Skip to content
AutoLens logo
← All articles

Duplicate Leads Cleanup: A Dealership Playbook

Duplicate Leads Cleanup: A Dealership Playbook

Dealer compares two customer inquiry records

To clean up duplicate leads, decide what counts as a duplicate, find matches by normalized phone and email, merge only the clear ones into the record with the fullest history, send the uncertain ones to a review list, and then fix the entry points so the same shopper stops getting three records. The order matters. Most damage comes from a blunt “delete everything with the same email” rule that throws away a second vehicle inquiry along with the true duplicate.

This guide is written for independent and mid-size car and powersports dealers who take leads from a website, Facebook Marketplace, phone calls and text messages, and who want one clean record per shopper.

Autolenstudios
Keep Dealership Leads Moving
AutoLens helps dealerships manage Marketplace listings and buyer inquiries, while tracking leads and performance in one platform.
Visit

TL;DR

  • A duplicate lead is a second record for the same shopper. A second inquiry about a different unit is not a duplicate, so write that rule down before you merge anything.
  • Find matches by normalizing phone numbers to the last 10 digits and lowercasing and trimming emails, then group by phone.
  • Keep the record with the fullest timeline, the right source and the newest contact details. Merge clear matches, review uncertain ones, and log every merge.
  • Prevent repeats: search before creating a record, attach new inquiries to the open lead, and run every channel into one pipeline.
  • Do the cleanup in a quiet week, not mid-weekend rush, and repeat a light version monthly.

Table of Contents

What counts as a duplicate lead (and what does not)

A duplicate lead is two or more records that represent the same prospect. In practice, the same person ends up in a CRM more than once because they came through different channels, someone imported a list, or a salesperson typed a new record by hand instead of searching first.

In a dealership it usually looks like this: a shopper submits a form on your site on Monday, calls from a different number on Wednesday and replies to an old text on Saturday. That produces three records with three owners, and nobody sees the whole conversation.

The harder question is what is not a duplicate. A customer asking about a used truck and, a week later, about a side-by-side is one person with two inquiries. Finding their existing contact does not make the second request a duplicate: the right result is one contact, two inquiries and a next action for each.

A lead that went cold last year and comes back is another judgment call. Depending on the team’s rules, it may be merged into the old lead or kept closed while the new one is worked fresh. Neither is wrong. Choose one policy and apply it everywhere.

What counts as a duplicate lead (and what does not)

Three things to separate before you touch the data

It helps to keep three identifiers apart:

Unit What it is Dealership example
Person (contact) The human being One shopper, one phone number
Inquiry One request “Is the 2021 Tacoma still available?”
Source event One delivery of that request The same web form pushed twice by your website

The three-way distinction of contact, inquiry and source event matters because each needs a different decision: match the person, keep each genuine inquiry, and recognize a repeated delivery before it creates more work. Another guide makes the same point about writing down your units before merging: a customer asking about two vehicles should not lose the second opportunity because their email address matches.

Why duplicates cost a dealership more than tidiness

Duplicates hurt in three places.

Response. When one shopper has three records with three owners, two salespeople may text the same person within an hour, or each assumes the other has it. Neither looks organized, and the shopper picks the dealer who answered once and clearly.

Reporting. Lead counts inflate, source attribution splits, and your cost per lead looks better or worse than it is depending on which copy survived. Speed-to-lead and aging numbers become unreliable when the clock runs on a shadow record nobody works.

Consent. Text messaging depends on a clean record of who opted in and who opted out. If the opt-out sits on one copy and a salesperson texts from the other, you have a problem a tidy database would have avoided. Our guide to SMS compliance for dealerships covers opt-in, opt-out and records.

Step 1: Back up and set your rules

Before you merge anything, export the full lead table as a file you can restore from. Then write a one-page rule sheet. Four decisions go on it:

  1. What is a match? For most dealers, the same normalized phone number is a strong match; the same email is a strong match; the same name alone is never enough.
  2. Which record wins? Use the survivorship rule below.
  3. What goes to review? Anything where phone matches but names differ (a spouse or a parent calling about the same vehicle), or where the vehicle of interest differs.
  4. Who is allowed to merge? Limit merge permission to one or two people. The same planning applies if you hand the cleanup to someone else: they need the definition of a duplicate, the winning-value rule, the review list and a clear permission level before they start.

The survivorship rule most dealers can live with: the surviving record should have the most complete activity timeline, correct source attribution, the most recent contact details and the strongest consent history. “Oldest record wins” is tempting and often wrong, because the oldest one is frequently the empty shell from a single form fill.

Step 2: Find the duplicates

The mechanical method is simple and works in a spreadsheet. Export a week of new records, normalize phone numbers to their last 10 digits, lowercase and trim email addresses, and group by normalized phone to surface collisions.

Why normalize? The same number appears as (555) 123-4567, 555.123.4567, +1 555 123 4567 and 5551234567. A system that compares raw text sees four different people. Stripping everything but digits and keeping the last ten makes them one.

Work in this order:

  1. Group by normalized phone. This catches the largest share of dealership duplicates because shoppers text and call more than they email.
  2. Group by normalized email.
  3. Look at groups where phone matches but the email differs, and the reverse. These are your review list.
  4. Check each group against the sources it came from: website form, Marketplace message, phone log, text thread, walk-in entry.

Where the lead lives in your CRM changes what a duplicate looks like. In some systems a lead sits on a contact, so a second sign-up shows up as a duplicate contact first; in others it is a standalone record. Know which one you have before you search.

Start with a week or a month of data rather than the whole database. A small batch lets you test your rules and see the kinds of false matches you get before you apply them to thousands of records.

Step 3: Decide to merge, keep or delete

Sort every group into one of three buckets.

Situation Action
Same phone, same email, same vehicle, same time window Merge into the survivor
Same person, different vehicle or a request months later Keep one contact, keep both inquiries as separate opportunities
Same phone, different name Send to review; do not auto-merge
A web form delivered twice within seconds Remove the repeated delivery, keep one inquiry
Empty or junk records with no activity Delete after backup

Two cautions. First, an uncertain match belongs in a review queue, not in an automatic merge. Second, a blanket rule such as remove matching emails can damage sales history while leaving duplicate event delivery untouched. Your service department, your parts counter and a buyer’s spouse can all share one phone number or email for good reasons.

Worked example

A shopper named Dana fills out your site form about a used crew cab on Monday (record A, source: website). On Wednesday she texts from her work phone about the same truck (record B, source: SMS, different number). On Saturday she messages about a trailer through Marketplace (record C, source: Marketplace).

  • A and B: same name, same vehicle, three days apart, one shared email on the form and in the text signature. Merge, keep A’s form details and B’s newer phone number, and keep the full text thread.
  • C: same person, different unit. Keep it as a second inquiry on Dana’s contact with its own next step, not as a second contact.

Result: one contact, two opportunities, one owner, and the whole conversation in one place.

Step 4: Merge without losing history

A merge should never reduce what you know. Before you confirm each one:

Do the first fifty merges by hand and read the survivors afterward. If the rule picks the wrong record more than once or twice in that batch, change the rule before running a bulk pass.

Timing also matters. Run the cleanup on a slow weekday morning. Merging a lead while a salesperson is mid-conversation with it is how replies get lost. If you take weekend leads, make sure someone owns them until the cleanup is done; our weekend lead coverage plan shows one way.

Step 5: Stop duplicates at the door

Cleanup without prevention is a monthly chore forever. Three habits remove most of the cause.

Search before you create. Validate data at entry and search for an existing customer before creating a record. The frontline version is a rule for every salesperson and BDC rep: search by phone and email first, create only if nothing comes back.

Attach, do not spawn. Configure your system to attach a new inquiry to an existing open lead when the phone number matches rather than creating another record. If your CRM cannot do this on its own, the same logic applies to your intake process, and a guide on CRM configuration to prevent duplicate leads walks through the response-time and consent side of capture automation.

Do not discard the next sale. A rule that ignores every message from a known email address removes duplicates by throwing away repeat business. Match the person, then decide whether the message is a new inquiry, a continuation or a repeat delivery. For handling returning shoppers without losing the new opportunity, this piece on managing repeat inquiries is worth a look.

Put every channel in one pipeline

Duplicates multiply when each channel has its own inbox. Marketplace messages live in one app, texts on a salesperson’s phone, form fills in an email alert. The fewer places a conversation can start without being logged, the fewer records you have to reconcile.

Answering fast helps too. When a shopper gets one prompt reply on the channel they used, they are less likely to call, text and fill out a form in parallel hoping someone responds. Our five-minute response test explains how to measure that, and our notes on Facebook Marketplace messaging cover the channel where many powersports and used-vehicle shoppers start.

A monthly routine that takes under an hour

Once the backlog is cleared, keep it small:

  1. Export the last 30 days of leads.
  2. Normalize phone and email and group collisions.
  3. Merge the clear ones, review the rest.
  4. Check the count of new records created by hand versus by a channel; a rising hand-entered share means the search-first habit is slipping.
  5. Skim the merge log for anything that looks wrong.

Powersports dealers should also run this check at the start and end of the riding season, when inquiry volume jumps and last year’s cold leads come back through new forms.

Where AutoLens fits

AutoLens is not a deduplication tool, and it does not try to be. It is focused software for the front end of the funnel: it captures inventory from your website or feed and creates AI-written Facebook Marketplace listings through a Chrome extension, and its AI Responder answers buyer inquiries by SMS in your dealership’s voice, handing off to your staff when negotiation or personal follow-up is needed.

For duplicate control, what matters is that leads are worked through one nine-stage sales pipeline with role-based dashboards and task management, so each shopper has one place in the process and one stage at a time. Performance metrics such as speed-to-lead and inventory aging then read from that same pipeline. You can see how the stages work on the sales pipeline page and the four views on the dashboards page.

What we think

We build for independent and mid-size dealers, and the pattern we see in how the work is organized is consistent: duplicate records are a symptom of conversations starting in too many places. Fix the entry points and the cleanup shrinks.

Our view is that a dealership is better served by one focused tool for Marketplace listings and SMS than by a pile of features it never opens. That is why AutoLens does two things on purpose: it creates listings from your actual unit specs and answers inquiries by text in your voice, so the first reply is fast and consistent whoever is off the clock.

The AI Responder replies right away and hands the conversation to your people when it turns to price or a personal follow-up. A shopper who gets a real answer in minutes has little reason to submit a second form, call a second number or message a second salesperson. Fewer parallel threads means fewer records to merge, and a pipeline with nine clear stages means everyone can see where a shopper stands.

Our advice: write down what a duplicate is, merge the clear cases, review the rest, and give every channel one door.

Adam

Put one front door on your leads

If you list inventory on Facebook Marketplace and take inquiries by text, AutoLens gives those two channels one workflow: listings generated from your own unit data, SMS replies in your voice, and leads tracked through a nine-stage pipeline. Plans are priced per month: Extension at 79 CAD, Individual at 130 CAD and Team at 219 CAD, with extra users at 100 CAD per month each. Dealer Groups has no published price. See the pricing page, or book a demo to see it with your inventory.

Autolenstudios

If your main need is qualifying shoppers once they reply, our list of lead qualification questions pairs well with the cleanup above, and the AI Responder page shows how SMS replies work.

FAQ

What is a duplicate lead?

A duplicate lead is a second record for a shopper who is already in your system, usually created when the same person comes in through a different channel. It is not the same as a new inquiry about a different vehicle, which belongs on the existing contact as its own opportunity.

How do I find duplicate leads in my dealership CRM?

Export a recent batch, normalize phone numbers to the last 10 digits, lowercase and trim emails, and group by phone, then by email. Review groups where only one of the two matches. Start with a week or a month of records to test your rules before running the whole database.

Should I delete duplicates or merge them?

Merge them whenever the records hold any history, so notes, texts and consent status survive on one record. Delete only empty or junk records with no activity, and only after a backup. Log each merge so you can reverse a mistake.

How often should I clean up duplicate leads?

Do a full cleanup once, then a short monthly pass on the last 30 days. Add a check at the start and end of your busy season, when inquiry volume and returning shoppers rise.

How do I stop duplicates from coming back?

Train staff to search by phone and email before creating a record, attach new inquiries to an open lead with the same phone number, and route every channel into one pipeline. Fast first replies also reduce the number of parallel inquiries a single shopper sends.

Sources