FC Health Index1,428.60+0.42%Novo NordiskDKK 812.4+1.10%Intuitive SurgicalUSD 546.9−0.30%EU AI Act — Art. 6in forceM+7FDA 510(k) AI clearances (YTD)312+18 w/wNHS AI Diagnostic Fund£123mcommittedKarolinska trials open48+2Reimbursement CPT codes (AI)17+1 QFC Health Index1,428.60+0.42%Novo NordiskDKK 812.4+1.10%Intuitive SurgicalUSD 546.9−0.30%EU AI Act — Art. 6in forceM+7FDA 510(k) AI clearances (YTD)312+18 w/wNHS AI Diagnostic Fund£123mcommittedKarolinska trials open48+2Reimbursement CPT codes (AI)17+1 Q
Tuesday, 17 November 2026 · Oslo · London · New York

Enterprise · Analysis

Why American Businesses Are Rethinking How They Train Customer-Facing Employees

Product knowledge is easy to test. Explaining value to a frustrated customer is not. Employers are moving from courses and completions toward practice, coaching and measured readiness, and the research explains why.

AI-generated editorial illustration of a branch employee rehearsing a customer conversation with a coach. Not a photograph of a real person or company.
AI-generated editorial illustration of a branch employee rehearsing a customer conversation with a coach. Not a photograph of a real person or company.

Brand-requested analysis

F

By Future Chronicle

Published 16 November 2026

7 min read

Evidence: Analysis

A new hire at a regional bank in Ohio can pass every module on checking accounts, overdraft rules and the new rewards card in her first week. On day nine a customer walks in angry about a fee, mentions that a credit union down the street does not charge it, and asks why he should stay. The quiz never asked that question. The customer just did.

That moment, repeated across retail floors, call centers, insurance agencies, car dealerships, clinics and sales territories, is why many American employers are reconsidering how they prepare people who talk to customers for a living. The problem is rarely a shortage of content. It is the distance between knowing something and being able to use it in a live conversation.

Why classroom training and online courses leave gaps

Formal training does work in general. A 2003 meta-analysis by Arthur, Bennett, Edens and Bell found meaningful effects of organizational training on learning, behavior and results. But researchers have long distinguished between learning in a session and applying it at work, a problem called transfer of training. Burke and Hutchins' 2007 review found transfer improves with realistic practice, feedback, supervisor support and a chance to use new skills soon after training.

Many programs offer few of those conditions. A two-hour webinar followed by weeks of unobserved work is efficient for delivering information and weak at building habits. Common measures do not help much either. Alliger and Janak's 1989 review found that how much trainees liked a course was a poor predictor of later job behavior, yet satisfaction surveys and completion rates remain the default scorecard in many learning systems.

Timing matters as well. Cepeda and colleagues' 2006 analysis of distributed practice found that spacing learning over time generally improves retention compared with cramming it into one block. An annual kickoff is the definition of a block.

Knowing the product versus explaining its value

An employee can recite that a phone plan includes unlimited data and international texting. A customer wants to know whether it will cost less than what she pays now, and whether switching is worth an afternoon. Those are different skills. The first is recall. The second requires connecting a feature to a specific person's situation, in plain language, while that person is skeptical.

This is where traditional assessments mislead managers. A multiple choice test measures recall well. It says almost nothing about whether someone can translate a feature into a reason to buy, stay or trust the company.

Why listening, questioning and objection handling matter

The strongest customer conversations usually depend less on what an employee says than on what they ask. Good questions surface the real concern: the angry banking customer may care less about a $12 fee than about feeling ignored. Listening long enough to hear that changes the conversation from a defense of policy to a solution.

Objection handling is the same skill under pressure. Price, timing, a competitor's offer and a past bad experience are predictable objections in almost every industry. Because they are predictable, they can be rehearsed. Most organizations simply do not rehearse them often enough for responses to become natural.

Roleplay and realistic customer scenarios

Roleplay is an old tool, and the evidence for behavior-focused practice is reasonably encouraging. Kauffeld and Lehmann-Willenbrock's 2010 field study found sales training built around behavioral practice was associated with improved performance. Practice is not a cure-all; Macnamara, Hambrick and Oswald's 2014 meta-analysis found deliberate practice explains only part of performance differences across fields.

Scenarios work best when they are specific. Instead of "practice handling a difficult customer," a home insurance agency might script a homeowner whose premium rose 20 percent after a hail season and who has a quote from a competitor. A pharmacy might rehearse a patient confused by a prior authorization denial. The closer the scenario is to a real Tuesday afternoon, the more useful the practice.

The traditional barriers are time and awkwardness. Manager-led roleplay needs a manager, and many employees find performing in front of peers uncomfortable, so it happens rarely.

How AI is changing practical training

AI simulation addresses the scaling problem. An employee can rehearse a conversation with a simulated customer at any hour, repeat it until it feels comfortable and receive structured feedback without an audience. That makes spaced, repeated practice realistic for large or distributed teams.

The evidence base is still young. Academic work such as Habel, Ahearne, Tirunillai and Vandaveer Novak's working paper is beginning to study AI-based sales training, but it has not completed peer review. AI feedback can also be wrong or miss context a skilled manager would catch. Employers should treat it as a practice partner, not a judge.

One example of AI-supported sales practice

Several vendors now offer AI roleplay and coaching software for sales and service teams. One example is Practis, which applies AI roleplay, sales coaching and readiness assessment to sales development. According to its own published materials, practice conversations are scored across multiple skill dimensions within a staged process intended to move reps from learning content toward demonstrated readiness. Future Chronicle has not tested the product, and these descriptions come from the company. Like others in the category, it has not published independently audited evidence linking its scores to sales results.

Coaching, feedback and measuring readiness

Practice without feedback can simply reinforce bad habits. The research on transfer consistently points to supervisor support as a factor, which means AI tools complement managers rather than replace them. A useful pattern is for employees to practice independently, then for a manager to review a small sample of recorded attempts and coach on one or two specific behaviors.

Readiness also needs a definition. For a call center agent it might mean resolving a billing dispute without escalation and without promising something policy does not allow. For a medical device rep it might mean explaining a product change to a surgeon accurately and briefly. Writing these standards down turns readiness from a feeling into something observable.

If AI scores influence pay, promotion or hiring, governance matters. NIST's AI Risk Management Framework emphasizes validity, fairness and transparency, and New York City's Local Law 144 requires bias audits for certain automated employment decision tools.

What leaders should consider when updating programs

First, audit what is measured today. If the main metrics are completions and satisfaction, the program measures activity, not capability. Second, identify the five or six conversations that matter most for each role and build realistic scenarios around them. Third, space practice across weeks instead of concentrating it at onboarding. Fourth, give managers time and simple guidance to coach, since tools rarely help when nobody reviews the results. Fifth, pilot any AI tool with one team, compare practice results with real outcomes such as resolution rates or conversion, and keep human review of any consequential scoring.

Limitations

Much of the training research predates current AI tools, and some studies focus on sales rather than service roles. Vendor descriptions reflect intended capabilities, not measured results. This analysis was brand-requested and prepared with AI assistance under editorial review; it did not include interviews or product testing.

The bottom line

Customer-facing work is a performance skill, and performance skills are built through rehearsal and feedback. Businesses that keep courses for knowledge, then add regular practice, real coaching and clear readiness standards, are likely to see more of their training show up where it counts: in the conversation with the customer.

Frequently asked questions

Why does traditional training often fail customer-facing employees?

It is good at delivering information but offers little realistic practice or feedback, which research identifies as key to applying skills on the job.

What is the difference between product knowledge and selling skill?

Product knowledge is recall of features. Selling or service skill is connecting those features to a specific customer's needs, in plain language, while handling doubts.

Does roleplay training work?

Behavior-focused practice has research support, particularly when scenarios are realistic and paired with feedback. Results vary, so outcomes should be measured.

Can AI replace managers in coaching?

No. AI makes frequent practice easier, but its feedback can miss context. It works best alongside manager review and coaching.

How should companies measure employee readiness?

Define role-specific skills in observable terms, assess them through recorded or observed practice, and check whether those results line up with real customer outcomes.

Published 16 November 2026