Reconciling the CRM record with what the customer actually said

Every customer at your store exists twice. Once as a set of entries in the CRM, typed by a salesperson between calls, and once as a series of conversations: the inbound call, the text thread, the voicemail nobody returned, the email that went to a shared inbox. Managers run the store from the first version. The customer lived the second one. Reconciling them is not an audit, it is how you find out whether your record is worth reading.

Two records of the same customer

Here is a real shape, with invented names. The CRM says: spoke with customer, interested in the black one, will call back Tuesday. The conversation says the customer asked twice what the monthly payment would be, said he needed to check with his wife before committing to anything, asked whether the store took his truck on trade, and was told somebody would text him photos that evening.

The note is not false. It is thin in a way that costs money. Nobody knows there is a trade. Nobody knows a second decision maker exists. Nobody knows photos were promised, so nobody sends them, and on Tuesday a rep calls a customer who has been waiting three days for something and has already assumed this store is not organized.

The four questions

Did the contact happen at all

Start here, because everything after it depends on it. The record says contacted on Monday. Is there a call of any length, a text exchange, an email reply. A logged activity is a claim that something happened, and a claim is not the thing.

Attempts recorded as contacts are the most common single distortion in a dealership CRM, and they matter beyond tidiness: they stop the clock on a lead that has not actually been reached, and they make a customer look handled when they are waiting.

Does the note match the conversation

Read the note, then listen to the call or read the thread. You are looking for three things: facts the conversation contains that the note left out, facts in the note that the conversation does not support, and anything the customer said that changes what should happen next.

The middle one deserves the most attention. A note claiming the customer is coming in Saturday, when the recording has the customer saying maybe Saturday, I will let you know, sends the whole store down the wrong path. Somebody holds a unit, somebody schedules a manager, nobody calls to confirm because the appointment looks set.

Was the promise kept

Find the promises in the conversation, then look for evidence they were kept. Photos said and not sent. A payoff figure that was going to be checked. A callback at a named time. This is the check that reps dislike most and the one customers feel most directly, because a broken promise is the clearest signal a store gives about how the rest of the transaction will go.

Be careful about what counts as kept. A text sent at 11:40 at night is not the photos tonight the customer was promised at five in the afternoon, but it is much closer than nothing, and the coaching is different.

Is the next step real

An action, a date, an owner. Follow up is not a next step. Follow up next week is not a next step. Call Thursday at 10 to confirm Saturday, owner Marcus, is a next step, and it is the only part of the record that causes anything to happen.

Check the dates that have already passed too. A next step dated last Wednesday that still sits open tells you the record has been abandoned while continuing to look active.

Twenty records, not two thousand

The reason most stores never do this is that they imagine the whole database. Do not. Twenty records a week, chosen deliberately, will show you everything a full audit would and will actually get done.

1. Take five records per rep from a small group, rotating the group each week.
2. Weight the sample toward records where something was supposed to happen: appointments set, quotes given, promises made.
3. Read the note first and form an expectation, then open the conversation. Doing it the other way round makes you read the note charitably.
4. Mark each record with the four questions, yes or no, and nothing more elaborate.
5. Keep the mismatches by rep, because the pattern is per person almost every time.

Forty minutes a week. The first week will be uncomfortable and the fourth will be quick, because the floor adjusts once they know it happens.

What the mismatches usually mean

Assume time pressure before you assume anything else. A rep who takes nine calls in a morning and writes notes at the end of the shift is not concealing a trade, they have forgotten it. That is a process problem with a process fix: notes before the next call.

Other common causes worth naming, because each has a different remedy.

• Template notes. Spoke with cust, will follow up, pasted for years. The rep is filling a required field, not recording anything.
• The field is not required. If the trade field can stay empty, it will stay empty on a busy Saturday.
• The rep did not hear it. Newer people miss buying signals entirely. That is coaching, not compliance.
• The note is written for the manager. A rep who expects to be questioned about a pile of open records writes notes that look active. Fix the pressure, not just the note.

Three kinds of mismatch, three responses

Missing information. The conversation had a trade, a timeline or a second decision maker and the note does not. Coach the note template, and check whether the field was even there to fill in. Low urgency, easy fix, and it is the most common finding.

Overstated information. The note claims more than the conversation supports: a maybe recorded as a yes, an attempt recorded as a conversation. Handle this in the same week, privately and directly, because the next person to touch the record will believe the note. Ask first rather than accusing. Sometimes there was a second call that never got logged.

Broken promises. The customer was told something would happen and it did not. Fix the customer's situation first, today, before you coach anybody. The rep conversation comes after the text goes out.

Make it a habit, not an audit

A one time reconciliation project produces a long list, a tense meeting, and a floor that starts writing defensive notes. A weekly sample of twenty produces a store where reps assume the record gets read next to the conversation, which is the only thing that keeps entries honest over time.

Two rules keep it survivable. Say out loud that most findings will be missing information and that missing information is a process issue, not a character issue. And when you find that a whole group has the same gap, fix the process, not the people. If nobody records the trade, your form is wrong or your standard was never stated.

What reconciliation cannot settle

It tells you what was said and what was written. It does not tell you the customer's intent, whether they would have bought, or whether a kept promise would have produced a visit. A customer who said maybe Saturday might genuinely have come. Use reconciliation for the gap between the record and the conversation, which is a fact you can act on, and leave the guessing out of it.

What to do this week

1. Block forty minutes on a fixed day and treat it as unmovable.
2. Pull twenty records, weighted toward appointments set, quotes given and promises made.
3. Read the note first, then open the conversation, and mark the four questions yes or no.
4. Fix any broken promise to a customer today, before you speak to the rep about it.
5. Give each rep one finding, in private, and say which of the three kinds it is.