Retail Sales Incentive Plans That Reward Quality, Not Just Volume

What an incentive plan is for

An incentive plan does three things: it directs attention, it pays for outcomes the store cannot get from salary alone, and it tells advisors what the company thinks good work is. The third is the one most plans forget. A plan that pays only on revenue says that a sale made by misquoting the warranty is as good as one made by answering the objection. Advisors believe what the plan says over what the trainer says.

In Indian retail the variable part is typically 15 to 40 percent of total pay for floor advisors, higher in automobile and jewellery, lower in telecom and durables, with brand promoters in multi-brand stores often paid by the brand on a separate plan from the store's own staff. These are working ranges from what chains describe publicly and in hiring, not a survey figure. The design question is what that variable part rewards.

The six designs, compared

Each design below is used somewhere in Indian retail today. None is wrong on its own; each breaks a specific way when it is the only component.

The pitfalls

1. Paying for the outcome only. The month-end number is the result of thirty days of conversations, and a plan that pays only on the number cannot say which conversations to change. Advisors learn to hit the number the easiest way, which in assisted retail is the customer who was going to buy anyway.
2. Self-reported inputs. Number of demos, calls made, CRM entries. Every self-reported input becomes a filled-in field within a month. If quality is to be paid for, it has to be measured from something the advisor did not write.
3. Brand spiffs that override the store plan. A promoter paid Rs 500 per unit by one brand and Rs 200 by another sells the first, and the store's own plan is irrelevant. Multi-brand stores need the handover and the spiff schedule agreed with each brand, or the floor sells the spiff.
4. Cliffs. A plan that pays nothing below 90 percent of target and a lot above it produces sandbagging in bad months and pulled-forward deals in good ones. Smooth the curve: pay something from 70 percent, accelerate gently.
5. Paying for finance and warranty attach without a disclosure check. The attach rate goes up, and so do the complaints six months later. The record of what was disclosed at the counter, described at /learn/how-to-prevent-mis-selling-in-retail-with-disclosure-records/, is the input that makes an attach incentive safe.
6. Changing the plan quarterly. Advisors stop believing any plan will last, and behave as if there is none. Change the weights once a year and the targets monthly.
7. Ranking as the only recognition. A leaderboard rewards the same three people every month and tells the other twelve they are losing. Fair recognition, as at /learn/how-to-recognise-top-performing-store-advisors-fairly/, needs the input to be something everyone can move.

Adding quality inputs fairly

The objection to quality inputs is always the same: "who decides what is quality?". If the answer is the manager's opinion or a mystery shopper's one visit, the objection is right. The quality inputs that survive contact with an incentive plan are the ones measured from real conversations across the whole month, on steps the playbook already names, where each advisor can see the conversations behind their own score.

1. Choose two or three steps, not ten. Number and date taken from non-buying customers, the objection answered rather than acknowledged, disclosure said before the finance form. Steps that predict sales in the store's own data, not steps that sound good.
2. Measure from consented conversations, not from CRM fields. Coverage matters: an advisor scored on 8 conversations in a month has not been measured. Set a minimum, say 40 scored conversations, below which the quality component is paid at the store average.
3. Weight it to matter but not dominate: 20 to 30 percent of the variable pay on quality, the rest on sales and mix. Enough that skipping the steps costs money; not so much that a good seller with a rough style is punished.
4. Show each advisor their own score with the conversations behind it, before payout. A score that can be inspected is trusted; a score that arrives as a number is fought.
5. Allow an appeal on a scored conversation. The occasional mis-scored recording is fixed in five minutes, and the right to challenge is what makes the rest accepted.
6. Pay the quality component monthly, with the sales component. Delayed quality pay is forgotten pay.

A worked plan for an assisted retail store

Illustrative weights for a durables or mobile store with six advisors and measured conversations. The multiplier structure means the quality score scales the sales payout rather than sitting beside it, which stops an advisor treating quality as optional once the sales target is hit.

Brand promoters and dealer networks

A brand paying promoters in multi-brand outlets has the harder version of the problem: the promoter's outcomes are visible in sell-out data weeks later, and the promoter's conversations are invisible to the brand entirely unless captured with consent. Brand plans default to per-unit spiffs because units are all the brand can see, and the spiff wars between brands on the same floor are the result. Promoter plans that add a quality input, the pitch delivered as designed, the rival's claim answered, the launch demonstrated, need the conversation, and the method is at /learn/promoter-performance-tracking-for-brands/.

Dealer networks in automobile and telecom have a further layer: the dealer's plan for its own staff and the OEM's or operator's plan for the dealer. When the OEM pays the dealer on volume alone, the dealer's advisor plan follows. An OEM that adds a conversation-quality input to the dealer scorecard, even a small one, changes what the dealer asks of its floor.

Measuring quality honestly

Every quality input above depends on measuring real conversations, in the language the floor speaks, with consent shown to the customer, and with the advisor able to see the conversations behind the score. Borentis scores consented conversations against the store's own playbook, in Hindi, English and Hinglish, and produces the per-advisor number-taken, objection-answered and disclosure rates with the transcript lines behind each, which is the form an incentive input has to take to be argued about fairly and then accepted. Coverage is reported with every score, so a low-coverage month is paid at the store average rather than on a guess.

A plan that pays for quality without measuring it is a volume plan with extra paperwork. A plan that measures it and shows the evidence changes what advisors do in the conversation, which is the only place sales are actually made.