A SKU rationalization effort becomes real only when analysis is translated into action. The work so far has shown where demand is weak, where margin is poor, where complexity is costly, and where substitutes exist. Leaders now need a recommendation set that is commercially sound, operationally feasible, and specific enough to implement.
This chapter explains how to build that recommendation set. It shows how to identify SKUs to eliminate, how to find items that should be consolidated or redesigned rather than simply removed, how to simplify common forms of portfolio complexity, and how to sequence the resulting actions so the company captures value while controlling risk.
13.1 Identifying SKUs to Eliminate
The most visible output of a rationalization program is the list of SKUs that should leave the portfolio. Weak lists usually come from one of two errors. Either the business uses crude cutoffs and targets items that are small but still necessary, or it becomes so cautious that only obviously dead items are removed and most of the value remains untouched.
A strong elimination recommendation begins with convergence of evidence. The best candidates are not merely low sellers. They are SKUs that perform poorly across several dimensions at once. They contribute limited demand, weak profit, and little strategic value. They create disproportionate complexity, inventory burden, or servicing cost. Customers have a credible substitute path. No regulatory, contractual, or technical barrier meaningfully protects them.
It helps to separate elimination candidates into three groups. The first is inactive clutter: items that are effectively dead already, with no meaningful demand, old inventory, obsolete listings, or system residue from past launches and legacy codes. The second is active but nonessential clutter: SKUs that still sell, but sell poorly, transfer easily, and create little unique value. These often generate the first wave of benefit. The third is managed end-of-life items: products that should exit, but only after inventory is run down, customers are notified, or a contract window is reached. Lumping these together creates confusion. Separating them improves both decision making and implementation.
Elimination recommendations should also be built at the right level. Some items look weak only because the real problem is duplication inside a family. If the team evaluates them one by one, it may produce a list that is technically correct but strategically incoherent. A better method asks first whether the customer’s needs should remain in the portfolio at all. If the answer is no, the item is an exit candidate. If the answer is yes, but the current offer is fragmented, the right action may be replacement or consolidation rather than pure elimination.
For each SKU proposed for exit, the team should specify the path, not just the label. A recommendation should state whether the item is a clean discontinuation, an account-specific removal, a regional exit, or a phased end-of-life with a substitute. It should also make clear where demand is expected to go and what stock disposition is required. Many programs fall short here. They produce a list of “cuts” but do not show how each cut will actually happen.
The burden of proof should change as the evidence strengthens. Once a SKU is shown to be low value, high burden, and readily substitutable, the default should exit unless a clear exception is documented. Without that shift, weak SKUs survive because no one wants to take responsibility for removing them. With it, the portfolio begins to reflect deliberate choice rather than accumulated tolerance.
13.2 Identifying SKUs to Consolidate or Redesign
Pure elimination is only one form of rationalization. In many portfolios, the larger value opportunity lies in simplification through consolidation or redesign. This is especially true when the customer need is valid but the portfolio contains too many ways of serving it. In those cases, removing an item without improving the surviving range would create unnecessary customer pain. The smarter move is to reduce fragmentation while preserving the underlying need.
Consolidation: reducing the number of SKUs that serve essentially the same purpose by concentrating demand into a smaller set of survivors. Redesign: changing the attributes, packaging, configuration, or architecture of the surviving offer so that more demand can flow through fewer SKUs without damaging customer acceptance.
Consolidation is usually the right answer when several items are splitting demand that could be pooled. Common patterns include overlapping pack sizes, minor flavor or scent variations, duplicate channel variants, excessive colorways, overlapping configurations, or customer-specific versions that differ only superficially. The goal is not to force standardization everywhere. It is to identify where the company is paying for distinctions that customers do not meaningfully value.
The first step is choosing the survivors. This should not be based only on history or internal preference. The surviving SKU should be the one that best combines customer acceptability, healthy economics, manageable complexity, and scalability. In some cases, that means keeping the item with the highest current volume. In others, it means migrating volume into a nearby item that has better margin, better manufacturability, or cleaner fit with the long-term portfolio architecture.
Redesign becomes important when none of the current items is the obvious winner, or when a slightly changed item could absorb demand from several weaker variants. A company may standardize ingredients, materials, or components so that one product can replace several near-duplicates. It may simplify a configuration menu so customers choose from fewer, clearer options. It may redesign a surviving SKU to better match the needs of customers currently buying an item marked for removal. In these cases, the action improves the portfolio while reducing complexity.
The recommended action set should therefore include not only elimination candidates, but also family-level simplification moves. These often produce better long-term economics than removing obvious tail items.
13.3 Simplifying Pack, Size, Flavor, Color, and Configuration Complexity
Pack sizes multiply to hit different price points. Flavors, scents, colors, and finishes are added to create choice. Configurations and options expand to meet special requests. A strong action plan restores discipline by showing where the range has become more granular than customer value justifies.
Pack and size complexity: this is usually best tackled by looking at the full ladder rather than isolated items. The key question is whether each step serves a distinct need or whether adjacent sizes are cannibalizing one another. Many portfolios contain too many “almost meaningful” sizes: enough difference to create another SKU, not enough to create clearly separate demand. Simplification may mean removing a redundant middle size, standardizing case counts, harmonizing pack formats across channels, or collapsing local variants into a core set of sizes.
Flavor, scent, and variety complexity: consumer-facing portfolios often accumulate these variants because they appear commercially harmless. Yet small sensory differences can create a large long tail of low-volume items. The right approach is not to strip choice mechanically. It is to decide which variants are true destinations, which are occasional traffic builders, and which are near-duplicates with weak incrementality. A healthier range usually includes clear hero variants, a small number of differentiated alternatives, and a disciplined policy for limited editions and channel exclusives.
Color and finish complexity: in apparel, footwear, home goods, and accessories, color proliferation is a common source of hidden burden. Individual colors may look attractive in line reviews, but the portfolio effect is often poor: split demand, high markdown risk, and scattered inventory. The right action is often to narrow the palette to the colors that truly drive volume or brand identity, while eliminating shades that are too close, too seasonal, or too account-specific to justify the cost.
Configuration complexity: in industrial products, electronics, automotive, and medical categories, complexity often sits in option combinations rather than visible consumer variety. Too many feature bundles, fitments, accessories, or specification combinations can create an illusion of responsiveness while overwhelming the operating model. Simplification may involve creating standard configurations, reducing optionality, separating true custom orders from standard catalog items, or moving low-frequency variants into engineered-to-order processes rather than stocking them as finished SKUs.
Across all of these categories, the important move is from item-by-item thinking to architecture thinking. The question is not simply which specific SKU is weak. The question is whether the overall structure of the offer is too fine-grained. A good recommendation therefore explains not only which items leave, but what the simplified architecture will look like afterward.
13.4 Sequencing Actions by Ease, Impact, and Risk
Once the action set has been built, the business still faces one more challenge: deciding the order in which to act. A rationalization program can have the right recommendations and still underperform if the sequence is poor. Too many hard moves at once can overwhelm the organization. Too much focus on easy moves can create activity without meaningful value.
Sequencing: ordering recommended SKU actions so the business captures value efficiently while controlling operational and commercial risk. The best sequence usually balances three factors. Ease: how straightforward the action is to execute from a systems, inventory, and customer standpoint. Impact: the value expected from the action in revenue retention, margin improvement, inventory reduction, or complexity removal. Risk: the likelihood that the action will create customer loss, service disruption, implementation failure, or internal resistance.
In practice, this means grouping actions into waves. Early waves usually include inactive clutter, clear exit cases with high transfer confidence, and small consolidations that require limited customer management. These actions build momentum and free capacity for harder work. The next wave often includes family consolidations, pack or size simplification, and channel-specific cleanup where communication and inventory runout need more coordination. Later waves typically include the most sensitive changes: major account transitions, redesign-driven migrations, and actions that depend on tooling changes, packaging changes, qualification work, or contract timing.
Dependencies are especially important. A SKU may be an excellent candidate and still need to wait because the replacement is not ready, inventory levels are too high, promotional calendars are locked, or a contract renewal is upcoming. Conversely, some items should move early precisely because those dependencies take time and the notice period must begin now. The point is to treat sequencing as a business decision, not as an administrative afterthought.
A useful sequencing test asks two questions for every action. What value does this unlock, and what must be true before we can execute it well? That framing prevents both impatience and drift. It recognizes that good rationalization is neither a mass purge nor a slow academic exercise. It is a staged program that removes low-value complexity while strengthening the portfolio the company intends to keep.
The finished recommendation should therefore look less like a static list and more like a portfolio of actions: some immediate, some phased, some contingent, some structural. The organization can see not only what should change, but how the changes fit together and in what order they should happen.
