CRM With Dealership Lead Source Attribution

Most stores cannot say which marketing produced last month's sales. The reason is almost never the reporting. It is that the source was lost at the front door.

When a dealer says their attribution is broken, they usually mean the reports are wrong. Sit with the data for an hour and you find the reports are accurate. They are accurately reporting garbage, because the source was already lost by the time the record existed.

The five ways it gets lost are the same at nearly every store. Everything from the website lands in one bucket called Internet, so five different campaigns share one label. Phone leads are attributed by whatever the salesperson guessed on the intake form, which is to say attributed to nothing. Walk ins get typed as walk in even when the customer drove there because of an ad they saw Tuesday. Duplicate records split one customer across three sources so no single source gets credit for the deal. And the third party lead providers all post through the same channel, so the feed itself becomes the source rather than the vendor.

None of that is fixed by better dashboards. It is fixed at the door, by making sure each inbound path arrives already labeled and stays labeled through every subsequent step. That is a CRM design question rather than an analytics question, which is why it belongs on this page rather than a reporting one.

The reporting side, once the capture is right, is covered at the lead source dashboard.

The foundation is unglamorous. Each way a lead can reach you should arrive through its own defined path with its own label, rather than everything pouring through one pipe.

The platform handles this with lead feeds and custom source feeds. A third party provider gets its own feed. Your website form gets its own. A specific campaign gets its own. A partner or affiliate arrangement gets its own. When a record is created, the source is set by the path it came through rather than by a person choosing from a dropdown at the end of a busy Saturday.

That single change usually produces the largest improvement in attribution accuracy a store will ever see, and it costs nothing but setup time. Human entered source fields are wrong at a rate that surprises people, not because anyone is dishonest but because the salesperson does not know and picks the first plausible option.

Structured inbound formats matter here too. Standard automotive lead formats carry source and vendor information in the payload, and preserving it rather than flattening it is what keeps a vendor comparison honest. More on that at ADF XML integration. Once feeds are separated, distribution can differ by source, which is usually what you want anyway. A third party lead and a referral should not be worked the same way. Routing is covered at lead routing rules.

Here is a technique most stores have never been offered, and it is the most durable attribution tool available to a dealership.

Instead of pointing every campaign at your homepage with a tracking parameter attached, give each campaign its own destination page. The platform includes a landing page builder we call lead pages, with per page URL settings so each page has its own clean, readable address rather than a query string.

Why this beats parameters is practical. Parameters get stripped when somebody copies a link into a text message. They vanish when a shopper closes the tab and searches your name three days later. They break when a platform rewrites the URL. A destination does not, because the destination itself is the signal. If the lead came from that page, it came from that campaign, full stop.

It also makes offline attribution possible, which parameters cannot do at all. A URL short enough to read aloud in a radio spot, print on a service drive sign, or hand out on a card at a community event, is a measurable offline channel. Most dealers have never measured a single offline campaign in their lives, and this is why.

The website side of the same problem, including what survives the trip from ad to form, is at website lead attribution.

At a typical store the phone produces an enormous share of real opportunities and almost none of the attribution. Somebody calls, an advisor or salesperson answers, and whatever happens next exists only in that person's memory.

AutoMail addresses this with number management, so a campaign, a listing or a landing page can carry its own phone number. A call to that number is attributable by definition. IVR routes callers to the right department instead of a general line, call forwarding escalates rather than dropping, and call logs keep the record.

Then call recording with transcription, which is the part dealers underestimate. A transcript turns a phone lead into a searchable text record, which means a manager can review a twelve minute call in thirty seconds and quote the exact sentence where it went wrong. It also means the reason a call did not convert is knowable rather than a matter of opinion. Click to call and a VoIP softphone keep outbound activity inside the system too, so the outbound side is recorded rather than reconstructed.

Two warnings. Recording law varies by state and your policy needs to reflect where you operate, so build that deliberately. And a tracked number that rings into a queue nobody answers has produced perfect attribution for a lost customer, which is accurate and useless. Fix the answering before you admire the reporting.

Two categories break every model and both matter more than the channels people argue about.

The referral. Somebody bought from you, told their brother in law, and the brother in law walked in and asked for the same salesperson. That deal will be typed as a walk in, and the marketing that produced it was a customer experience three years ago. Referral and affiliate accounts exist in the platform with affiliate automation, so a referral relationship can be a tracked object with a person attached rather than a story told in the sales meeting. Stores that formalize this find it is one of their better performing sources, which is not surprising and is almost never measured.

The dark walk in. A customer saw four things over five weeks, clicked nothing, and drove to the store. No model can attribute this correctly, and every model will assign it to something. The honest handling is to ask, at the desk, with one specific question rather than a dropdown, and treat the answer as directional rather than fact.

Which brings up the broader point. Attribution is a model, not a measurement. First touch overvalues awareness spending. Last touch overvalues whatever runs closest to the sale, which is usually branded search taking credit for demand something else created. Anyone selling certainty here is selling a model and calling it a fact. Use one model consistently, understand which way it is biased, and make decisions on trend rather than on absolute credit. Model choices are covered at marketing attribution software.

Capture is half of it. The other half is making sure the label survives eleven weeks, four salespeople and two duplicate records.

Global lead history keeps a record of where every lead went and why, which is what you need when a manager insists their team never received something. Duplicate management matters more than people expect, because the single most common cause of a source looking unproductive is that its leads merged into records created by another source and the credit went elsewhere. There is a dedicated page on that at duplicate lead management.

Then the deal itself. When a customer is desked, the record carries forward into the deal rather than starting over, so the source is still attached at the point where the money is. Deal jackets, deal chats and the deal visit log keep the paperwork and the conversation attached to the same object.

Reporting sits on top in three layers. Activity reporting for what individual people did, company reporting for store level performance, and management reporting for the roll up across rooftops. Because communication happens inside the platform rather than on personal phones, the underlying activity data is real rather than self reported, and that distinction decides whether the reporting gets used or quietly ignored.

A short opinion that will save you money. Raw lead counts by source are close to useless as a budgeting tool, because they reward whichever channel produces the most volume regardless of what that volume is worth. A source delivering four hundred form fills where nobody answers the phone will outrank a source delivering forty conversations, and the budget will follow the wrong one.

Report further down. Conversations started, appointments set, appointments shown, and deals. Compare cost per appointment rather than cost per lead, and you will find the ranking changes, sometimes dramatically.

Also report speed by source, because it is a confounder people miss constantly. A source that looks weak is often a source that arrives at 9pm and gets touched at 11am the next day. Time to first response varies enormously by channel and by hour, and until you control for it you are comparing your own responsiveness rather than the source's quality.

Worth being honest here about our own leads too, since they are one of the sources you would be measuring. Leads inside your territory are delivered to you exclusively and are not resold. Nothing is filtered or scored. We ask pre screening questions at capture and deliver every submitted lead in the zone, and problems go through post delivery replacement review rather than through a quality claim made up front. What to expect from that is on this page.

A sequence that works, in order, without a consulting project attached.

1. Inventory your actual sources. Write down every way a lead can reach the store today, including the ones nobody manages. Most stores find two or three they had forgotten, and at least one that has been posting into a shared bucket for years.
2. Split the feeds. Give each one its own lead feed or custom source feed so the label is set by the path rather than by a person. This is the highest return step and it is mostly configuration.
3. Build a destination per campaign. Lead pages with their own URLs, one per campaign, including at least one offline campaign so you finally have a measurement for it.
4. Put a number on the campaigns that deserve one. Then verify that the number is answered, because attribution on an unanswered line just documents a loss precisely.
5. Deduplicate before you report. Run duplicate management first, or your first month of reporting will understate whichever source arrives second.
6. Report on appointments, weekly, with an owner. A metric without a meeting is decoration.

Pricing is month to month with no long term contract. CRM Only starts at $199 a month if you already have your own sources and just want the software and the reporting. Programs that include exclusive local buyer leads start at $799, seller programs at $999, and combined at $1,599. Full detail on the pricing page. No DMS integration or inventory feed is required to operate any of it.

Why is our current lead source reporting wrong?

Almost always because the source is set by a person choosing from a dropdown, or because every web path lands in one bucket. Fix capture first by giving each inbound path its own feed. Better dashboards on top of bad capture just report the same error more precisely.

How do you attribute phone calls?

Through number management in AutoMail, so a campaign, listing or landing page can carry its own phone number. Calls to that number are attributable by definition, and call logs with recording and transcription keep the record of what was said.

Can offline advertising be measured?

Partially, and better than most stores manage. Give the campaign its own lead page with a short readable URL and its own phone number, then count what arrives through both. It is not perfect, but it turns an unmeasured channel into a measured one.

Which attribution model do you use?

The platform holds the source captured at first contact and carries it through to the deal. Beyond that, model choice is yours. First touch overvalues awareness spend and last touch overvalues branded search. Pick one, stay consistent, and read trends rather than absolute credit.

Do duplicate records break attribution?

Yes, and it is the most common hidden cause of a source appearing unproductive. When one customer exists as three records, the deal credits whichever record closed. Run duplicate management before you draw conclusions from a report.

Do we need a DMS integration to make this work?

No. Neither a DMS integration nor an inventory feed is required to operate the platform. If you do have a feed, My Inventory Link can ingest it so vehicle detail sits alongside the source data.

See per source feeds, campaign lead pages, tracked numbers and appointment level reporting running together. Month to month, no long term contract.