A portfolio can look clean in a spreadsheet and still fail in the market. That is why SKU rationalization cannot rely only on internal economics. The company also has to understand how customers buy, why they choose one item over another, and what role each SKU plays in a specific channel. Some variants are genuinely incremental. Others merely divide demand that would have gone to a neighboring item. Some customer-specific SKUs are strategically justified. Others are historical exceptions that have been carried forward long after the original reason disappeared.
This chapter moves the analysis from the inside of the business to the outside. It explains how to read customer needs and buying behavior, how to distinguish incrementality from cannibalization, how to evaluate customer- and channel-specific SKUs, and how to account for shelf, assortment, and presentation requirements before making rationalization decisions.
9.1 Understanding Customer Needs and Buying Behavior
The first task in customer and channel analysis is to understand what the customer is actually trying to accomplish. Too many rationalization efforts assume that if two SKUs are technically similar, they are functionally interchangeable. In practice, customers do not buy products as abstract product codes. They buy solutions for specific occasions, price points, use cases, and constraints. If the company does not understand those drivers, it will either protect too much complexity out of caution or remove items that matter more than the numbers suggest.
Customer need: the underlying requirement or job that a buyer or user is trying to satisfy when selecting a SKU. That need may be practical, such as the right dosage, fitment, size, or compatibility. It may be economic, such as affordability at a specific price point or value in a larger pack. It may also be experiential, such as flavor preference, convenience, portability, aesthetics, or a brand cue. Good rationalization work begins by identifying which differences customers truly value and which differences they barely notice.
Buying behavior helps separate those two things. A customer who repeatedly buys only one specific size or format, despite nearby alternatives, may be signaling a real requirement. A customer who buys across several neighboring variants may be showing flexibility. That is why behavior should be read over time, not from a single order. Repeat purchase patterns, switching behavior, basket composition, and seasonality all help reveal how much a SKU is actually differentiated in the market.
It is also important to distinguish between the buyer and the end user. In many channels, the item is chosen by a retailer, distributor, procurement team, or physician, but consumed or used by someone else. Their needs are related but not identical. A retailer may care about shelf productivity, price architecture, and visual assortment. The end consumer may care about flavor, size, or convenience. A hospital buyer may focus on formulary or contract terms, while clinicians care about format and ease of use. Rationalization must understand both sides if it is to predict what will happen after a change.
Customer and channel analysis therefore should combine data with informed field input. Historical sales alone rarely explain why a SKU matters. Sales teams, category managers, and customer service teams often know which variants are truly requested, which are merely accepted because they exist, and which could be transitioned if the company handled the conversation properly. The key is to capture that input in a disciplined way. Anecdotes are valuable only when they help interpret behavior, not when they become a blanket argument against change.
9.2 Measuring Incrementality vs. Cannibalization
One of the most important questions in SKU rationalization is whether a product is bringing new demand into the portfolio or merely taking demand from nearby items. This is where many portfolios become bloated. New variants are often approved because they are expected to create growth, but once launched, they mainly redistribute existing sales across a broader set of SKUs. The topline may look stable, yet operational complexity increases.
Incrementality: the share of a SKU’s demand that represents genuinely additional sales the company would not have captured through another item in the portfolio. Cannibalization: the share of demand that is taken from existing company SKUs rather than from competitors or entirely new demand. Rationalization depends heavily on this distinction. A cannibalizing SKU is often a candidate for consolidation, while a genuinely incremental SKU deserves stronger protection.
Incrementality is rarely visible from one simple metric. It has to be inferred from a combination of patterns. One useful signal is overlap in customer purchasing. If the same customers frequently buy several very similar SKUs interchangeably, the variants may be substituting for one another rather than expanding the market. Another signal is launch behavior. If a newly introduced variant quickly gains sales while neighboring SKUs decline in parallel and total family sales barely change, the item is probably cannibalizing more than it is adding. Promotional patterns can also be revealing. If demand shifts sharply between close variants depending on promotion or availability, the portfolio may be carrying more choice than the market truly requires.
Customer-level and channel-level analysis improves this assessment. A SKU may be incremental in one channel and cannibalizing in another. For example, a smaller pack may attract a convenience shopper who would not buy the standard size, while doing little more than divide shelf demand in a supermarket. A channel-exclusive color, flavor, or accessory may attract a retailer-specific shopper, or it may simply offer one more version of the same choice. Rationalization should therefore examine incrementality in the context where the SKU is sold, not only at total-company level.
Commercial judgment matters here as much as data. Incrementality can be overstated by teams that champion launches and understated by functions focused on simplification. The solution is to force the question using evidence. If the SKU disappeared, what portion of demand would likely leave the company, what portion would move to a substitute, and what portion might have been diverted internally even while the item remained active? When teams debate this question explicitly, the portfolio becomes easier to manage because new and existing items are judged by the same standard: not whether they sell, but whether they add.
9.3 Identifying Customer-Specific and Channel-Specific SKUs
Many portfolios contain items that exist for one customer, one banner, one distributor, or one route to market. Some are commercially justified. Others are artifacts of years of unstructured concession-making. A disciplined rationalization effort must separate the two.
Customer-specific SKU: an item created primarily for a particular account or small set of accounts, often involving a dedicated pack size, label, configuration, or specification. Channel-specific SKU: an item tailored to the requirements of a route to market, such as club, e-commerce, convenience, wholesale, specialty retail, medical distribution, or direct sales. These SKUs are not automatically bad. In many businesses, they are necessary to win with important customers or to fit the economics of a channel. The danger is that they accumulate without periodic review.
The analysis should begin by asking why the SKU was created and whether that reason still exists. Was it tied to a formal listing requirement, an exclusive agreement, a regulatory need, a minimum shelf presentation, or a true end-user preference? Or was it introduced to close one negotiation, satisfy one buyer, or support a short-term promotion that became permanent by inertia? This history often determines how much flexibility the company actually has.
It is also essential to test dependency honestly. If a SKU is sold almost entirely to one account, that does not automatically make it strategic. The account’s total value, bargaining power, growth potential, and willingness to accept substitution all matter. Some customer-specific items protect large, profitable relationships. Others generate low-margin volume that the company continues to serve mainly because no one has reopened the discussion. The same logic applies to channel-specific SKUs. A warehouse club pack may be perfectly valid if the economics are strong and the format clearly fits the channel. A lightly modified version for a minor subchannel may be needless fragmentation.
The right output is not just a list of special-case SKUs, but a classification of them: required, commercially justified, potentially migratable, or likely removable. That classification gives sales and customer teams a practical starting point for negotiation and transition rather than forcing every exception into a binary keep-or-cut debate.
9.4 Evaluating Shelf, Assortment, and Presentation Requirements
Even when a SKU looks economically weak, it may still influence how the brand or category appears to customers. Rationalization decisions therefore have to account for physical and digital presentation. A portfolio is not just a manufacturing and planning construct. It is also something customers see, compare, and navigate.
Shelf requirement: the need to maintain sufficient breadth, facings, or visual presence to compete effectively in a retail or merchandising environment. Assortment requirement: the logic by which a channel or customer expects a product range to cover key price points, occasions, or use cases. Presentation requirement: the way items must appear, individually and together, to create clarity, conversion, and brand credibility.
In physical retail, shelf architecture matters. Retailers often want a range that signals choice but remains easy to shop. Too little breadth can weaken presence and reduce conversion. Too much breadth can fragment facings and reduce productivity. The company should therefore evaluate not only whether a SKU sells, but whether it helps create a coherent shelf set. In some cases, two adjacent variants may be redundant and one can be removed with little consequence. In other cases, removing an item may leave a visible gap in size ladder, flavor ladder, color spectrum, or price architecture that makes the entire set harder to understand.
The same principle applies online. E-commerce may reduce physical shelf constraints, but it introduces its own presentation rules. Search filters, hero images, ratings, product titles, and assortment pages all influence conversion. Too many near-identical variants can confuse the shopper, dilute reviews, and weaken search performance. In digital channels, rationalization often improves customer experience by reducing clutter and making the offer easier to navigate.
Presentation analysis should also account for packaging and portfolio design. Sometimes the right answer is not to keep every weak SKU for visual completeness, but to redesign the range so that the remaining items form a clearer architecture. A cleaner pack hierarchy, better naming logic, or more deliberate size ladder can allow the company to reduce count while preserving customer confidence. That is the underlying principle of customer and channel analysis: simplify where the market will accept simplification, and preserve variety only where customers truly recognize and reward it.
