Engage, Qualify, Convert, Activate, Re-engage, Retain. Six stages and five handoffs. Most dealer networks measure the stages, own them in separate departments, and lose the volume in between.
Sit in a quarterly review at a large European dealer group and you will hear five numbers, reported by five different people.
Web sessions, from marketing. Lead volume, from the digital team. Contact rate, from whoever runs the BDC. Test drives and units, from the sales director. Service retention, from aftersales, usually last and usually with the fewest slides.
Every number is accurate. Nobody in the room can tell you what happened between them.
I spent a long time treating that as a reporting problem, something a better dashboard would fix. It isn’t. It is an ownership problem that shows up as a reporting problem. Five teams own six stages, and nobody owns the moments where a buyer moves from one stage to the next. Those moments are where the volume goes.
What follows is the methodology we use when we walk into a network and try to work out where the buyers went. Six stages: Engage, Qualify, Convert, Activate, Re-engage, Retain. The stages are the easy part. Naming the five handoffs between them is what changes the conversation.
A dealer group we met in Germany last year had a genuinely good website. Fast, properly photographed, a configurator that worked, and a stock feed updated hourly. Their marketing team had earned every session they were reporting.
Their test drive numbers had been flat for three quarters.
The reflex in that situation is to question traffic quality. We pulled the numbers apart instead and the traffic was fine. What was broken was the thirty seconds after a buyer decided they were interested. A four-field contact form, then nothing until someone from the BDC called back the following afternoon. By then the buyer had configured the same car on two other brands’ sites.
No individual stage was failing. The joint between two stages was failing, and nothing in their reporting could see a joint.
94.7% of German car buyers already know they can book a test drive.
Car-Buyer Experience Report 2026, n=310. Awareness is not the constraint.
Most conversion models in automotive stop at the sale. They run from awareness to purchase, the diagram gets narrower left to right, and the last box is a signed order.
That shape describes about two thirds of what actually happens in a dealer network. The buyer who did not convert this quarter is still in the CRM. The buyer who did convert is still driving a car that needs servicing, and is statistically the most likely person to buy the next one.
Four of these six stages run from first contact to booked appointment. The other two, Re-engage and Retain, are the ones that feed the top of the model again. A funnel ends. An engine returns its output to its input, which is the only reason the word engine is doing any work here.
In practice the last two stages are also the ones with the weakest ownership. Re-engagement usually sits with whoever has capacity. Retention sits with aftersales, who report separately and are rarely in the same conversation as the people who paid for the leads in the first place.
Engage is the first interaction, and the standard most networks hold it to is far too low.
A buyer lands on the site at 22:10 on a Sunday. What happens next is either a conversation or a form. If it is a form, the network has decided that the buyer’s intent can wait until Monday, which is a decision very few people in that network would defend out loud.
The same applies to launch events. A live product launch that generates questions nobody answers in the moment produces sentiment, not pipeline. Real-time Q&A, polls, and a booking route inside the stream turn an audience into named buyers while they are still watching.
What we are measuring at this stage is not sessions. It is the share of arrivals that entered a real conversation, on whichever channel they chose, at whatever hour they arrived.
HANDOFF 1 — ENGAGE TO QUALIFY
An anonymous visitor becomes someone you know something about. If nothing on the page invites that, the stage ends here for the overwhelming majority of your traffic.
Most dealer sites qualify after capture. The buyer fills in a form, the form lands in the CRM, and someone calls to find out whether the buyer is real.
That order is expensive. It puts the cost of qualification onto the BDC, and it puts the friction of a form in front of a buyer who has not yet been given a reason to complete it.
Qualifying inside the conversation reverses the sequence. Budget, timeline, model interest and trade-in get captured while the buyer is still talking, because that is when they are willing to say. The record that reaches the CRM arrives with the context attached, and the BDC starts from a position they would otherwise have spent two calls getting to.
This is also the stage where the routing decision gets made. A buyer scoring high on intent goes to a live expert. A buyer three months from purchase goes into a nurture track. Sending both to the same call queue is how a good lead and a cold one end up receiving the same treatment.
We published the five metrics we use to score this in the Lead Quality Scorecard. The short version: a qualified lead is one where the next person to touch it does not have to start over.
90% lower cost per lead, and a 3:1 conversion ratio against traditional web leads.
Measured against the network’s own prior form-based baseline, not against an industry average. The ratio comes from qualification happening before capture rather than after it.
HANDOFF 2 — QUALIFY TO CONVERT
A known buyer reaches something that can move them forward. This is the handoff with the tightest clock, and the one most European networks still run on office hours.
Speed at this joint has more effect on outcome than almost anything upstream of it, and it degrades fast. A buyer who qualifies at 21:40 on a Sunday and waits until Tuesday morning for a callback is not the same buyer by the time the phone rings. Nothing about the quality of the lead changed. The window did.
Convert is where the purchase actually gets decided, and it is the stage that resists automation the hardest.
A buyer who has said they want a specific configuration, on finance, with a trade-in, inside six weeks, has questions that need a person. What the monthly figure looks like with their deposit. Whether the trade-in valuation holds once someone sees the car. Whether the colour they want is in the pipeline or eleven weeks out.
I used to think more of this could be handled without a human than turned out to be true. On a purchase this size, buyers read the presence of a competent person as a signal about whether the brand is worth trusting. Take the person out and the conversation stalls at the exact point it was about to become a sale.
A live video session does something a phone call cannot, which is show the car while the question is being asked. The expert walks the buyer around the specific vehicle, on camera, from the buyer’s sofa. What the AI layer contributes here is making sure that expert arrives already briefed on everything captured in stage two, and arrives at the moment the buyer is ready rather than the moment the roster allows.
We went into the architecture of that split in more detail in Chatbot vs AI Sales Agent.
HANDOFF 3 — CONVERT TO ACTIVATE
A good conversation becomes a specific appointment, at a specific dealership, at a specific time. Interest that never gets a date on it is not a result.
Activate is the booking, and it is the stage where European networks quietly lose more volume than almost anywhere else.
Our 2026 research found that 69% of German buyers do their research on mobile, while 59% still complete the booking by phone or by walking in. That is not a preference. It is what happens when a mobile research journey arrives at a booking flow that was not built for one.
The second half of this stage is attendance. A confirmed booking that nobody attends costs the dealership a slot and the buyer nothing. Confirmation and reminder sequences are unglamorous work, and they move the number more reliably than most things that get board attention.
The same research found that 45% of buyers take exactly one test drive before deciding. One. Whichever brand gets that slot is competing against nothing.
45% of German car buyers take exactly one test drive before deciding.
Car-Buyer Experience Report 2026, n=310. The booking that gets confirmed first often has no competition at all.
This is also the stage that produces the most useful data in the whole model, and the stage where most networks collect none of it. What the buyer said after the drive, by model, by dealer, by market, is the closest thing to unfiltered product feedback a brand ever gets.
We wrote about why this stage is still broken in most markets in Why Booking a Test Drive Online Is Still Broken in 2026.
HANDOFF 4 — ACTIVATE TO RE-ENGAGE
A buyer who did not buy stops being written off. This is the handoff almost nobody has assigned, and it is the cheapest one to fix.
We number this stage five. In practice it is not a stage at all, it is the return path for everyone who fell out of the first four.
The buyer who qualified and never got called back. The buyer who booked and did not attend. The buyer who test drove, said they wanted to think about it, and was never contacted again. Every one of those is sitting in the CRM with acquisition cost already spent against them.
Most dealer groups are over-invested in lead generation and under-invested in lead conversion. They would rather buy ten thousand cold contacts than properly work the two hundred they already have. The reason is not stupidity. It is that new leads have an owner and a budget line, and dormant ones have neither.
Reactivation rates of around 35% are achievable on a well-maintained CRM, and that number varies a great deal with data quality, original lead source, time since last contact, and whether the original consent terms allow the outreach at all. In Europe that last condition is not a footnote.
We built the CFO-facing version of this argument in The Hidden Math on Your Dormant CRM Data, including the four hidden costs an honest reactivation model has to carry.
HANDOFF 5 — RE-ENGAGE TO RETAIN
A buyer who finally purchased enters the post-sale lifecycle, or falls into the gap between the sales system and the service system. Most networks have a literal system boundary at this exact point.
The relationship starts at delivery, and in most networks that is precisely where the attention stops.
Service reminders, satisfaction check-ins, upgrade conversations at the point in the contract where an upgrade is actually possible. None of this is new thinking. What is new is that it can now run at the same level of personalisation as the pre-sale conversation, rather than as a quarterly email blast that the customer has learned to ignore.
The commercial argument for this stage is the one aftersales directors have been making for years and losing. Retained service customers cost less to sell the next car to than any lead marketing can buy. The reason the argument keeps losing is that the saving lands in a different budget from the one paying for acquisition.
We made the longer case for this in Sales AI Got the Hype. Service AI Will Get the Margin.
THE LOOP — RETAIN BACK TO ENGAGE
A retained customer re-enters the model as the cheapest, warmest, highest-converting arrival in it. This is the handoff that turns six stages into an engine
Stage totals tell you how big each bucket is. Stage-to-stage ratios tell you where the leak is. These are the five numbers worth putting on one slide.
|
Handoff |
What moves |
The ratio to track |
What to look for |
|---|---|---|---|
|
1. Engage → Qualify |
Anonymous visitor becomes a known buyer |
Qualified conversations ÷ sessions |
Split it by hour. If the out-of-hours figure collapses, the problem is coverage, not the site. |
|
2. Qualify → Convert |
Scored buyer reaches a human |
Median minutes from qualification to first substantive response |
Report working hours and out of hours separately. The second figure is where most networks find the surprise. |
|
3. Convert → Activate |
Conversation becomes a dated appointment |
Confirmed bookings ÷ qualified conversations |
A strong conversation rate with a weak booking rate usually means the booking step sits on a different page. |
|
4. Activate → Re-engage |
Non-buyers return to the pipeline |
Dormant contacts entering an active sequence ÷ total dormant contacts |
Most networks score close to zero here and have never measured it. |
|
5. Re-engage → Retain |
Buyer enters the post-sale lifecycle |
Service retention rate at 12 and 24 months, by acquisition source |
Segment by how the customer was originally acquired. The pattern is usually not what sales expects. |
Note on benchmarks: the useful comparison is your own network against itself, market by market, quarter over quarter. Cross-network benchmarks in automotive are unreliable because almost nobody defines a “qualified lead” the same way twice.
Six numbers, five divisions, one decision.
Pull last quarter’s sessions, qualified conversations, live conversations with an expert, confirmed and attended test drives, dormant contacts re-entered into a sequence, and 12-month service retention. Divide each by the one before it.
You now have five stage-to-stage rates instead of six totals that never touch each other. One will be visibly worse than the others, and fixing a joint is almost always cheaper than buying enough extra traffic to compensate for it.
Then run the same calculation per market. A group operating across eight countries will usually find the weak joint sits somewhere different in Poland than it does in Spain, which is the whole argument for measuring by market rather than by network average.
We went through what that variation actually looks like in What “Multi-Market” Really Means When Deploying Across 20+ Countries.
Every network we have worked with had at least one stage running well. Most had three. The reason their quarterly numbers looked flat anyway is that a strong stage feeding a broken handoff produces the same result as a weak stage.
And the two stages that keep the engine turning, the ones that take a buyer who did not convert and a customer who already did and put both back at the top, are the two with the least ownership in almost every organisation we walk into.
Pull the six numbers and divide them. Whichever of the five ratios embarrasses you is next quarter’s work
If you want the data underneath this methodology, the Car-Buyer Experience Report 2026 is where most of it comes from: 310 German car buyers on awareness, booking friction, channel preference, and what actually decides the purchase.
Download the Car-Buyer Experience Report 2026
Or if you would rather see the three handoffs running live on a site like yours, book a demo and we will walk through your own four numbers.