OBI: 'too expensive' is not an instruction to issue a discount

A customer says 'too expensive', and somewhere inside the associate a discount window opens. Yet the person may mean something quite different: they do not understand the scope, expected a smaller project or were startled by the total. Training objection handling without investigating the reason is rather like choosing a tool before looking at the job.

OBI offers home-project advice and planning. That is the factual starting point for my independent AI sales-training proposal, not a description of an OBI deployment. I would train a conversation about the project budget rather than stage a contest between an associate and somebody else's wallet.

Open up the objection

In the first scenario, AI plays a customer with a limited room-refresh budget, revealing the amount after an appropriate question. The learner must establish scope, essentials and what can wait. Offering an immediate discount without understanding the constraint should not earn a good assessment simply because it sounds commercially energetic.

The second scenario is about an unclear estimate rather than an unaffordable total. The associate explains materials, preparation and services in ordinary language. A third customer compares incomplete quotes. The task is to align scope, not declare that another retailer is morally deficient and owns an unreliable tape measure.

After the dialogue, AI identifies the point where the learner assumed instead of asking. It then replays that section with a different customer response. This is more actionable than an empathy score out of ten. The associate needs a behaviour to practise, not a mysterious psychological verdict generated by software.

Saving money cannot remove essential safeguards

The exercise library contains approved cost categories, service terms and mandatory technical constraints. The assistant must not remove necessary protection or required work to achieve an attractive price. Complex technical questions go to specialists. Fictional cases are preferable; real estimates must be anonymised and checked before becoming training material.

The first trial could cover a few familiar project types with brief practice twice a week. Managers review a sample of exercises and discuss disputed assessments. Simulator scores should not automatically determine bonuses. Otherwise colleagues will soon become exceptionally good at selling the correct answer to the AI instead of helping customers.

Then test the conversation in practice: conversations identifying the actual budget objection, estimates revised because requirements were initially missed, and conversion of agreed quotes. Compare similar projects rather than a room repaint with a major renovation. Reduced discounting matters only if the customer still understands the offer and receives an appropriate solution.

The purpose is not winning an argument. It is building a realistic project together. Sometimes that means a smaller purchase now and another stage later. The customer understands what they are paying for and does not leave feeling defeated. For a long-term retail relationship, that is more promising than one heroically closed transaction.