Beyond Lead Volume: Why Call Outcomes Matter More Than Call Counts
Call counts were automotive's most trusted vanity metric. Here's what changed.
Lead volume tells you how many buyers called. Lead outcomes tell you whether any of them converted and why. Most automotive marketplaces track the first. Almost none systematically track the second. That gap is where revenue disappears.
Kaisa’s position is straightforward: a call count tells a dealer how many leads their marketplace subscription delivered. Outcome metrics tell them what those leads were actually worth, how many reached a qualified buyer, progressed to a test drive, or ended in a purchase commitment. For a marketplace, that is the difference between reporting demand and proving value.
What lead volume actually measures (and what it hides)
A lead volume metric answers one question: how many?
How many calls came in. How many leads were generated. How many times a buyer actually called. These numbers tell you whether your listings are attracting attention and whether your marketing channels are generating inbound enquiries.
What they cannot tell you is what happened on the call.
A volume metric just treats any phone call or form lead equally,” says Henrik Lenerius, Chief Product Officer at Kaisa. “It doesn’t matter if it was an actual fraudster, a spammer, or someone who wanted to talk about spare parts — those are treated exactly the same as someone who actually agreed to buy a car.
That bluntness runs deeper than most organisations realise. Without AI, the standard proxy for call quality is duration — typically a 30-second cutoff, with calls above it counted as qualified leads and calls below it written off as noise.
The problem is that the cutoff fails in both directions.
I know from some of the dealers we work with closely that if I call them just to test what that experience looks like, it can take more than a minute just to get through the IVR piece of it,” says Lenerius. “So counting 30 seconds as a qualified lead is really a false positive.
James Morris, CX and Data Specialist at Kaisa, sees the same pattern from the data side.
The dealer will have an IVR and a very long wait time — so it looks like this dealer is engaging with customers really well,” he says. “But that’s not true.
Volume metrics are also vulnerable to a more deliberate distortion. Dealers routinely leave sold or unavailable vehicles listed on marketplaces — continuing to generate call volume after the car has gone. Those calls are logged identically to calls about available stock. The buyer’s experience is completely different. The conversion potential is completely different. Volume reporting cannot tell them apart.
That invisibility has a cost. In automotive, where up to 42% of inbound calls never reach a human, and where 75% of used-car buyers purchase from the first dealer they successfully speak to, a missed call is not a minor operational friction. It is a lost sale — one that volume reporting does not log as a loss at all.
Three outcome metrics worth tracking instead
Shifting from volume to outcomes does not require replacing your existing reporting. It requires adding a layer beneath the call count that tells you what actually happened.
Three metrics are worth prioritising first.
Appointment rate — whether a test drive or physical visit was agreed — is the metric Henrik Lenerius recommends starting with. “All dealers I’ve met, almost without exception, feel like they’re very good at closing the deal if they can just get the lead inside their building,” he says. “So if you can prove you have a high level of leads that get to a test drive — that’s intuitively a very strong outcome.” It is also one of the most accurately detectable signals in a call, making it a reliable foundation for outcome reporting.
Human connection rate measures whether a call reached a human, reached a voicemail, hit an IVR menu, or was missed entirely. This single metric, mapped consistently across dealers, immediately reveals where calls are being lost — not because of poor lead quality, but because of operational gaps. Staffing, after-hours coverage, call routing: human connection rate exposes all of it.
Qualification rate is the share of inbound calls in which a human answered, the enquiry concerned an available vehicle, and the conversation reached a concrete buying signal — a buyer who asked about price or availability, requested a test drive, or agreed to an in-person visit. Unlike call duration, which can be inflated by hold times and IVR menus, qualification rate measures whether a real buying conversation actually happened. It is the metric that answers the question marketplace operators are usually arguing about without evidence: were these actually good leads?
“What happened in the call and are they qualified or not — there is nothing at that level,” says Fatemeh Iranmanesh, Analytics and Insights Coordinator at Kaisa. At the basic level of most call reporting — missed calls, answered calls, duration — buyer intent is entirely invisible.
Why volume-based reporting survives inside organisations
If outcome metrics are more useful, why is volume reporting still the default?
The honest answer is a technical one. “It’s just been a technological limitation,” says Henrik Lenerius. “Only in the last couple of years has the ability to actually analyse what goes on in a call with AI been sufficiently accurate, at a sufficiently low price point to make it realistic.”
Before AI, the alternative was manual listening — sales reps sitting through recordings, hours of work that couldn’t scale. Or organisations ran deterministic call-topic classifiers: complex if/then pipelines that were, in practice, blunt, inaccurate, and expensive. Large language models have made sophisticated call analysis affordable and accurate in a way that simply was not possible five years ago.
There is also an organisational dimension. Outcome metrics are more exposing: they make it harder to hide operational problems behind a good-looking call count. Volume metrics are relatively safe to report. High volume looks like success. Low volume prompts a conversation about marketing spend. Outcome data prompts a harder conversation about what is being done with it.
What changes when outcomes replace volume in reporting
The shift has practical consequences at two levels: inside the marketplace’s own reporting, and in the conversations it has with its dealer customers.
“The age-old debate for marketplaces has been to say: look how many leads we create for you,” says Lenerius. “And the dealer says, most of them are poor quality anyway. Now you have a much stronger evidence base to say, actually, this was valuable to you. It’s quality over quantity — proving your return on investment to your customers.”
Beyond the internal conversation, diagnosing exactly where the buyer journey breaks. “The digital handoff of a lead from the marketplace — the online part of the buyer journey to the offline journey that continues with the dealer — that’s a critical breaking point where a lot of leads drop out,” says Lenerius. “Now you can start to look at why. Are they getting stuck in IVRs they can’t navigate? Is there a voicemail that doesn’t prompt people to leave a message? What’s actually the problem here?”
Kaisa’s data shows that 78% of buyers who reach a voicemail don’t leave a message. Voicemail is not a safety net — it is where high-intent buyers disappear.
At the seller level, the change is equally concrete. A marketplace that can show a dealer their connection rate, their appointment rate, and how their calls compare to peers on the platform has a completely different renewal conversation than one presenting traffic and click data alone.
How AI call analytics makes outcome measurement practical
Outcome measurement at scale was not practical without AI. Listening to calls manually works for coaching a handful of sales reps. It does not work across thousands of dealers and millions of calls.
AI call analytics changes that. Every call is automatically analysed and structured: call disposition (how the call started and how it ended), appointment detection (whether a test drive or visit was agreed), and a factual call summary (what was discussed, what the buyer said, what the outcome was). No manual tagging. No sampling. Every call, every time.
“You’ve gone from the ability to just count leads to proving value through the quality of those leads — from outcomes rather than pure counting,” says Lenerius.
For marketplaces, this creates a new category of platform-level intelligence: not just how many calls each seller received, but what happened on them — aggregated across every seller, comparable across the network, actionable at both the seller and the platform level.
Three questions for your next platform review
The shift from volume to outcome reporting does not require a systems overhaul to begin. It starts with three questions a marketplace operator can ask now.
What is our connection rate? Of every call generated through the marketplace, what proportion reached a human? If you do not know the answer, that is the first gap to close.
What is our appointment rate? Of the calls that connected, what share progressed to a test drive or physical visit? This is the metric dealers understand intuitively — and the one that most directly predicts revenue.
What is our after-hours exposure? What proportion of calls come in outside business hours, when no human is available to answer? This is often where the largest share of missed calls concentrates — and the most addressable gap.
“When we bring the 42% figure — the share of calls that never reach a human — to customers, you can see people’s eyes widening,” says Lenerius. “That’s a bit of an aha moment. They had a hunch that call duration was a blunt tool, but they didn’t know the scale.”
Lead volume will always matter. It tells you whether your marketing is attracting buyers. Outcome metrics tell you whether those buyers are converting. Both numbers belong in the same conversation — and AI has finally made it practical to have it.