Merchandising in eCommerce is the practice of controlling product visibility across categories, search results, and product pages. It determines which products users see first and how those products are prioritized.
Table of Contents
- What Merchandising Actually Is
- The Core Merchandising Techniques
2.1. Manual Boosting (Rule-Based)
2.2. Algorithmic Sorting
2.3. Personalization
2.4. Contextual Merchandising (Query & Category Intent)
2.5. Inventory-Driven Merchandising
2.6. Promotional Merchandising
2.7. Cross-Sell & Upsell Placement - When Each Technique Breaks
- How Merchandising Works as a System
- Real-World Constraints
Different techniques are used to achieve different goals, such as improving conversion, promoting specific items, or aligning results with user intent. Because these goals can vary by context, merchandising is best understood as a set of approaches rather than a single method.
In this post, we’ll look at the main merchandising techniques and when each one is most effective.
1. What Merchandising Actually Is
In ecommerce, merchandising is the process of controlling how products are presented and prioritized across the site. It defines which products appear first, which are more visible, and how product lists are ordered in categories, search results, and recommendations.
At its core, merchandising is about ranking and placement. Every product list, whether it’s a category page or search results, is not neutral. Someone (or something) decides the order.
This decision can be driven by different factors:
- relevance to a query or category;
- popularity or conversion rate;
- business priorities such as margin or inventory;
- user context or behavior.
Merchandising brings these factors together into a single outcome: what the user sees first.
In practice, merchandising sits at the intersection of product, marketing, and search. It translates business goals and user intent into concrete product visibility decisions across the entire site.
2. The Core Merchandising Techniques
2.1. Manual Boosting (Rule-Based)
Manual boosting is the most direct form of merchandising. It involves explicitly controlling product positions using predefined rules, such as pinning items to the top of a category or prioritizing specific products in search results.
Instead of relying on automated ranking, this approach gives full control over visibility. Common implementations include:
- fixed positions (e.g., “always show in top 3”);
- boost rules (e.g., “increase ranking score by X”);
- inclusion in featured sections or curated lists.
Manual boosting is typically used when there is a clear business priority. For example:
- launching new products;
- promoting high-margin items;
- supporting marketing campaigns;
- highlighting seasonal assortments.
Its main advantage is precision. It allows guarantee visibility for specific products regardless of performance data.
However, this control comes with limitations. Manual rules do not scale well in large catalogs, require ongoing maintenance, and can conflict with relevance if overused. As the number of rules grows, managing them becomes increasingly complex.
2.2. Algorithmic Sorting
Algorithmic sorting uses data to automatically rank products instead of relying on manual rules. Products are ordered based on signals such as popularity, conversion rate, revenue, or other performance metrics.
Common sorting logic includes:
- best-selling;
- most viewed;
- highest converting;
- top revenue generators.
This approach is most useful in large catalogs where manual control is not practical. It allows rankings to adjust over time as user behavior changes.
The main advantage is scalability. Once set up, the system continuously updates product positions without manual input.
At the same time, algorithmic sorting depends on data quality. New products may struggle to appear due to a lack of history, and strong performers can dominate results even when they are not the best fit for a specific query or category.
In practice, algorithmic sorting works best as a foundation layer that can be adjusted with other techniques when needed.
2.3 Personalization
Personalization adjusts product ranking based on individual user signals. Instead of showing the same results to everyone, the order changes depending on behavior, preferences, or context.
Typical inputs include:
- browsing history;
- past purchases;
- location or device;
- affinity to brands or categories.
This approach is most effective for returning users, where there is enough data to influence ranking in a meaningful way.
The main advantage is relevance at the individual level. Users are more likely to see products that match their interests, which can improve engagement and conversion.
At the same time, personalization has limits. For new users, there is little or no data. It can also reduce product discovery if the same types of items are shown repeatedly.
In practice, personalization is usually applied as an additional layer, not the primary ranking logic.
2.4 Contextual Merchandising (Query & Category Intent)
Contextual merchandising adjusts ranking based on query or category intent. Instead of one logic, priorities change depending on what the user is looking for.
For example:
- “cheap” favors lower prices;
- “best” favors ratings or popularity.
The main advantage is better alignment with user expectations.
In practice, it acts as a layer that adapts ranking logic to different contexts across search and category pages.
2.5 Inventory-Driven Merchandising
Inventory-driven merchandising adjusts product visibility based on stock levels and availability.
Common use cases:
- boosting overstocked items;
- suppressing low-stock products;
- hiding out-of-stock items.
The main goal is to align product ranking with inventory needs.
In practice, this works as a supporting layer that helps balance user experience with operational priorities.
2.6 Promotional Merchandising
Promotional merchandising adjusts product visibility during campaigns and marketing activities.
Common use cases:
- highlighting discounted items;
- boosting products in seasonal sales;
- supporting campaign-specific assortments.
The main goal is to align product ranking with active promotions.
In practice, this is a temporary layer that should be easy to apply and remove without affecting baseline ranking.
2.7 Cross-Sell & Upsell Placement
Cross-sell and upsell placement focuses on recommending additional or higher-value products alongside the main item. This can include complementary products, higher-priced alternatives, or bundles that extend the purchase.
The main goal is to increase AOV (average order value).
In practice, this is applied on product and cart pages as a separate layer from main listing and search ranking.
3. When Each Technique Breaks
Each merchandising technique works well in specific conditions, but none of them is reliable on its own. Problems usually appear when a method is applied too broadly or without considering context. This is why merchandising works best as a combination of layers rather than a single approach.
| Technique | Scales Well | Needs Data | Can Override Relevance | Main Limitation |
|---|---|---|---|---|
| Manual Boosting | ❌ | ❌ | ✔️ | Hard to maintain at scale |
| Algorithmic Sorting | ✔️ | ✔️ | ✔️ | Favors existing winners |
| Personalization | ✔️ | ✔️ | ✔️ | Weak for new users, limits discovery |
| Contextual Merchandising | ✔️ | ✔️ | ✔️ | Depends on correct intent signals |
| Inventory-Driven | ✔️ | ❌ | ✔️ | Can conflict with user expectations |
| Promotional | ✔️ | ❌ | ✔️ | Temporary but can affect baseline |
| Cross-Sell & Upsell | ✔️ | ✔️ | ✔️ | Low impact if not relevant |
4. How Merchandising Works as a System
Merchandising is not a single technique applied in isolation. In practice, product ranking is the result of multiple layers working together.
Each layer adjusts the output of the previous one. Instead of replacing each other, techniques are combined into a single ranking process.
A simplified version of this system can look like this:
- Base ranking (algorithmic sorting);
- Contextual adjustments (query or category intent);
- Business rules (manual boosting, promotions);
- Inventory signals;
- Personalization.
The order matters. Earlier layers define the foundation, while later layers make targeted adjustments.
For example, algorithmic sorting provides a scalable baseline, contextual merchandising aligns results with intent, and manual or promotional rules apply specific business priorities. Personalization, when used, is typically applied last to fine-tune results for individual users.
5. Real-World Constraints
Merchandising is limited by factors that are not always visible in a clean model.
Large catalogs introduce scale issues. With thousands or millions of products, it is not possible to manually control visibility, and even automated systems struggle with long-tail queries that have little or no data.
Data is often incomplete or noisy. Conversion rates, popularity, and other signals can be unstable, especially in low-traffic areas. This makes ranking less reliable than it may appear.
Different teams may have conflicting goals. Marketing, product, SEO, and merchandising can prioritize different outcomes, which leads to competing rules and constant adjustments.
Technical limitations also play a role. Ecommerce platforms and search tools may restrict how flexible ranking logic can be, especially when combining multiple signals.
These constraints do not prevent effective merchandising, but they shape what is realistically achievable.
