Baselining Your Marketplace Message-to-Sale Funnel

The internet director wants more posts. The used-car manager wants more appointments. The GM wants to know why Marketplace "isn't producing." Nobody can draw the path from a first Messenger line to a sold unit with numbers the store actually recorded. Without that path, every argument is a theory: we need more volume, we need better photos, we need a different price, we need a different channel. You cannot choose among those theories until you baseline the funnel you already have.

Baselining your Marketplace message-to-sale funnel means writing the stages, counting only events you can verify, and sitting with a month of ugly truth before you change posting cadence or pack mix. It is not a claim that Facebook publishes a close rate. It is not a CRM implementation project. It is a one-page funnel with definitions a floor rep can follow on a Tuesday.

Marketplace inquiries do not behave like website form leads. If you dump both into one "internet lead" bucket, the baseline is already wrong. Keep the comparison honest with Marketplace leads versus dealer website leads.

Why the Message-to-Sale Funnel Needs a Baseline First

Scaling a leaky process multiplies the leak. A store that posts twice as many units without knowing where conversations die will hire more people to ignore more threads.

What "baseline" means here

A baseline is a frozen definition plus a time window plus counts:

• Stages named in language the floor already uses.
• An owner for each handoff.
• A start date. Nothing before that date is "data."
• No invented percentages. If a stage has no log, the count is "unknown," not zero and not a guess.

Run the baseline long enough to include weekdays, a weekend, and at least one messy recon week. A three-day snapshot during a sale event is a story, not a baseline.

What this funnel is not

• It is not Meta's ranking system. You cannot see distribution as a stage.
• It is not a promise that more messages produce more deals.
• It is not legal advice about advertising, deposits, or out-of-state deals. Confirm those with counsel and your state regulator.

The funnel measures your store's handling of conversations that already arrived.

Draw the Stages in Language the Floor Uses

If the stages only exist in a slide, reps will not log them. Use names that match the conversation.

A practical Marketplace funnel

Keep it short enough to count by hand:

1. Live listing. The VIN is in posted status and findable. This is inventory hygiene, not a lead stage, but it is the gate. No live listing, no funnel.
2. First inbound message. A buyer-initiated message on that listing.
3. First reply. A human response from your store, timestamped.
4. Qualified conversation. The buyer answered a real question (who they are, when they can visit, what they are trying to do) or asked a specific vehicle question beyond "still available."
5. Appointment set. A time on the calendar, with a name.
6. Show. They arrived, or they completed a video walkaround you actually ran.
7. Sold or lost. Outcome on that VIN or a different unit, recorded without shame.

You can add "deposit" if your store uses it, with the same honesty: a deposit is not a sale until the deal is done. This is operational tracking, not a legal definition of a contract.

Define the messy middle

Most stores lose the baseline in the middle:

• "Still available" only. Count it as first inbound. Do not count it as qualified until there is a second exchange with substance. Scripts for that first reply live in responding to is this still available.
• Ghosted after reply. That is a stage exit, not a missing stage. Record it.
• Switched units. If the buyer came in on a Civic and bought a Corolla, credit the conversation, then record the sold VIN separately. Do not pretend the Civic listing closed itself if you cannot trace it.
• Phone call after Messenger. If the thread moves to a call, the appointment still needs a calendar row. A "we talked" in a rep's memory is not a baseline event.

Owners and clocks

Write who owns each clock:

If two people own a stage, nobody owns it. Name one.

Count a Month of the Message-to-Sale Funnel Without Cooking the Books

You are building a baseline, not a press release.

Capture rules

• One row per inbound conversation, tied to a VIN when you know it.
• First reply time on the same row.
• Appointment time on the same row, or a clear "none."
• Outcome: sold this VIN, sold other VIN, lost, unknown.
• Source field if you have one. If you do not, see source capture habits and do not invent "all of these were Marketplace."

Unknown is allowed. Fake precision is not.

A simple count sheet

Consider a store that logs every Marketplace thread for four weeks. At month end the sheet might show (illustrative structure, not a benchmark):

• Live unique units posted during the window.
• First inbound messages logged.
• First replies within the same day vs. later vs. never.
• Qualified conversations.
• Appointments set.
• Shows.
• Deals tagged to those conversations.

The useful pictures are the drop-offs, not a blended "close rate" you will quote in a meeting. If 40 first messages produced 8 appointments, the argument is about qualification and calendaring, not about posting 80 more units into the same hole.

Separate process failure from market failure

Before you conclude Marketplace "doesn't convert":

• Check response time. A funnel with slow first replies is a desk problem.
• Check listing accuracy. Wrong mileage and bait prices create messages that cannot become deals.
• Check sold-unit hygiene. Conversations on cars that already sold will never show.
• Check whether appointments are written down. Many "no-shows" were never on a calendar.

Only after those checks do you get to segment, price, and mix.

Use the Baseline to Change One Thing

A baseline that does not change behavior was a homework assignment.

Pick the widest leak you control

Typical patterns, each with a different fix:

• Inbound exists, replies do not. Staffing and inbox ownership. Do not buy more inventory for a store that does not answer.
• Replies exist, qualification does not. Scripts, training, and a rule that "still available" gets a question and a visit offer, not a one-word yes.
• Appointments exist, shows do not. Reminders, clearer lot instructions, and fewer fake times ("come anytime").
• Shows exist, deals do not. Desk, pricing, and vehicle condition. That is not a posting problem.

Change one leak for a full cycle before you change posting volume. If you change three things at once, you will not know what worked.

Freeze the definitions when you scale

When you later add associates, a second lot, or a faster posting cadence, keep the same stage names. New volume on a new definition is not a comparable funnel. Put the frozen definitions next to the monthly scorecard so the GM is not looking at two different languages.

What to tell the GM in one paragraph

"Here are the stages. Here is what we counted for [month]. Here is the largest drop we control. Here is the one process change for next month. We are not claiming a Marketplace close rate. We are claiming we can see our own handoffs."

That paragraph is the baseline. Everything else is commentary.

How PostToMarket Automates This

You cannot baseline a Marketplace message-to-sale funnel if listings have no status and messages never leave private inboxes. The operational gap is a durable record of what was posted, what was inquired on, and what was marked sold.

PostToMarket inventory statuses (draft, posting, posted, sold, failed) tell you which VINs were actually live, which is stage one. The leads inbox gives managers a single place to see inquiries and coach first-reply speed instead of reconstructing threads. Team analytics by rep and source help you see whether a leak sits with a person or with a feed. Per-rep queues, manager assignment to associates, and a team queue view keep ownership of posting (and therefore of the listing that started the thread) from dissolving into a shared login. Mark as sold and archive from inventory or the queue so "sold" is a status, not a rumor, which is the last stage of the funnel you can record without a full CRM. AI titles and descriptions, edited before posting, reduce the junk inquiries that look like top-of-funnel volume and then go nowhere. It is independent software, not affiliated with or endorsed by Meta, and it does not guarantee that Facebook will deliver messages or deals.

See the Handoffs Before You Buy More Volume

More posts into an unmeasured funnel is overtime, not a strategy. Baselining your Marketplace message-to-sale funnel with definitions the floor will actually log is how you fix the widest leak you own.

PostToMarket gives you unlimited vehicles in inventory, bulk CSV and Excel import with a row-level error log, AutoPost (one vehicle at a time, human-like delays, server-enforced daily limits), queues by role, a leads inbox, and analytics. Starter is $59 per month for a solo seller or independent lot with one login. Dealership is $299 per month when you need up to 13 logins.

Start your 10-day free trial and complete checkout with Stripe. Get posted units and inquiries into one dashboard before you argue about volume. Cancel anytime.