We are not far into the year 2025 and it feels like we are heading for a repeat of 2024 when it comes to the retail and ecommerce markets. 2024 was hard for everyone in the retail sector, we saw some of the usual suspects doing well and some bigger names taking a hit. With yet more uncertainty around the corner, its more vital than ever that brands search for every competitive advantage they can, to improve fortunes and make the most of even the smallest opportunities.
The ecommerce space has always been competitive, with new challenger brands emerging all the time and frequent races to the bottom, when it comes to pricing. There are so many levers when it comes to marketing, advertising, pricing, merchandising and productisation - juggling those is hard and monitoring success or failure, even harder.
Combine these complexities with current market conditions and the ever-increasing expectations of online customers. Each of your customers has an expectation of what their ecommerce experience should be, across sales, delivery, support, marketing and returns. keeping track of those touch points is vital to not only earn a new sale, but keep customers returning time and again. So how are some of these companies finding success? It’s in the data….
Making the case for analytics
So, what has analytics got to do with everything discussed so far? Data is a vital tool for understanding anything and everything about your business, so key decisions can be made and issues resolved before they do any harm. It’s the one of the reasons why some brands won out in 2024 as it can be such a differentiator.
Below we break down some important analytical insights and talk about how a brand might leverage each.
What’s good?
Keeping track of what is working well in a commerce organisation is vital to sustain performance and encourage growth. Being reactive and agile in this space if far more impactful that typical legacy reporting techniques, where brands will slowly pull together reports once a month and attempt to react to issues potentially weeks after an event occurred, meaning opportunity can be lost or missed.
Product Performance
The question asked here is "What products are doing well, and why?". This enables brands to focus in on such products and look at ways to either replicate that success with other products or further incentivise potential customers to purchase those stand out products.

The optics we can put over product performance are:
- Revenue: How much money the product actually made
- Order Count: How many baskets did the product appear in
- Launch Status: How long did it take for a product to first sell
- Repeat Purchases: How many times does the product sell again in a time period
Lifting these metrics up to a product category, brand, SKU or geolocation level can provide compelling evidence and direction for brands so they can forward plan marketing, purchasing and development strategies.
Why this is valuable: Understanding which products are resonating with customers can directly inform product development and marketing efforts. By identifying patterns in successful products, businesses can strategically allocate resources to replicate success, increase profitability, and ensure long-term relevance in a highly competitive market.
Lifetime Value
The concept of Lifetime value (LTV) in commerce analytics is a crossover of both customer and product analytics. As the name suggests, LTV focuses on several levers where a brand wants to understand the value of customer overtime.

The insight here becomes granular and precise when you start to overlay products, category's, brands and regionality features, which can reveal some discerning trend information regarding the behaviour of customers after first purchase. The sweet spot is finding the products that see loyalty, growth and future purchases. Understanding which of these features produce higher or lower lifetime values helps to facilitate data-driven decision-making in customer retention and acquisition strategies.
You can read more about LTV in our focused blog post.
Why this is valuable: Knowing which customer segments are most valuable over time allows for more effective customer acquisition and retention strategies. Brands can invest in acquiring high-LTV customers while nurturing existing ones to maximise profitability.
Customer Demographics
Most brands will have a general idea of who its customer are, in some case they will cater to a very specific demographic, in others the customer landscape could be highly diverse. Brands collect 1st party customer data, which can be used to understand how customers interact and behave via the brands various sales and marketing channels. This information is valuable but can be greatly enhanced by augmenting that data with 3rd party sources such as Acorn segments or The Data Refinery’s Open Data Sets.

So how can a commerce brand with a broad demographic customer base start to understand who its customers actually are? By using 3rd party data, a picture of customers world beyond brand interactions can be created. Leveraging such 3rd party data enables brands to do the following:
- Understand regionality and the geo-demographic make-up of a customer base.
- Identify trends in the 3rd party data that enables growth into similar demographics.
- Understand typical household characteristics.
- Create more useful segments and marketing materials.
This concept is explored in much greater detail in our previous blog post
Why this is valuable: Augmenting first-party data with third-party insights allows brands to refine their targeting, expand into new demographics, and improve their messaging to resonate more effectively with different segments of their customer base.
Marketing performance
Built a killer ad, campaign or outreach strategy? Great, but how do you know it’s truly working? Measuring the success of your marketing efforts is essential to ensuring you’re maximising return on investment (ROI). Analytics allows brands to track key performance metrics such as click-through rates (CTR), conversion rates, return on ad spend (ROAS), and customer acquisition costs (CAC). With the right data in hand, ecommerce businesses can identify which marketing campaigns, channels and tactics drive the most engagement and sales.

Advertising platforms provide this data as standard, combining this data with actual figures from sales and finance systems, means you can see
Why this is valuable: By continuously measuring and optimising marketing campaigns, businesses can allocate their marketing budget more effectively, focusing on what works and eliminating waste. This helps increase customer acquisition, retention, and overall brand visibility while ensuring a sustainable marketing strategy.
Future Forecasts
Knowing what is (potentially) just around the corner, is obviously a super useful place to be. For ecommerce brands, looking back at historic data can un-earth really important trends in product, purchase and customer behaviours that can provide concrete evidence of what lie ahead.
The challenge faced by most brands is that typical reporting tools found across the ecommerce stack don't have in built forecast capabilities, and building this capability requires Data Science expertise. Inspecting a historic view in one tool might not produce obvious trends without overlaying information from other tools or channels. This is where The Data Refinery's ability to ingest years of historical data really comes into its own.

Getting ahead of seasonal and emerging trends ensures brands don’t miss out on potential opportunities, it also helps de-risk crucial decisions such as stock production or purchasing.
More importantly, what's bad?
Encouraging growth by grasping opportunities early is great, but putting the brakes on a major issue is often more important, this is even more relevant today given the stagnant nature of the current commerce market. Most businesses might use typical year on year (YoY) metrics to track performance over time, but what happens when a business is under performing? How do you find out why that’s the case?
The Data Refinery offers a top-down view that allows users to drill into specific levers that contribute toward overall business performance. This view can be further filtered by region, site and channel, to isolate the levers further.

We break performance down into the following groupings:
Revenue
At the top level, knowing the general direction of travel when it comes to business performance ensures action can betaken immediately rather than at the end of the week/month/quarter. By tracking revenue in real-time, brands can identify shifts in buying patterns and respond to changes swiftly. If revenue is down (or up), then being able to follow a breadcrumb trail to find out why, means the root cause can be identified.
Orders
Two factors that can have a dramatic impact on revenue are the number of orders received and the average value of those order(AOV). Whilst simplistic at a high level, there can be multiple reasons why this might be the case. When order volume dips, it could indicate that potential customers are finding your offerings less compelling, or customers are shopping elsewhere. While a drop in AOV could mean that customers are spending less per transaction, potentially because one product is driving popularity, but there is no complimentary.
Average Order Value (AOV)
In an ideal world a brand wants both Order count and AOV to increase over time, but depending on the goods sold its possible to receive less orders but have much higher order values. The value of an order can be impacted by both the number and value of items it contains. Brands must find the right balance of item price and the potential for cross sell and basket expansion to improve overall AOV. An example might be, that by lowering an items price, AOV actually improves, because the lower price encourages customers to purchase additional items, that push the basket size up and thus a marginal gain overall. By identifying the products that drive higher AOV via either price point or follow on sales, businesses can tailor their product offerings and promotional strategies to boost their overall sales revenue.
Customers and Purchase Patterns
Understanding customer behaviour is crucial, especially when performance is under pressure. The number of customers in a period will obviously impact order volumes. If your analytics is suggesting there are fewer returning customers, then digging into previous purchases patterns to understand a possible reason or by triggering a marketing prompt, may help to encourage a return. Alternatively, if the number of new customers is trending down, then it may suggest a potential product or marketing issue where a price alteration or change in campaign strategy might be the source of the issue.
Knowing that there is an issue by looking at the numbers (e.g. revenue) can trigger an action to improve, but knowing whereto start means there is a much higher chance of the course of action being positive.
AI to the rescue?
Well kind of…
AI now provides some interesting use cases when we talk about commerce analytics, here are some applications of AI that can help users do more with their data:
- Predictive Analytics: AI can analyse past purchasing behaviour and forecast future trends, helping brands make more informed decisions.
- Customer Segmentation: AI can help brands segment their customers more accurately by identifying patterns in purchasing behaviour that may not be immediately obvious.
- Personalisation: AI can be used to tailor product recommendations, promotional offers, and content to individual customer preferences, improving engagement and sales.
At the Data Refinery, we use AI in a few different ways. The primary use case is data accessibility, where our users can ask questions of their data much like an interaction with Chat GPT. This provides a really simple way for user to get answers without needed to become data experts. Take a look at our AI Blog post to learn more.
Good old facts for the win?
The output of well thought out ecommerce analytics can often be seen as validation of what a brand might already understand about its current position. The difference between a gut feel and actual factual data can be huge for a business when making crucial decisions, data provides confidence and lowers the risk of any key decisions made.
Even if your analytical output is isolated to a single system or a report that has been built in a spreadsheet, having some form of intel that can detect an issue or point to a success can give brands pointers to investigate.