The one metric you decide

An independent used car dealer running thirty listings knows his lead volume. He far more rarely knows how long he takes to answer the first message, and almost never knows how many of those contacts got no answer at all.

It is the most thoroughly documented measurement in the trade. In 2011, Harvard Business Review published an audit of 2,241 US firms by James Oldroyd, Kristina McElheran and David Elkington: an average response time of 42 hours, 23 % of firms never responding, and roughly 7 times better odds of holding a qualifying conversation when the reply came within the hour. Four years earlier, a study run by James Oldroyd for MIT Sloan and InsideSales (2007, more than 15,000 leads and 100,000 calls) worked at the scale of the minute: contacting a lead within 5 minutes rather than 30 multiplies the odds of reaching them by roughly 100, and the odds of qualifying them by roughly 21.

The two studies are constantly conflated. They do not say the same thing, and they say the same thing: time is the first filter, ahead of price, ahead of the vehicle, ahead of the salesperson’s skill.

That is why first response time belongs on the first line of the dashboard. It is the only metric on the list that depends neither on stock quality, nor on the market, nor on negotiating ability. It is decided. Osmosia answers in under 15 minutes, 7 days a week from 7am to 11pm, which is not a sales performance but a setting.

The five metrics that make up the chain

Five measurements are enough to see where the sales go.

First response time, counted from the arrival of the message to the first real reply, not the platform’s automatic acknowledgement. The share of leads left unanswered, a figure almost nobody tracks and the single most revealing one. The qualification rate, the share of leads whose buying intent is established. The appointment rate, the share of qualified leads that turn into a dated slot. And the conversion rate, the share of attended appointments that end in a sale.

A sixth metric deserves separate treatment: the no-show rate. A booked slot is not a customer standing in the yard, and a conversion rate computed on appointments booked rather than appointments attended flatters the dashboard.

A median, never an average

An average response time is a boardroom number. Across a week of thirty leads, two messages received on a Sunday evening and handled on Monday morning are enough to push the average from twelve minutes to four hours. The dealer concludes he is failing everywhere, when he is excellent twenty-eight times out of thirty and absent twice.

The median says the opposite, and it says the truth: it measures ordinary behaviour. Track the median as your routine indicator, then look separately at the tail, the leads handled after more than two hours. That is where the lost sales hide. Never in the average.

”Qualified” does not mean “mentioned financing”

Plenty of dashboards borrow a grid inherited from B2B: a lead is qualified once budget, timeline and need are known. Applied to used cars, that grid produces false positives by the dozen.

A buyer who writes “interested, I have 7,000 euros down” has talked budget. He has committed to nothing. A buyer asking for a photo of the engine bay or the date of the last roadworthiness test has expressed a need. He has committed to nothing either. Conversely, “could someone call me” contains no budget, no timeline and no need, and signals a far stronger intent: the prospect wants out of the messaging thread.

Osmosia’s scoring separates three families of signals: the action taken (a confirmed appointment with a date and time, a request for a phone call, a document sent such as a registration certificate), the maturity of the prospect (precise technical questions on engine size, fuel consumption, service history, references to market valuation) and the trajectory of the conversation (moving forward, stalling, retreating). An intent score aggregates the three, and the qualification threshold is adjustable. No lead reaches the top of the scale without a concrete action.

The operational consequence is direct: measure your qualification rate on actions, not on declarations of interest. Otherwise you will spend the week calling polite browsers.

The KPI missing from almost every dashboard: days in stock

A lead has no value in itself. It has value because it moves a vehicle off the lot.

The quarterly barometer published by Mobilians and Avere-France gives the order of magnitude on the electric segment: in the fourth quarter of 2025, the average time to resell a used electric vehicle stood at 161 days across all ages, against a professional stock of 61,613 units. No equivalent public figure exists for used combustion vehicles, but every dealer knows his own.

Translate it into euros. An 18,000 euro vehicle sitting for five months is frozen capital, a floor plan loan running, depreciation advancing. Every day cut from turnover time is worth more, mechanically, than a point of margin won in negotiation. It is the only calculation that ties response time to cash flow, and the one to put in front of a banker.

Read the chain backwards

These metrics form a funnel, and a funnel is diagnosed from the bottom up.

A strong conversion rate paired with a poor response time does not mean things are fine. It means you sell remarkably well to the people you manage to reach, and that the leakage happens before the first exchange. Polishing the sales pitch in that configuration will produce nothing. The leak is upstream.

The opposite reflex costs just as much. Excellent response time, high qualification rate, and a collapsing attendance rate: the problem is neither speed nor sorting, it is confirmation. There is no point accelerating a machine that leaks further down.

So read the funnel whole, and always fix the highest stage that is failing. Fixing a lower stage while an upper one leaks is painting a room whose roof lets the rain in.

What Osmosia measures, and what it does not

Osmosia answers the first message, applies an intent score to every conversation, triggers follow-ups according to the delays you set, and books appointments when the prospect is ready. The interface reports three things: actual response time, the score of each lead with its history, and the appointments booked.

Osmosia does not know whether the sale happened. The attended appointment, the negotiation, the order form belong to you, and the conversion rate is computed in your management system, not in ours. A vendor claiming to measure the whole chain through to signature is in fact measuring whatever you choose to declare.

Data does not replace a salesperson’s instinct. It relieves him of guessing what time he replied, and who he never replied to at all.