Market pulse
What is this? Models ranked by an activity score built from listing turnover (appearances, disappearances, price changes). A disappearance means the ad left the market β sale, removal, or window rotation; it does not prove a sale. Click a row for the numbers behind a model.
| Model | Active | Disappeared | Turnover | Median β¬ | Days on market (med) | Disapp. 7d | Heat |
|---|
What is this? Cars priced at least 15% below the median of comparable listings β same model, year Β±1, mileage Β±20.000 km. Far-below-cohort prices are flags to investigate, not guaranteed bargains β condition is not observable from a listing. Click a row to open the listing on the source site.
| Listing | Price | Comparables median | Below median | Comparables | Location |
|---|
What is this? The trader column β local deals, imports and exports in one ranked table. LOCAL = priced below its home-market cohort median; IMPORT = foreign listing whose landed cost (price + known transport/plates/COβ tax) beats the destination median; EXPORT = home listing cheap vs a foreign median. Every row carries its comparable sample (n) and confidence β potential market spread, never guaranteed profit. Click a row to open the listing. Country-driven, not source-driven β the market switcher does not apply here.
| Type | Listing | Price | Landed | Market median | Spread/Discount | n | Confidence | Route |
|---|
What is this? Market radar β which vehicle cohorts (model Β· generation) are systematically cheaper in one market than another, after known costs (transport + plates + COβ-band tax). This is your search list: cohorts worth hunting, not one-off cars. Ranking requires β₯3 comparable asking prices on both sides β thin cohorts are never shown. Asking prices only; transaction prices unknown. Cross-market by design β the source switcher does not apply here.
| Cohort (model Β· generation) | Direction | Source median | Landed cost | Dest median | Spread | Samples | Confidence | Search hint |
|---|
What is this? Median asking price per model over time, built from every snapshot the tracker stores. Falling line = sellers cutting prices; rising = demand pulling up. Per-model is the meaningful view β the whole-market line mixes segments and is shown only for orientation. Needs a few days of tracking; resolution is daily until two weeks of history exist, then weekly.
What is this? Every tracked listing in the database β inspect the raw records behind the analytics. Filter, sort, click a row to open the ad on the source site.
| Listing | Price | Year | km | Location | Status | First seen | Last seen | Admin |
|---|
What is this? The canonical car registry β every vehicle cohort AutoPulse has ever classified, kept independently of individual ads: if every listing vanished tomorrow, this catalog survives. Body segments come from a centrally curated model map + title rules, not from whatever a single ad claims.
| Cohort | Segment | Fuel | Transmission | Observations | Active now | First seen | Last seen |
|---|
On-demand targeted scrape. Run parsers right now with dedicated filters β the same filters the CLI uses, validated against each source's config. Tick several sources to run them in parallel (one run per source at a time, 30s per-source cooldown). Targeted runs never expire other listings (sweep skipped by design). Run history lives in the database β refreshing the page never loses it.
Sources: 0 of 0 βΎ
| Started | Source | Filters | Pages | Status | Result |
|---|
What is this? The raw market heartbeat β every change the tracker detected, newest first: NEW listings appearing, price changes (old β new), EXPIRED (disappeared β sold, removed, or left the search window), REAPPEARED (back on the market). Click an item to open the listing.