, Cluster 1 advertisers had both their Sponsored Products and Sponsored Brands campaigns on, for a higher median number of weeks than other advertisers.
Recommendations
If you’re a vendor: Use the Amazon Vine programme. The programme was created to provide customers with more information including honest and unbiased feedback from some of Amazon's most trusted reviewers.
If you’re a seller: Register with Amazon Brand Registry and use the Early Reviewer programme to help boost the number of reviews of your products.
2. Top-performing advertisers run campaigns throughout the year
Insights
Always-on campaigns can help improve campaign performance, as the algorithms gain more insights over time to improve the campaigns. Cluster 1 advertisers, on average, have Sponsored Products campaigns that are continuously live for a duration of 3X compared to lower clusters.
Always-on advertising performs best when awareness, consideration and conversion tactics are implemented together. Awareness and consideration tactics, such as display ads through Amazon DSP, can help encourage new customers to move down the funnel; and conversion tactics, such as display remarketing, may help them decide to purchase.
Recommendations
When selling products in multiple countries, advertisers should consider the peak and non-peak periods in each country individually. Many advertisers, including top performers, invest in always-on campaigns during off-peak periods and run seasonal advertisements during on-peak periods. To maximise campaign planning (for each locale), you should consider an always-on campaign approach off-peak and tailor your on-peak advertisements to the seasonal calendars of each region.
3. Top-performing advertisers use negative keywords
When launching in a new international region, keywords tailored to the locale performed better.
Negative keywords are words or phrases that prevent your ad from appearing on shopping results pages that don’t meet your performance goals. Using the right keywords and avoiding negative keywords can help ensure that customers are able to find your products when performing searches.
Here is an example of how negative keywords work:
| ASIN catalogue | Negative keywords | Rationale |
|---|---|---|
| Patio furniture | Dining room furniture | Ad won't show when shoppers look for dining room furniture |
| Patio furniture covers | Sofa furniture covers | Ad won't show when shoppers look for sofa covers |
| Both patio swing sets and sofa products | Sofa products in campaigns promoting swing sets | Ads promoting the sofa product line won't show next to the ads from campaigns promoting patio swing sets |
Recommendations
When launching in new marketplaces, you can source keywords from product reviews. Use phrase match to reach large audiences and balance it with exact match to improve product discoverability. Test, learn and optimise keywords for your products.
Advertisers with access to the Amazon Ads console can use the keyword localisation tool to translate to their language of preference belonging to EU5 countries, helping to remove language barriers that may have hindered their ability to launch manual-targeted campaigns in the past in another region. To get started, advertisers can go to the advertising console, launch a new campaign, select manual targeting and scroll to suggested keywords. There, they will see relevant translations below each suggested keyword.
4. Top-performing advertisers combine advertising with locale expansion and product launches
Insights
An Amazon Shopper Behavior Study in 2018, run by CPC Strategy, found that 80% of Amazon customers use Amazon to discover new product and brands.1 The study also found that early advertising in new locales may help with product discovery, and in turn an increase in product sales.
84% of the advertisers within this study that opted to use Sponsored Products or Sponsored Brands to support a product launch or new locale observe sales growth within the first year. Cluster 1 advertisers, on average, launch their first Sponsored Products campaign within six days from entering a new region. These advertisers also tend to adopt other products (Sponsored Brands) faster than other advertisers. The median number of days for Cluster 1 to launch Sponsored Brands is 40 days.
Recommendations
Once you have launched your first campaign, keep learning with us. Chat with our specialists and come to our intermediate and advanced webinars on targeting, budgets and bids, keywords and reporting to get trained on improving your Amazon Ads performance.
Before adding products to your Amazon store, consider using A+ Content to showcase your brand story and product features by using rich text and images on the Amazon detail page. Building your brand on Amazon can help drive conversions, and potentially increase traffic and sales.
Methodology
We analysed about 3,000 advertisers in the Hardlines category selling beyond their home countries and expanded to at least three countries amongst UK, Germany, France, Italy and Spain from 2013 through to 2020. We excluded advertisers originating from China for this study.
We aggregated the total sales and glance views (GVs) for each advertiser, in months 0 to 6 and months 7 to 12 post-Sponsored Products launch. We then created a composite score by measuring the growth in sales and glance views between those two time periods. We then identified the top advertising and retail strategies that corresponded to increased composite scores with machine learning algorithms.
How are advertisers distributed across the clusters?
We used machine learning algorithms to automatically classify advertisers into clusters based on their advertising and retail attributes.
Cluster 1
Cluster 2
Cluster 3
Cluster 4
Cluster 5
How does the clustering work?
We created a binary composite score using a combination of return on ad spend (ROAS), year-over-year growth of retail sales, and year-over-year of product page views. We labelled advertisers that ranked in the top 50% across all three components as “one,” and others as “zero”. We then applied an XGBoost classifier to identify which features and with which weights best predict these labels. In doing so, we considered advertising or retail actions as features such as ad product usage intensity and mix, timing of advertising support, tactics of targeting, creatives and placements, customer review counts and ratings, percentage of products with quality product pages and the types of products promoted in ads, etc.
Using the identified features and weights above, we then applied a k-medoid clustering algorithm to classify advertisers into clusters. Note that we classified advertisers by their actions rather than by the components of their composite score. Next, we ranked the final clusters by their composite scores from high to low.