Assortment Planning 2.0: Balancing Data Science with Merchant Intuition

Assortment planning is no longer just about filling shelves with the right products. It’s about creating the right mix for the right customer at the right time—across multiple channels, often under pressure from dynamic market trends. Assortment planning in retail involves increasingly leaning on data science and predictive analytics to guide these decisions. But while algorithms offer precision and scale, the merchant’s intuition remains irreplaceable. The challenge is finding the right balance between the two.

Advancements in machine learning and data analytics have transformed how assortments are built. Retailers now use demand forecasting, customer clustering, and predictive modeling to determine which SKUs to stock and in what quantities. These tools provide powerful insights derived from sales history, seasonality patterns, local demographics, and even social media sentiment.

Data science brings several key advantages:

• Speed and scalability: Systems can generate localized plans across hundreds of stores within hours.
• Accuracy: Predictive models can flag underperforming products or overstock risks early.
• Granularity: Sophisticated tools can segment customer preferences down to a postcode or store level.

However, these models have limitations—especially when they rely solely on historical data.

Where Algorithms Fall Short

Data models, no matter how advanced, are inherently backward-looking. They analyze what has happened to predict what might happen. But the retail landscape changes fast—particularly in sectors like fashion, beauty, or consumer electronics. What worked last year might flop this year due to shifts in trends, consumer sentiment, or macroeconomic factors.

Also, algorithms can sometimes misinterpret outliers. A one-time promotional spike, a supply chain disruption, or even weather anomalies can skew recommendations. This is where merchant judgment plays a critical role.

The Enduring Value of Merchant Intuition

Merchants bring something algorithms can’t: context. They understand brand positioning, upcoming trends, and the nuances of customer behavior. A seasoned buyer knows when to take calculated risks on new product categories—even if there’s limited historical data to support the move.

Merchant input also helps:

• Interpret gray areas that models struggle with, like “aspirational value” or “trend relevance.”
• Advocate for innovation by introducing untested, trend-forward items that algorithms would normally exclude.
• Adapt assortments based on in-store feedback and customer conversations.

In essence, merchant intuition helps soften the sharp edges of data-driven decisions.

Best Practices for Blending Data with Intuition

To succeed in Assortment Planning 2.0, retailers need a hybrid approach. Here are four best practices:

1. Build Collaborative Tools: Invest in planning platforms that surface data-driven recommendations but also allow merchant overrides and commentary. Human-in-the-loop systems improve decision quality.
2. Use Data for What It’s Best At: Let algorithms handle the complexity—like clustering, demand forecasting, or size curve optimization—while merchants focus on assortment curation and storytelling.
3. Invest in Training: Teach merchants how to interpret and challenge model outputs. This upskills the team and builds trust in the system.
4. Create Feedback Loops: Post-season reviews should assess both algorithmic accuracy and merchant calls. This continuous learning loop strengthens both components.

Looking Ahead

The future of assortment planning is not about choosing between data and instinct. It’s about enabling both to coexist in a structured, informed way. Retailers who can harmonize algorithmic intelligence with human experience will be best positioned to delight customers, optimize inventory, and outperform competitors. The era of Assortment Planning 2.0 is here—not just smarter, but wiser.