What is the “Auto Pulse Rank (APR)” on our platform

Definition (non-negotiable):

Car Page Rank = the probability that a listing is shown on page 1 for relevant searches.

Not clicks. Not quality. Not beauty. Visibility. If a car sits on page 2, it is effectively invisible. Dealers know this instinctively.

Key insight: cars don’t compete in one dimension — they compete in search moments

A car does NOT have one rank. It has many ranks, depending on:

• the user’s filters
• the user’s intent
• the sorting logic

So the correct question is not:

“What is my car’s rank?”

But:

“In how many relevant searches does my car appear early?”

That’s the breakthrough.

The core concept: Search-Space Page Rank (SSPR)

Instead of comparing a car to “similar cars”, we measure:

How much of the relevant search space this car dominates.

Think of it like this:

Every possible search creates a ranking list. Your car either:

• appears early
• appears late
• or doesn’t appear at all

Your Page Rank is an aggregation of all those outcomes.

Step 1 – Let’s define the dominant search axes (user reality)

From real behavior, 90% of users search by combinations of:

1. Price
2. Mileage
3. Year
4. Body size / seats
5. Fuel / consumption
6. Location

Brand is secondary in many searches. So we build rank around price–value equivalence, not brand equivalence.

Step 2 – Build “Comparable Value Buckets” (this replaces peer groups)

Our Value Buckets: A bucket is defined by:

• Price range (e.g. €28k–32k)
• Mileage range (e.g. 60–100k)
• Year range (±2 years)
• Size class (not model: compact / midsize / large)
• Fuel type

Inside a bucket you may have:

• Audi A8
• BMW 7 Series
• Mercedes E/S-Class
• Volvo S90
• Lexus LS
• Large SUV alternatives

This matches buyer substitution logic.

Step 3 – Simple classification that everyone understands

• Page 1 (positions 1–10)
• Page 2 (11–20)
• Invisible (>20 or excluded)

Step 4 – Translating numbers into “human language”

Example for a specific car listed on Auto Pulse has the following page rank:

• Page Rank: 30%

What does that mean?

• “Your car appears on page 1 in 3 out of 10 relevant searches.”

Why the rank is low or high

Now we decompose rank blockers. For each car, identify:

• Searches where it just misses page 1 (e.g. ranks 11–15)

Then analyze:

• How much cheaper cars rank above it
• How much newer
• How much lower mileage

This gives you delta explanations:

“In most searches, your car ranks just behind offers that are • €/CHF 800 cheaper or • ~15,000 km lower.”

What actually moves page rank:

Primary levers:

• Price
• Image quality (affects sorting tie-breakers)
• Data completeness
• Freshness (update cadence)

“Brand does not affect ranking. Price–value position does.”

Our Auto Pulse Rank (APR) ✔ Reflects real user behavior ✔ Cross-brand comparable ✔ Scales with inventory growth ✔ Hard to manipulate ✔ Aligns with Google PageRank intuition This is not a “score”. It’s search gravity and we call it “Auto Pulse Rank (APR)”