Ecommerce Analytics Reports: Decision-Driven Data Analysis & Conventions That Mean More Than Benchmarks

The first step in any new digital experience optimization program is to build a strong understanding of the digital journey.

The reason is pretty simple. Whether it’s a software registration experience or an ecommerce path to purchase, our goal is always to identify challenges and present a clear roadmap to address them. But we first need to understand the experience as it is today.

So, how do we get started with “learning” a website and formulating an improvement strategy? We start by querying the data. And no, we don’t benchmark.

Benchmarks are bullshit

Many external advisors offer services to benchmark your data against industry numbers. They’ll deliver findings like:

• Your cart abandonment rate is X%, which is X% above average
• You have an average time spent on page per session of XX seconds, which is X% below average

Supposedly, these benchmarks inform their strategy.

But, as we’ve written about before, benchmarks are bullshit. Public metrics can be faked. One-for-one comparison is nearly impossible. And knowing whether your metrics are higher or lower than a competitor’s is simply not enough to build a decent strategy.

Our job is to help you make decisions that earn you more money and solve user problems. And benchmarks are a red herring when it comes to better digital experiences. That’s why we use a decision-driven query approach.

Using a decision-driven approach to data queries

As we already established, the first step in any new digital experience optimization program is to build a foundational understanding of the digital journey. To start “learning” and setting up an improvement strategy for an ecommerce website, we go through an initial data analysis.

But our approach is probably different than your marketing team’s. We look at everything through the lens of decision-driven data analysis.

Why? Data in itself is not the end-all-be-all. It’s the decisions that you’re able to make with the data that matter.

That means that rather than looking at the data and deciding what to do with that information, we’re looking at the data to make specific decisions. For example:

• Decide who to target for usability testing
• Define the pages, device types, users, and journeys that are the highest priority for optimization
• Understand product mix (by category, by revenue)
• Understand on-site behavior (search, top pages)
• Identify pages and audiences in need of deeper research
• Look for any red flags or indicators of unusual user behavior
• Define key contexts we need to know as we work on a new website (device types, purchase readiness, channel mix)

Instead of benchmarking, analyze with key conventions in mind

So, what exactly are we looking at with this lens of decision-driven analysis? We generally start a digital experience optimization program by reporting on 6 ecommerce analytics categories:

1. Demographics
2. Seasonality & promotions
3. Acquisition strategy
4. Device strategy
5. Site search
6. Products

Along with the lens of decision-making, we run and analyze our reports with some key conventions in mind.

You already know we don’t believe in benchmarks, but we do look at tons of data sets and have certain conventions we use to assess the data. Generally, and I mean very generally speaking, some truisms tend to hold across websites.

One example? Mobile tends to convert worse than desktop. While this truism isn’t completely universal, having seen so many data sets tells us this is more than likely to be the case. When we see a data set that breaks convention, it highlights an area that deserves further investigation. It means something interesting is going on with the unique attributes of your customers or vertical, so we keep an eye out for data sets that break from convention.

6 Ecommerce analytics report categories and key conventions to look out for

All of our data analysis is in service to digital journey optimization strategy. Later in our program, we might report on the complex intricacies of your site. But first, we need to set the foundation of the strategy with decision-driven data analysis.

Let’s take a look at:

• The specific reports essential for information-gathering

  What conventions we look for when assessing a new site

• How these reports and conventions help us craft a foolproof digital journey optimization strategy

1. Demographics: Check assumptions about your audience

Teams often make assumptions about user gender and age, but it doesn’t always match 1-to-1 with reality.

Often, site traffic differs widely from what teams projected in the startup phase. So it can be helpful to check assumptions about your audience with a few reports.

Reports:

• Unique visits by age group
• Unique visits by gender

Questions to ask:

• How do visitor demographics align with our assumptions?
• How might people experience our site differently based on whether they’re purchasing for themselves, a significant other, or a friend?

Callout:

Don’t be surprised if you have higher conversion rates for a gender that doesn’t match your target audience. While most analytics platforms aren’t able to give us complex reports on the gender identities of site visitors, they do often give an estimation of what portion of an audience is likely to identify as male or female. We generally see that when a brand has a target audience that more broadly appeals to one gender, it’s typical to see higher conversion rates for the other, especially around gifting seasons.

It’s also normal to see higher conversion rates among older audiences than younger ones, especially for products where the price point is considered on the higher end for its category.

2. Seasonality & promotions: Uncover when to run optimizations

Understanding seasonality helps you decide when to design, when to build, and when to push code live (or hold off on production changes).

For experimentation programs, it also gives great indicators for when to run tests. You can easily determine when you have enough traffic to run tests and also identify important peaks in traffic for when you may not want to run experiments. For example, many ecommerce brands choose to enter a code freeze from before Black Friday and Cyber Monday through January.

Reports:

• Sessions over time
• Transactions over time
• Average purchase value over time
• Conversion rate over time

Questions to ask:

• Are there traffic peaks and valleys that repeat YoY?
• As traffic increases, do transactions increase proportionately, making conversion rates hold steady? If not, traffic quality might not be steady throughout the year.
• How, if at all, does the average purchase revenue change with time? Taking a look at this information can help you understand if your free shipping thresholds are elastic and whether or not promotional periods impact cart value.
• How do promotions, emails, and campaigns impact sales? Looking at your traffic trends also helps you understand the motivations of your customers. Are they only purchasing when you run promotions? Does traffic go up when you have limited-edition items?
• What offline conditions might be impacting traffic and transaction levels? For example, paint companies often see traffic and conversions increase in the spring and summer months. Identifying that can help you connect with your customers about the time-based urgency of completing their paint job.

Callouts on conventions:

We often see ecommerce site traffic dip starting in October or as early as September. As the holiday season becomes closer, people defer purchases that aren’t urgent in hopes of receiving a discount.

3. Device strategy: Decide to optimize for desktop or mobile-first

Understanding the mix of traffic and revenue from device types is foundational to your strategy.

We look at several factors to determine whether the site is appropriate for a mobile-first or desktop-first strategy and what mindset users might be in depending on the device they are visiting from.

Reports:

• Sessions by device category
• Revenue by device category
• Conversion rate by device category

Questions to ask:

• Where does the largest portion of traffic come from? What about revenue?
• How do conversion rates differ between desktop and mobile?
• How might the price and product complexity warrant these differences?

Callouts on conventions:

In general, mobile conversion rates tend to be about half that of desktop. But as average purchase value creeps up into the $100’s, the difference between desktop and mobile conversion rates tends to stratify.

Conversion rates are often closer together for lower price point products (that are more prone to impulse purchases) and farther apart for high complexity, high price products like home stereo systems and build-your-own customizable products.

4. TOFU vs BOFU: Gathering details on acquisition strategy & purchase readiness

Channel mix is a great place to start when understanding acquisition strategy, what sources of traffic convert best, and what the typical journeys look like.

We use this information to tell data stories: generalized anecdotes that use channel, device, and landing page to group users into a few categories based on brand awareness and purchase readiness.

Reports:

• Channel & source/medium mix
  • Volume and portion of traffic per channel
  • How those channels convert
• Source/medium by device type
• Top landing pages
• Channel & landing page group

Questions to ask:

• How much traffic and revenue is each marketing channel generating? This can help you identify which channels are most effective at converting (versus which channels might be functioning as awareness channels), where to focus testing initiatives, and which customer journeys are the most important to optimize.
• What channels and landing pages have particularly high bounce rates? Isolating low-intent traffic (often to blog or ancillary pages) can help you quickly de-prioritize page types where even a 5x increase in conversion rates would yield a low return on investment.
• Which sources are more likely to attract mobile sessions vs desktop sessions? In the previous step, you already identified which device types convert better. Understanding which channels bring in the high and low-converting device types will help you form strategies for those channels.
• What are the top landing pages, and which channels are likely to direct traffic to each page type? There’s a big difference between a mobile user landing on a product page from an ad and a desktop user landing on the homepage from a direct source. Understanding the interplay of channel groupings and landing pages is essential to a tailored digital strategy.

Callouts on conventions:

It’s normal for paid and organic social not to convert very well. Often, we see that brands use it as more of a TOFU awareness play.

For brands doing CPC, It’s common to see that traffic is generally mostly mobile, and it often lands on product pages. Contrast that with known audiences who might go directly to your homepage, and they are having a pretty different experience!

5. Site search: Establish what shoppers are looking for

We use site search ecommerce analytics reports to understand the prominence of search as a navigation method on-site and which products and categories people are searching for.

Reports:

• Percent site search
• Top search terms
• Conversion rate by use of search

Questions:

• What portion of users are using site search?
• What products and terms are people looking for?
• Are those using search more or less likely to convert?

Callouts on conventions:

For websites with many thousands of SKUs, it’s common to have upwards of 20% of sessions include a site search. But for smaller brands with few products, it’s common to see as little as a single-digit percentage engage with search.

Generally, we see that if a user searches, they are 5-7x more likely to convert than the average user. If conversion rates for users who engage with site search are comparable to or lower than average, you may want to run a heuristic analysis of site search to understand if the results are satisfactory or if they are hindering conversions.

6. Product analysis: What is driving revenue for the brand

These reports outline which products and product categories are purchased in the highest quantities and which account for the largest portions of revenue.

Reports:

• Product sales by revenue & volume
• Product category by revenue & volume

Questions:

• What products sell in greatest quantities?
• What categories of products make up the largest portion of revenue?
• How do popular products map to average order value? Are they higher or lower?

Callouts on conventions:

For websites with many thousands of SKUs, it can be difficult to attribute revenue to a single product. Instead, start by isolating the largest-impact product categories.

What does this mean for kicking off a digital experience optimization program?

The data can tell us so much. So, given the right data, a specialist won’t need to ask many questions at the start of an optimization program. They should be able to simply run an analysis and quickly learn a lot about a website: marketing strategy, user groups, channel mix, etc.

The report and the analysis will all be in service to the digital optimization strategy. It won’t necessarily look like a marketing report because a digital journey consultant doesn’t care about 100% of the same metrics as a marketing professional.

There’s overlap, of course, but while marketers tend to care about efficient spend, looking at customer acquisition cost (CAC), etc, a digital journey specialist from The Good is hoping to understand things from a different perspective. They are trying to understand the people visiting the site today and how to convert them. The metrics are bound to be different, which is where commerce analytics reports can come in.

We aren’t marketers, so what we measure and report on won’t be the same as a marketing advisor. We are here to improve your digital experience, and that starts with a select set of ecommerce analytics reports that we scrutinize through the lens of decision-driven data analysis and key conventions learned from decades of repetitions.