Training Needs Analysis for Retail: Is It a Skill Gap?

By awareretailgroup.com Editorial· Edited by Tom Walsh· Published 2026-10-15

A retail performance gap is a store number that misses its target, such as price overrides running above plan at one location. A training needs analysis for a store runs in three steps. Pull the metric that missed, match the pattern against the decision table below, and approve a course only when associates truly lack a skill or knowledge. An override key that needs no manager approval calls for a register change.

Why a training needs analysis should check the environment first

Rummler and Brache put it plainly: pit a good performer against a bad system, and “the system will win almost every time.” They report that usually 15 to 20 percent of performance opportunities sit in the Skills and Knowledge area, and about 80 percent in the environment the performer works in (Rummler & Brache); by their own account the figures come from consulting experience rather than a sampled study and vary somewhat across jobs, industries and countries. When Christensen and Wallace looked for the research behind the claim for eLearn Magazine, they wrote: “We have not been able to locate any specific research to demonstrate that these claims are evidence based, only assumptions” (eLearn Magazine). Don’t quote them to a CFO as proof.

Roger Chevalier’s update to Thomas Gilbert’s Behavior Engineering Model says environmental factors “are the starting point for analysis because they pose the greatest barriers to exemplary performance” (Chevalier), and that they are “usually more cost-effective to fix than individual factors” (Chevalier). Thomas Gilbert’s own sequence, described by Boise State, checks data, then instruments, incentives, knowledge, capacity and motives, in that order. Knowledge sits fourth on that list.

How do I tell whether a shrink or LP performance gap is a skill problem, or a tools, incentive or process problem?

An exception report that clusters around one register, one shift or one SKU points toward a process or tool problem or, as Famous Footwear’s sorting allowed, possibly fraud. Chevalier frames the underlying logic: changing individual factors “will most likely not improve” performance if environmental factors remain unresolved.

What Famous Footwear’s coupon exceptions revealed

Famous Footwear used exception reporting to sort each problem before fixing it, aiming to “understand whether we have a training problem, a policy procedure issue or possibly fraud, and follow up from there” (LP Magazine). It found two coupon problems this way. Treat this as one retailer’s own account, first published in 2005 and updated in 2016 (LP Magazine), not an industry-wide study.

The first was a loyalty coupon. In almost one third of cases, the associate redeemed the coupon rather than explain that it couldn’t be used on that transaction (LP Magazine). The LP department worked with other departments on three fixes: associates were trained in how to respond to the request, the POS system was altered to help prevent the discount, and the brochure and coupon were reworded to explain the terms (LP Magazine). Infractions dropped 40 percent in a matter of months (LP Magazine). Because training came with the POS and wording changes, the drop cannot be credited to training alone.

The second was “Coupon Zero,” a register code associates used to honor expired coupons under a policy that was communicated poorly in the procedures (LP Magazine). Famous Footwear store associates generally were “doing what they were measured on”, trying to accommodate customers and drive sales, and in this case, because the POS system and the manual supported the override, the stores used it (LP Magazine). The home office groups set the policy, revised the procedures and re-educated the associates, and the new procedure reduced that trend by 62 percent (LP Magazine).

When a job aid beats a course

The U.S. Coast Guard training standard makes two separate speed claims. It says the Coast Guard “can develop job aids three to four times faster than developing training materials designed for memorization” (Coast Guard). Training its people to use a job aid, by the same standard’s estimate, is three to four times faster than training the same task for storage in memory (Coast Guard).

For some tasks, the guidance goes further: “plan to develop only a job aid for the task; training support is NOT needed” (Coast Guard). A rare refund exception or a seasonal price-override rule is a candidate for a laminated card or a quick-reference screen at the register (more on associate training). Our gift card scam register checklist is one example of that format. A task performed daily, under real consequence, with no time to look anything up, like de-escalating an angry customer, still needs practiced training.

Why turnover favors process discipline over more training

A study of retail stores found that increasing turnover does not hurt performance at stores with high process conformance, but has a pronounced negative effect at stores with low process conformance (Ton & Huckman). Where performance depends on repeating known tasks, the researchers suggest managers can reduce turnover’s effect by imposing process discipline through standard operating procedures (Ton & Huckman). An SOP only helps if it is followed, and why store procedures get ignored covers that side.

That result is limited to repeated, known tasks and does not reach judgment calls. A manager weighing what to do when a repeat offender walks in, or an associate holding a tense exchange together, is doing something process conformance can’t standardize away. In a high-turnover store, tighten the SOP and put a job aid at the register for routine-task gaps, and keep judgment tasks like de-escalation in the training budget regardless of how fast the roster turns over.

The process-failure share of shrink

NRF’s 2023 National Retail Security Survey, covering fiscal year 2022, found that retailers attributed 36% of shrink to external theft, 29% to employee theft, 27% to process, control failures and errors, 6% to unknown sources and 1% to other sources, on average (LPRC). Those are the newest figures of their kind and they are dated: NRF did not publish the survey in 2024, saying “a broad study about retail shrink is no longer sufficient for capturing the key challenges and needs of the industry” (Retail Dive). The same survey put industry-wide shrink at 1.6% in fiscal 2022, up from 1.4% the year before, roughly $112 billion in losses against $94 billion the prior year (LPRC).

Adrian Beck and a co-author believe that most internal and external theft, vendor fraud, and nearly all process failures trace back to operational failures within the business (LP Magazine). LPRC’s Dr. Cory Lowe cautions against conflating the shrink problem with the harms of retail crime, since shrink includes both criminal and non-criminal sources (LPRC). A store manager handed a shrink number without a cause breakdown should ask for the store’s own split between internal theft, external theft and process error before approving any training spend against it.

How to know the fix worked

Famous Footwear measured success in the same infraction data that exposed the problem (LP Magazine). Before anyone orders a course, decide which number will prove the fix worked: whatever store data flagged the gap (the exception rate, shrink by store, an audit score) is the one to re-run.

For the metric work itself, see how to measure whether associate training actually worked.

What to do this week

• Pull last quarter’s exception report, audit score, or shrink-by-store number tied to the problem that triggered the training request.
• Run the symptom against the decision table above before approving a course.
• If the cause is tools, incentives, or process, fix that first and skip the course.
• If the cause is a rare task that can be looked up at the point of use, build a job aid instead of a module; add a short walkthrough if the task is new, complex, or performed daily under pressure.
• Re-run the same metric in 30 to 60 days and report the resulting change.