Lead Source Conversion at the Group Level

Lead Source Conversion at the Group Level

AI Summary

The lead source mix at a franchise dealer group is inverted from what most general managers optimize for. The highest-volume sources, third-party internet leads and OEM digital programs, convert at the lowest rates and produce the thinnest gross. The lowest-volume sources, repeat customers, referrals, and unsolicited walk-ins, convert at the highest rates and carry the gross profit. Group-level data across the franchise dealer category routinely shows sixty to seventy percent of gross profit coming from thirty to forty percent of lead volume. Most groups optimize the wrong end of the funnel because the dashboards surface lead volume as the headline metric and treat gross-weighted attribution as a quarterly afterthought. The dealer principal making allocation decisions off the dashboard has been aiming at the wrong number for a decade.

Source: Brevmont Labs, dealer lead-source category analysis, June 2025.

---

A category that measures volume and reports volume

The dealer CRM category produces a standard morning report. Lead count by source. Touch count by source. Response time by source. Conversion percentage by source, where conversion percentage is defined as a closed deal divided by the lead count from that source.

The report is universally available. The report is structurally misleading. It treats every lead as a unit and every source as comparable on a per-unit basis. The math at the group level does not work that way.

Industry research published over the last decade by NADA, JD Power, and Cox Automotive's own analytic arm produces a consistent range for lead source conversion across the franchise dealer category. Internet leads from third-party aggregators close in the three to seven percent range across the category. OEM digital leads close in the eight to twelve percent range. Phone-up leads close between twelve and twenty percent depending on lead quality. Walk-ins close between thirty and forty percent. Repeat customers and warm referrals close between fifty and seventy percent.

The conversion rates are not a secret. The implication of the conversion rates is the part the category does not surface clearly.

The volume-conversion inversion

Pull the same data set and add lead volume. The pattern inverts.

Third-party internet leads dominate inbound count. A typical 200-unit franchise rooftop in 2024 received between three hundred and a thousand third-party internet leads per month. OEM digital programs deliver another hundred to four hundred. Phone-ups arrive at one to two hundred. Walk-ins arrive at sixty to a hundred. Repeat customers and direct referrals arrive at twenty to fifty per month combined.

The numbers vary with rooftop size, brand, and market. The shape does not. The volume hierarchy runs in inverse order from the conversion hierarchy. The largest pile by lead count converts at the lowest rate. The smallest pile by lead count converts at the highest rate. Walk-ins, repeats, and referrals together represent ten to twenty percent of lead volume and produce a meaningfully larger share of closed deals.

This inversion is consistent across rooftop type, brand, and market. The closer a dealer group looks at its own data, the more cleanly the inversion appears. The CRM dashboard does not surface it because the dashboard's primary view is volume.

Gross profit attribution amplifies the inversion

Conversion rate is not the only inverted variable. Gross profit per closed unit is also inverted.

A closed deal that originated in a third-party internet lead has shopped, on average, six dealers before showing up. The buyer is rate-shopping on the unit, sometimes shopping on financing, sometimes shopping on trade. The closer's negotiating position is structurally weakened because the buyer holds quotes from competitors. The PVR (front-end gross plus F&I gross per retail unit) on these deals runs below the rooftop average across the category.

A walk-in close, by contrast, often arrives without competitive quotes in hand. The buyer's negotiating position is weaker. The PVR runs above the rooftop average. The repeat customer or referral closes at a still-higher PVR because trust is presumed and the buyer is not shopping aggressively.

When the dealer group calculates total gross dollars by lead source, the inversion compounds. Industry research consistently finds that walk-ins, repeats, and referrals together produce somewhere between sixty and seventy percent of total gross profit at the typical franchise rooftop, despite representing thirty to forty percent of total lead volume. Third-party internet leads, despite representing the largest share of lead volume, often produce the smallest share of gross dollars.

The dealer principal allocating his marketing budget against lead volume is allocating against the wrong target. The third-party aggregator that delivers the most leads is rarely the channel producing the most gross. The unglamorous channels that are nobody's favorite agenda item in the marketing meeting are the ones funding the rooftop.

Why dashboards do not surface this clearly

A dashboard that surfaced gross-weighted attribution as the headline metric would change rooftop behavior immediately. Rep allocation would shift toward the high-conversion, high-gross channels. Marketing budget would reallocate toward the channels delivering gross, not just lead volume. Aged-inventory pricing models would change because the channel mix on aged inventory would be visible as a structural variable, not a marketing puzzle.

The category does not produce this dashboard. The reasons are structural and they line up with the category's existing customer relationships.

The CRM vendor's largest revenue lines often run through OEM program enrollments and through third-party lead aggregator partnerships. A dashboard that explicitly devalued OEM digital leads and third-party internet leads against walk-ins and repeats would damage the vendor's commercial relationships with both. The vendor's incentive runs toward maintaining lead-volume parity in the dashboard treatment, not toward calling out the gross-weighted reality.

The dealer principal does not get the dashboard he would have built for himself if he were building the CRM. He gets the dashboard the CRM vendor's commercial relationships allow.

This is not a moral failure on the vendor's part. It is the logical product of selling to a customer (the dealer principal) while maintaining program revenue from a different stakeholder (the OEM and the lead aggregator network). The dashboard treatment reflects whose interests the vendor is balancing. The dealer principal is not the only voice in the room when the dashboard gets designed.

Rooftop-level variance the principal cannot see

The volume-gross inversion runs differently at rooftops with different market structures. A high-line luxury rooftop in a metro market sees a meaningfully smaller share of repeat-customer volume than an entry-level rooftop in a rural market. A truck-heavy rooftop sees more walk-in conversion and lower internet conversion than a passenger-car-heavy rooftop. A Hispanic-market rooftop sees a different referral pattern than a non-Hispanic-market rooftop.

A dealer principal running a multi-rooftop group cannot easily see the variance between rooftops because his CRM dashboard reports each rooftop separately and does not surface the variance as an actionable view. The principal who has eight rooftops sees eight separate dashboards, each averaging out the rooftop's mix into the same vendor-defined report. The variance is invisible because the analytic frame is rooftop-by-rooftop, not channel-by-channel across rooftops.

What he is missing is the question of where his marketing dollars and his rep allocation actually produce gross. That question requires a different cut of the data than any major vendor surfaces today. The cut is technically straightforward. It is not in any dashboard the principal has been sold.

What gross-weighted attribution would look like

The headline view in a gross-weighted dashboard would not be lead count. It would be gross dollars per source, per rooftop, on a trailing twelve-month basis. The view would rank sources from highest gross-dollar contribution to lowest. The view would surface, at each rooftop, the channels that are funding the operation and the channels that are absorbing rep attention without paying for it.

Below the headline, the view would show conversion rate and average PVR per source. The principal could see at a glance which sources are converting at low rates with high PVR (often walk-ins and referrals) and which are converting at high rates with low PVR (rare, sometimes specific OEM-incentive deals). He could see which sources are converting at low rates with low PVR (the third-party internet leads producing the bulk of the volume).

Below that, the view would show rep allocation by source. The principal could see which reps are absorbing the most third-party internet lead volume and producing the least gross. He could see which reps are allocated to repeat-customer reactivation and producing the most gross. The pay-plan implication of this view is significant. Most rep pay plans are a flat-percent commission against gross. They do not reward differential allocation toward higher-gross channels. The dashboard does not produce the data that would justify a differential pay plan.

This is the dashboard the dealer principal has needed for a decade. The category did not produce it. The reasons are structural.

What the next layer surfaces

The execution layer Brevmont is building above the dealer CRM treats lead source conversion economics as a first-class view. Gross-weighted attribution becomes the headline. Lead volume becomes a secondary metric, useful for capacity planning and not much else. The view is rooftop-aware, channel-aware, and rep-aware in a single composition.

The view's data model does not require the CRM vendor to cooperate. The data sits in the CRM. The witness layer captures what actually happened at the customer interface, including the source attribution, and writes the gross-weighted aggregation back into a view the dealer principal can read against the AP report and the marketing budget.

This is not a feature the incumbents will copy in the next release cycle. The incumbents have program-revenue reasons to keep lead volume in the dashboard headline position. The dealer principal will buy the new view from a different vendor because the existing vendors have demonstrated, across a decade of release cycles, that they will not.

The dashboard the category should have produced is the dashboard the next layer produces. The dealer principal who has been allocating marketing and rep attention against the wrong number for a decade gets the right number for the first time.

That is the architectural argument we are making with the work.

---

*Brevmont Labs publishes original research on the execution layer beneath relationship-driven sales. The conversion-rate ranges in this essay are drawn from publicly available NADA and JD Power industry research and represent typical category ranges, not single-rooftop observations.*