Sections:
/w=3840,quality=90,fit=scale-down)
eCommerce Marketing Techstack
Website Personalization & CRO Use-Cases/w=3840,quality=90,fit=scale-down){kind=icon}
Data Collection/w=3840,quality=90,fit=scale-down){kind=icon}
Bloomreach Engagement Use-Cases/w=3840,quality=90,fit=scale-down){kind=icon}
Analytical Frameworks & eCommerce AnalysesMerchandising in eCommerce: Product Placements on a Website
Why should eCommerce Web Managers select products on the website carefully?
Managing the product selection on a e-shop’s website is an important responsibility of the website manager. A store, ultimately, is judged by the product offering it has and the price at which they are offered. A visitor in a typical session won’t view more than 2 or 3 products. The decisions of the website manager regarding which products are displayed where have an impact on the bottom line of the company and the browsing experience of the shoppers.
It is a missed opportunity then, that most website managers do not make decisions about their product discoverability with the help of data.
Data driven website Merchandising
Website Managers need to Prioritize Products between multiple competing pressures:
- which products do we have on inventory?
- which products are “Go To Market” → Brand Presentation / Seasonality
- which products do we want to promote strategically (for instance, a brand that has no Women’s products is expanding into Women’s needs to “invest” website real estate into creating awareness about this completely new line of products)
- which products do we have creatives for and which ones should we invest into?
While these are important considerations, none of these pressures take into account customer feedback. The above are all internal company pressures. As a result many e-shops are not promoting products on the website that their customers have told them they like the most. This is especially true in e-shops that have more than 500+ products, where it becomes difficult for any human to keep track of the products. The more products your e-shop has the more likely you are leaving money on the table if you are not applying data-driven website merchandising in your e-shop.
This guide
In this guide we will go through how e-commerce companies can evaluate their products in a data-driven way thanks to the data tracked on the website and then cover 7x website decision touchpoints that website merchandisers can use these data-driven merchandising on.
Data-Driven Evaluation of Product Performance
What KPI do we recommend to follow when selecting products for merchandising?
If you're tasked with making merchandising decisions in an eCommerce business, this Notion page is designed specifically for you.
We argue that solely evaluating products based on the number of units sold can be deceptive and risky. This guide will elucidate our perspective on this issue and offer an alternative approach to product performance evaluation.
How we analyze Product Performance
At its simplest, analyzing the performance of your products is as simple as sorting all your products in a table by their total number of units sold or revenue.
Let’s say you want to analyze Month-To-Date shorts performance based on the revenue. You may be looking at the table similar to this one:
The biggest shortcoming of this analysis is that we analyze only output (meaning how many units of specific products we sold), but we don't see what input was required to get these results.
If we add Product Views (meaning how many times product was viewed on the website) into the picture, we suddenly have an excellent approximation of your total marketing effort across all marketing channels for the specific product.
We can see that “Shorts 2” and “Shorts 3” had way less exposure compared to “Shorts 1”. We can see that “Shorts 2” and “Shorts 3” have way better “Product CR%” metric, which is simply the ratio between “Unique views” & “Units Sold”.
Now the question is, how much more money we would make if “Shorts 2” and “Shorts 3” had higher exposure, and “Shorts 1” had less exposure?
If we know how much profit we make when we sell one unit of a particular product, we can calculate a metric called PPV (short for Profit per View).
Now let’s do the math here. Let’s say that instead of promoting “Shorts 1” & “Shorts 4” across all marketing channels (Email, better exposure on the Website, Paid Facebook, Organic Facebook,…) we would promote “Shorts 2” & “Shorts 3” instead.
Let’s say “Short 1” would receive 8,000 fewer Unique Views, “Shorts 4” 5,000 fewer Unique Views, and “Shorts 2” and “Shorts 3” would receive 6,500 more Unique Views. What would that mean for our bottom line?
Product | Change in Unique Views | Change in Profit | Old Profit | New Profit |
Shorts 1 | -8,000 | -$3,920 | $6,037 | $2,117 |
Shorts 2 | +6,500 | +$9,295 | $5,578 | $14,873 |
Shorts 3 | +6,500 | +$6,110 | $4,003 | $10,113 |
Shorts 4 | -5,000 | -$2,250 | $3,427 | $1,177 |
Total | 0 | +$9,235 | $19,045 | $28,285 |
Organic PPV
When excluding paid channel Product Views and Units Sold, we would end up with the Organic PPV metric.
Based on this kind of analysis website managers can make data-driven decisions about which products they promote on the website. The Organic PPV is directional as we can expect products that have a higher Organic PPV to bring in more profit per view.
What are the ways a website managers can use Organic PPV to increase the profitability of their store?
Web Product Decisions Points & Tactics
In the following section, we’ll cover common touchpoints on the website that can be optimized with the help of digital analytics, why they are important touchpoints and the recommended analytics method for making the best product decisions.
1. Sorting Products on Category Pages
Category pages on e-commerce shops almost always contain more products than is available to show at the top of the page. That digital real-estate is even smaller when you consider the same for mobile. That is why it is important to consider the default order of the products that belong to the category.
That is why we recommend implementing a process where every week the category pages are re-ordered based on the performance of Organic PPV of the products in the last week. The report would only include the products from the category in question. The products within the category that has the highest PPV last week would be ranked 1st, 2nd, 3rd and so on.
Data driven sorting of products on category pages ensures that the products most prominently displayed on category pages - are currently the ones that makes the most profit for the bottom line.
The process of ordering products on a category page can be very manual if an e-shop has a lot of categories and/or a lot of products in categories. As a result this is often accomplished semi-automatically or automatically by uploading a CSV document into the Web platform (such as Shopify) - with rank values for each product within its category. This rank values are then used by the website to rank the products on the category page.
The benefits of this practice should be multi-fold. The effectiveness and attractiveness of the category pages should be improved by showing the best performing products for that category for previous week. That will also provide a sense of “freshness” for customers who return frequently. Overall, this should improve the CR% and Total Revenue achieved from the same amount of browsing customers, compared to not changing the category page frequently or doing so based on intuition.
2. Sorting Products on Sales Pages / Custom Landing Pages
This is very similar to Point 1, except that Sales Pages and Custom Landing Pages typically mix products from various categories on a single page. As a result the PPV analysis would need only include products that are on the category page. That is fairly simple in Point 1 as that only requires the filtering of products based on the category field. In the case of a custom landing page or a sale, a CSV of selected products for the custom landing page or sale page needs to be compiled. From this CSV on the “back-end” of your dashboards, you would create a specific product attribute with the values either TRUE / FALSE for the product depending on whether they were selected for the landing page.
Based on this attribute the product within this selected list with the highest Organic PPV during the Sale period - products should be ranked 1st, 2nd, 3rd and so on. In case of Sale events we would recommend monitoring the performance of the page hourly and refreshing the order of the products daily or maybe twice a day, depending on the volume of sales you are experiencing and the length of the of Sale event.
- If we assume that the total cost of the product sourcing and shipping for the Slipper Pink Cats is 40% of the original 25 EUR, (10 EUR) at the discounted price of 14 EUR the profit is 4 EUR per unit sold.
- If we assume that the total cost of the product sourcing and shipping for the Slipper Pink Cats is 50% of the original 25 EUR, (12.5 EUR) at the discounted price of 14 EUR the profit is only 1.5 EUR per unit sold.
- If we assume that the total cost of the product sourcing and shipping for the Slipper Pink Cats is 60% of the original 25 EUR, (15 EUR) at the discounted price of 14 EUR the is losing 1 EUR per unit sold.
When marketers design sale events discounts on products without the knowledge of the cost of the products (both sourcing and shipping) then in some instances the company actually make a loss on each sold unit of a product that has been discounted too steeply. If possible invest into ensuring you have good data about product costs on product level.
3. Sorting Product Subcategories + Variants on a Category Page
Certain e-shops choose to display their category page as a rows of subcategories of the given category page. In the screenshotted example below, a fashion e-shop oriented on shoes, structures their category page based on the “styles” of product available within their category.
In the example below we can see the category of “Women’s Slippers” presenting a different style per row with final color variants of each product style as the product they offer. This is unlike the simpler category page design where all products from a particular page are simply ordered regardless of their subcategory, mixing.
In this case the data-driven performance sorting of the Styles and Products takes a different shape to accommodate the different structure of the category page.
First, the Organic PPV of the styles (subcategories) within the category needs to be evaluated as a whole, comparing all of the product variants of one style PPV against one another. The styles and their respective “carousels” with the highest “overall Organic PPV” in the last week would be ranked 1st, 2nd, 3rd and so on.
Second, the Organic PPV of the colors (variants) within the style (subcategory) need to be evaluated within the context of the style against one another. Then the colors with the highest “overall Organic PPV” in the last week should be ranked 1st, 2nd, 3rd and so on within the carousel of the product style.
4. Selecting the category “Hero” product
In many e-commerce shops, categories of products are represented by an image of a product from that category. The selected product that represents the category is known as its “hero product”. Many web marketers underestimate the value of optimizing the choice of the hero products. This is a common oversight. The choice of category to browse within is often the first real choice a browsing customer makes on an e-shop. For 1st time visitors, selecting a category they would like to browse within is one of the first touchpoints; where they see the kinds of products the shop is offering.
That is why we recommend implementing a process where the hero product selection is re-evaluated every 1 or 2 weeks. In a Filter on products only within the particular category. Shortlist the top 3 ranked products in terms of PPV during the past 1 or 2 weeks (depending on your time interval). Pick the 1st ranked product to be the hero of the category unless there are inventory or strategic concerns to override the 1st ranked and go with the 2nd ranked and so on.
It is important to keep the “Hero” product among one of the top results on the category page itself, in case the user clicks through the category page and was interested in that particular Hero product. If your e-shop has a lot of categories and subcategories (10+) we recommend automating the process, importing the PPV Rank of products into the backend of your web
5. Selecting the “Homepage Hero”
Most e-commerce shops invest into creatives of products that are used as campaign “flag-ships” that represent the campaign or season for the shop. These are changed at the frequency of every 2 to 4 weeks depending on the marketing calendar of the brand. The products that receive such prominent placement on the Homepage get a lot of views during the periods of the campaign. The costs of these creatives are not inconsiderable. Additionally, the decision to select the products for the campaigns often have to be done 2 to 3 months in advance of the campaign launching. As a result the, the choice of the product that is invested into needs to be done carefully.
Often these product creative investments are often made without data consultations. There are many strategic considerations that are usually taken into account. A very important consideration is often based on creative direction within the context of the campaign - how a certain product may fit the theme of the campaign or lifestyle communication or seasonal design. Sometimes the products are selected based on legacy reasons - “these products define us”. In the worse cases they are based entirely based on the intuition of key decision makers in the e-commerce marketing team.
While the choice of the homepage hero product is ultimately a brand decision and should be influenced by the creative direction of the brand - it is important to take into account what the data says about the performance of products as well. Once you have a selection of 5-10 products that you are considering to creatives into, check their performance:
- Recent Performance - What is the Organic PPV of these products in the last month in relation to the other 5-10 products? Are there any similar products to the 5-10 draft selected products with high Organic PPV that could be considered for selection?
- Performance during the previous Campaign / Season - Let’s say the campaign is concerning the launch of the summer season during the month of June. What was the Organic PPV of these products during June in the previous years? Are there any similar products to the 5-10 draft selected products with high Organic PPV that could be considered for selection? Are we choosing products from the best performing categories / subcategories during June in the previous years?
In the situation the products selected for the Homepage Hero, is new and there is no historical data for its performance on your website → look for products that are similar to the new products and try to estimate their PPV via their “similar products”. While it is more difficult to estimate what is going to happen with new products - use the PPV reporting to monitor the performance of new products when they are released to be able to change the “Hero” in case it does not perform well and replace it with a product from the new selection that is performing well.
6. Sorting Products Variants on the Product Page
E-commerce websites that have product variants, often have the option to click-through to other product variants of the same product on the product page. In fashion, furniture and electronics e-commerce these are often color variants of the same product.
Often there is high variation of product performance within a product’s variants. It is entirely possible that out of 8 colors, 2 are performing very well, 2 are performing very poorly and 4 are doing around average.
As a result, just like on the product style carousels on the category page - with the help the Organic PPV of the colors (variants) within the style (subcategory) order the colors with the highest “overall Organic PPV” in the last week should be ranked 1st, 2nd, 3rd and so on within the carousel of the product style.
7. “Trending Products Recommendations”
Website managers that are responsible for e-shops with large quantities of products make good use of “product shovelers” - placements on the site that display a row of products thematically grouped together. Sometimes they are personalized to the user → based on their transactional history or browsing histories. Often they are guided by simple rules such as - “Best selling in the last 7 days”. However, as we demonstrated earlier, only looking at the sales of a product is like not considering the costs that went into promoting it.
With the help of PPV it is possible to create “product shovelers” that are based on the approach of Organic PPV Rank, within desired category or set of products. These would then automatically give more prominence to products that are performing well from a PPV perspective, without the website’s team intervention. Organic PPV Rank Product Shovelers would be well placed on:
- category pages
- homepage
- cart (pre-checkout)
PPV Product Evaluation Example
The table below is an example of a table a web merchandiser would use to pick a “Hero” Image and the “Ranks” of products. The sample data below are an example of a product selection from the same category, a total of 23 products (A1, A2, B1, B2, etc.) within 6 styles (A, B, C, etc.) over a period of 2 weeks.
It shows the performance of these 23 products across 4 key metrics during that time period
- Unique Units Sold - How many customers have bought this product during this time period?
- Unique Views (Ordered by from Largest to Smallest) - How many customers have viewed this product during this time period?
- Unique Revenue $ - How much Unique Revenue was brought in by this product during this period? How many unique customers have bought this item at the avg. item price during this time period? Meaning, if a customer buys a single product 10x times, it is only included once. This is to avoid PPV being inflated for a product that was purchased a lot by a small number of customers.
- Organic PPV - Organic PPV means simply that Unique Views and Unique Units Sold that occurred in sessions that came through Paid Ad Channels such as Facebook Ads, Google Ads, Bing and TikTok would be excluded entirely from the calculation. This is because traffic generated by these channels are always less qualified and have lower CR% on average than “Organic Channels”. Additionally products that are being pushed hard on Paid Ad channels tend to have degraded PPV, as a result of getting a lot of “unqualified” exposure.
Interpreting the table above, notice how
Product C2 was the most viewed product during this period and also has one of the lowest Organic PPV performances. Product C2 was not the best use of marketing investment.
Product B1, on the other hand has the 4th most product views and the highest Organic PPV, making that Product B 1 a win for marketing investments.
Product D2 didn’t have a lot of views - 6th from the bottom, has also the second highest Organic PPV at 6.10$. This is a great candidate to promote more going forward. Product F 2 and Product F 1 have a similar story.
A website e-commerce manager that uses data-driven merchandising methods would pick the “Hero” product for this Category to be Product B1 as it has the highest Organic PPV for the time period. The rank of these products on the category page would go in this order:
- Product B1 ($6.14), Product D2 ($6.10), Product F2 ($5.72), Product F1 ($5.39), etc.
- The product may not be well stocked, or soon out of stock
- The product does not have previous performance history as it was not stocked before. Additionally, there is a strategic interest to get more attention to this new line of products.
Product Evaluation & Website Merchandising Checklist Google Sheet
Feel free to use the following Google Sheets
- Product Evaluation Table: as a template for calculating PPV for your eCommerce products. In addition
- Web Merchandising Checklist: as a template checklist for seeing how data-driven your touchpoints are at the moment
If you found this post valuable…
Should you find this post insightful, we encourage you to share it with friends and colleagues who might also derive value from it.
Additionally, we offer a complimentary 45-minute consultation if you're interested in delving deeper into the topic of merchandising decisions.
Thank you for investing your time in this read. We hope the information provided proves beneficial to you!
On this page
- Merchandising in eCommerce: Product Placements on a Website
- Why should eCommerce Web Managers select products on the website carefully?
- Data driven website Merchandising
- This guide
- Data-Driven Evaluation of Product Performance
- How we analyze Product Performance
- Web Product Decisions Points & Tactics
- 1. Sorting Products on Category Pages
- 2. Sorting Products on Sales Pages / Custom Landing Pages
- 3. Sorting Product Subcategories + Variants on a Category Page
- 4. Selecting the category “Hero” product
- 5. Selecting the “Homepage Hero”
- 6. Sorting Products Variants on the Product Page
- 7. “Trending Products Recommendations”
- PPV Product Evaluation Example
- Product Evaluation & Website Merchandising Checklist Google Sheet
- If you found this post valuable…