Toyota Kirloskar: A 23 Point NPS Gain in One Year

At a glance

• Toyota Kirloskar's NPS had plateaued, and the measurement process could not explain why.
• Feedback was collected manually and arrived too late and too thin to act on at dealer level.
• Real-time, dealer-level measurement lifted NPS by 23 points within a year.
• Feedback response rates rose 300 percent, giving dealer-level scores enough volume to be meaningful.
• The programme now runs across more than 520 dealerships.

The situation

Toyota Kirloskar Motor sells and services through a dealer network across India, and the Numr programme now runs across more than 520 dealerships. Ownership experience is largely delivered by those dealers, not by the manufacturer directly, which makes dealer-level performance the thing that actually determines whether a customer returns.

The company had a national NPS programme. The score had stopped moving.

The measurement itself was the constraint. Feedback was gathered through manual surveys, which meant low response volumes, slow collection and results that arrived aggregated at a level too high to act on. A national score tells the manufacturer how it is doing. It does not tell any individual dealer principal what to change on Monday.

A national NPS score is an average of hundreds of dealer experiences. Nobody in the network is responsible for the average, so nobody acts on it.

Why a plateau is a measurement problem

A flat score usually means one of two things, and they call for opposite responses.

Either performance is genuinely stable, or the measurement lacks the resolution to see movement that is already happening. Aggregation causes the second. Improving dealers and declining dealers cancel each other out, the national number sits still, and the organisation concludes nothing is working.

Toyota Kirloskar's variance was the point. Some outlets were delivering considerably better experiences than others, and the reporting structure made that invisible.

What we changed

Measurement moved to real time, at the dealer. Feedback was captured close to the service or sales interaction rather than gathered manually and reported later, and results were resolved to the individual outlet rather than rolled up nationally.

Response rates were addressed as a precondition, not a nice-to-have. Dealer-level scoring only works if each outlet generates enough responses to be statistically meaningful. Below that threshold the numbers move on noise and the network stops believing them.

Results went to the people who could act. Dealer principals received their own performance rather than a national figure, which converted the programme from a corporate report into an operational tool.

The result

Within one year, NPS improved by 23 points.

Feedback response rates rose by 300 percent, which is the number that made the rest possible. Higher response volume gave individual outlets scores stable enough to be held accountable against, and gave the manufacturer the resolution to see which dealers were moving.

Both figures were measured across the programme's more than 520 dealerships and aggregated over the 12 months of the first year.

At Toyota Kirloskar's request, baseline scores are not published.

A national score cannot be improved because nobody owns it. A dealer-level score can, because someone does.

Amitayu Basu, CEO and Co-founder, Numr

What transfers to other dealer and franchise networks

Measure at the level where the experience is delivered. In any franchised or dealer model, the customer's experience is created locally. A national roll-up describes it without ever locating it.

Treat response rate as infrastructure. Dealer-level accountability requires dealer-level statistical validity. Without the volume, local scores swing on noise and the network learns to dismiss them.

Give people their own number. The single largest behavioural change in this programme was not analytical. It was that a dealer principal could see their own performance rather than the company's.