Monitor soft-ask metros, message where demand clears

Cross-metro sourcing is one tight search on a few cities, not a vague nationwide keyword. You watch where asks run soft, and you message, pick up, or ship toward where you can clear the unit. Cities and dedupe replace twenty Marketplace tabs. Crawlbench surfaces the truck on its next check; the freight and the resale buyer are yours to arrange.

You will not find a thick public playbook that measures this query. The homepage already describes multi-city coverage. This post is how to use it without turning the queue into noise. The click-by-click city picker lives in the multi-city searches guide. Read that when you need the control. Read this when you need the sourcing model.

What is a soft-ask metro, and what is a demand metro?

A soft-ask metro has more supply, or softer stickers, for the exact thing you buy. A demand metro is where that thing clears faster, at retail or into a wholesale group you already use.

The job splits:

• Monitor the soft-ask side so new listings hit one queue.
• Message and move the unit toward the clear. That logistics is yours. A search does not book transport or auto-route inventory.
• Keep one set of filters. Same identity, same excludes, same price band in every city on the list. Forking filters per tab is how the twenty-tab mess starts.

Do not invent a dollar spread between two cities and call it a system. Asking medians move, and a homepage illustration is not your margin. If you quote a band, take it from Market Snapshot or a week of asks you actually saw, and label it as asking prices.

Do you need a separate search per city?

No. One search can cover a custom city list, or every city nationwide, with a radius on each city.

Use the modes this way:

• Single city when you only pick up in one metro. Extra cities are vanity.
• Custom city list when you have two to four metros you will triage this week. This is the normal sourcing setup.
• Nationwide when the identity is extremely tight (one nameplate, one year band, a hard price ceiling). Wide keywords plus all-cities is a noise machine.

Cities are the budget. A Day Pass includes one search and 20 cities for 24 hours, which is Pro-depth coverage for the day, not a reason to add twenty metros. Spend them on cities you can drive or ship this week. The Day Pass post lists the rest of the window, including a 7-day Snapshot.

Crawlbench checks busy cities every 10 minutes on Starter, every 5 on Growth, every 2 on Pro and every minute on Scale, as pricing shows on 18 October 2026. That holds whether the search is a custom list or nationwide.

Will the same listing ping you three times?

No. If Facebook surfaces the same listing against three cities, Crawlbench keeps one listing. That dedupe is the point of leaving the tabs behind. Three pings for one truck train you to ignore the channel.

Radius still matters. A city plus a radius can pull a town you will not drive. The multi-city guide covers the distance re-check. If a row is outside the trip you will make, pass it. Do not add another city to “catch” a town that was a radius spill.

What does a worked lane look like?

Pick one lane. Here it is midsize trucks. The same shape works for open-box laptops if that is what you move.

1. Write the identity in one line: nameplate, year band, condition you will actually buy.
2. Set a ceiling from recent live prices. A tight cohort median is enough to see which asks sit low. When the median ask is enough is the longer version of that.
3. Choose two or three supply metros where that truck is common, and name the place you clear (a retail city, a wholesale group, or the same region). You do not need a fourth city on day one.
4. Add the excludes you already trust so multi-city does not multiply salvage and parts ads. The excludes post is the pattern. Do not paste a vehicle exclude list onto laptops later and hope.
5. Success check: can you message the listings rated Great price and Good price the same day, from one queue?

If the queue is still junk, remove a city or tighten excludes before you add coverage. More geography multiplies whatever the gate allows.

The sourcing desk post is the weekly operating rhythm. This post is only the geography decision inside that rhythm.

When does multi-city fail?

• Vague keywords plus all-cities. You asked for the country and you got the country.
• More cities than you can triage. A listing you will not open is a city you wasted.
• Alerts turned on before your filters are clean. Mute is a symptom. Fix excludes first.
• Adding a city to catch radius spill instead of passing the row in triage.
• Chasing an arbitrage dollar figure you did not measure. If you cannot say the ask band out loud, you are not ready to add a metro.

How do you pick the first two cities without a fake spread?

Start from trips you already make.

Name the metro where you will stand in a driveway this week. That is a supply city only if the asks there are the ones you buy. Name the metro, group, or lot where those units leave your hands. If both are the same city, you do not have a cross-metro operation yet. You have a local search, and that is fine. Add a second city only when you have driven it or shipped from it in the last month.

Then look at asking prices, not a story about what sold. Open a handful of live listings in each city for the same year band. If one city is consistently softer on the ask, it earns a slot. If you cannot see a difference without inventing a number, do not add the city to sound sophisticated. Snapshot, when the sample is thick enough, shows the band. A thin sample should stay No comps yet, not a made-up gap.

Write the cities down as a list of four at most. If the fifth city is “maybe,” it is not on the search. You can add it after the first queue stays readable for a day.

How do radius and dedupe interact?

Radius answers “how far from this city will I go.” Dedupe answers “how many times will I see this listing.”

Set the radius to a drive you will actually do from that city. A huge radius makes a custom list behave like a sloppy nationwide search. Dedupe still collapses the same Facebook listing into one result, which is good, but it will not collapse two different trucks that happen to look alike. You still read the row.

If a listing is inside the radius and outside your real route, pass it in triage. Do not “fix” that by adding another city on top. The multi-city searches guide is the place for the control labels. Use it after this choice is made, not instead of it.

Prove two or three cities before you add the rest

A Day Pass is the right test: one search, up to 20 cities, 24 hours, no card. Use two to four cities, not the whole cap.

Put the soft-ask metros and the clear market on one search. Create the search, read the listings it finds, and see whether you can message from a single queue. If you can, keep the list. If you cannot, delete a city. Scale only after the first queue stays clean.

Pricing is where the city caps live if you move off the pass. Features states the dedupe in one line: the same listing in three metros is one match.