eCommerce data and analytics are essential to understanding and communicating insights needed to drive optimal decision-making and sales.
Brand manufacturers and retailers of all sizes use eCommerce data and analytics for various goals such as:
Improving shopping experiences;
Doing competitor analysis;
Defining a pricing strategy;
Optimizing the stock availability and distribution.
The 3 Vs. of Big eCommerce data
eCommerce has developed rapidly over the last 3 years. eMarketer forecasts that online shopping will continue growing, representing 20% of total retail sales in the U.S., nearly 40% in the U.K., and more than 50% in China.
Brands and retailers have had little choice but to transform following the eCommerce revolution and the resulting big eCommerce data and data-driven way of working.
However, marketing questions and decisions have remained unchanged since data has become the "new oil" in commerce.
But what has shifted with the emergence of the era of data?
Data are collected daily and driven by 3 essential properties, which we call the 3V's:
Volume - The vast amount of available data Some statistics that illustrate data volumes: - In 2020, each minute, $1.000.000 was spent online. - Also, in 2020, Amazon shipped 6.659 packages every minute. - In 2022, eCommerce analytics software SiteLucent monitors over 2.000.000 product detail pages daily.
Velocity - The speed at which new data is being produced As an eCommerce professional, you probably encounter that the digital shelf changes daily, sometimes even multiple times a day. Think about a sudden price drop or stockout event.
To give you an example: It's needed for SiteLucent to scrape millions of data points 8 times every 24 hours.
Variety - The types of data we have available Data is being collected at high volumes and speeds and can also be collected from various sources or covering multiple topics.
For instance, those millions of eCommerce product pages being monitored daily by SiteLucent, in turn, contain dozens of data points with which hundreds of different metrics can be specified, related to eCommerce topics such as: - Assortments - Prices - Stock availability - Search rankings - Ratings & Reviews - Product content, including images and descriptions - Buy box wins and losses
The rapid growth of data makes it difficult for eCommerce professionals to know where to begin and distinguish between valuable and useless data to extract insights. More insights can be uncovered when different data types and sources are combined, which makes, at the same time, eCommerce analytics more complicated.
The eCommerce Analytics Process
The process of answering practically any question in eCommerce (and marketing in general) follows a strategy that involves the following steps:
Define a challenge for your eCommerce business: For instance, a brand notices a decrease in the in-house share (the brand's market share, based on turnover) and wants to find out what is causing the decrease and how to make it grow again.
Find out what data is required to address the challenge: Types of eCommerce data are clickstream data (visits and behavior), conversion data (sales and cart abandonment rates), digital shelf data (product and consumer review data), and other related data categories used in eCommerce analytics and marketing. Data can be structured (e.g., prices) and unstructured (review texts).
Gather analytical techniques and tools: The era of big data goes hand in hand with an abundance of analytics tools — Tools that offer lots of interesting (but also useless) data. Choosing the right eCommerce analytics tool is essential to uncovering valuable insights.
Translate eCommerce analytics into insights and recommendations: Raw data is meaningless unless it can be transformed into an easy-to-understand format by creating visuals. Then, translate data and visuals into a narrative, uncovering important insights and recommendations to flow from the analyses.
Examples of how to move from eCommerce Analytics to Insights
Example 1: Stay aligned with your brand's pricing strategy
The challenge Manufacturer Batho Online, a fictive company in bathroom accessories, has implemented a price increase for its brands Batho and Silky for online resellers and wants to know the effect of the increase on the average selling prices in the market (of both their own brands as competitor brands). Do Batho's online prices still align with their eCommerce pricing strategy?
Required data To answer this question, we need daily selling prices for the past 6 months for 165 SKUs of their 2 own brands and 9 competitor brands on 15 retailers.
The techniques and tools eCommerce analytics tool SiteLucent collected the required pricing data using web scraping techniques.
Insights and recommendations
The graph below shows that Batho's average consumer selling price increased by 15% between October 1st, 2021, and March 31st, 2022.
Silky's average consumer selling price increased in the same period by 9%.
Both resulted in an average price increase of 12% for the brands in 6 months.
We do not see the same increase when looking at retail prices of competitor brands during the same 6-month period, shown in the graph above, except for one exclusive higher segment brand (which was excluded from the above-shown average).
Actions/recommendations: Batho online discovered that the increased selling prices don't align with their brand's pricing strategy. Based on the above insights (and decreased retailer sales figures), they can infer that the brand is pricing existing customers out of the market. They can eliminate the competition by negotiating margins with retailers, backed by the pricing data, and bringing back consumer prices below the average.
Example 2: Customer reviews and how to act on them
The challenge Music&I, a fictive consumer electronics manufacturer, wants to know what decisions could be made in response to customer reviews on Amazon.
Required data Average review amounts and star ratings of their brand and 4 competitor brands. Review texts of their brand.
The techniques and tools eCommerce analytics tool SiteLucent collected the required review data, and SiteLucent InSight services provided custom visualizations such as a benchmark table, and a scatter plot to quickly uncover underperforming products in terms of reviews.
Insights and recommendations As shown in the below table, Music&I's brands need more reviews to match the market's average.
In the table and graph below, the outlier products with a low number of reviews and low review ratings are visualized by red dots. These products need attention first.
The below graph illustrates the top 100 unique words in review texts and the number of times customers used them.
Actions/recommendations: - Incorporate the frequently-used 'customer language' in product page content to increase the chances of showing up in top retail search results. - Zoom in on underperforming products and early-stage reviews first, and run a review campaign to gain reviews for the products without, or with low numbers of, reviews. - Find out what shoppers say about products and if something is amiss with the product or the product content. Read here how Sony used customer review data collected with SiteLucent, to optimize the product content of a newly introduced product!
Example 3: Improve product search placements by developing a targeted SEO strategy
The challenge Brand manufacturer Music&I wants to know if products are in the top 10 retailer search results and, if not, what they can do to rank higher per retailer.
Required data Digital shelf data: e.g. search rankings, # of images, # of videos, # of characters in the product title, presence of keywords in product titles, # of reviews, and average star ratings.
To define the right keywords and SKUs that need attention, retailer search volumes, CTRs, and conversion rates per keyword are needed.
The techniques andtools eCommerce analytics tool SiteLucent helps collect and visualize the required search and digital shelf data (see below graphs).
By using web scraping techniques, required data can be collected.
Digital shelf data and regression analysis tell which content factors are important for search rankings specific to each retailer. Benchmark analysis and visualizing the data using graphing tools make it more understandable and actionable.
It is recommended to use platform-specific keyword research and other performance tools. Music&I gathers platform performance data such as search volumes, page traffic, and conversion data.
Actions/recommendations: First, select relevant keywords with preferably high search volumes and relatively low competition.
Scope Focus first on: - Underperforming brands and channels (the red-marked ones in the table below);
- Products with low page traffic and high conversions with which you can achieve the most impact; - Categories/SKUs with high-profit margins, flagship, and sponsored products.
Don't end up improving things that are not important to the retailer search algorithm. For retailer X, we found that # of characters in the product title and # of images impact the search rankings positively.
Use SiteLucent's Digital Content Completeness (DCC) score to uncover missing yet important content criteria and start optimizing!
What are your eCommerce analytics challenges?
Do you have eCommerce challenges, but you don't know where to begin or how to extract insights from your data? Your SiteLucent InSight team will help you uncover the answers!