Inventory Search UX for Dealer Websites

Most dealership search results pages are built for the dealer's mental model, not the shopper's. That gap is measurable in leads.

A dealership website has one job that matters and several that feel important. The job that matters is turning an anonymous visitor into a conversation. Everything else, the awards page, the staff photos, the finance explainer, is scenery around that.

Inventory search is where the conversation is won or lost, because it is where nearly all of your engaged traffic goes. A shopper who lands on your home page and leaves cost you nothing. A shopper who ran three filters, opened two vehicles and then left was interested, and you have their behavior and none of their contact information.

That is why search design deserves more attention than it usually gets. In most stores the results page was configured once, by whoever set up the site, using the categories a dealer thinks in: year, make, model, trim, mileage, price. Shoppers think in different terms. They think in monthly payment, in body style, in how many people fit, in whether it will tow the boat, in whether it will start on a January morning. A filter set that does not speak their language quietly sends them to a third party marketplace where the filters do.

We are not a website vendor. LeadLocate does not sell a full dealership website platform, we are not an OEM certified provider, and we do not build OEM compliant websites. If you need a new site, evaluate a website provider on its own merits and use the questions further down.

What we do build and can be held to is narrower and useful. My Inventory Link ingests and stores your inventory feed, exposes a public inventory API, and publishes per salesperson inventory sites from a template through a build and publish pipeline, with a data manager behind it and key provisioning for outside consumers. Lead pages is a landing page builder with per page URL settings, which is how most of our dealers run campaign specific vehicle pages without touching their main site. A free live chat widget drops onto the site you already have. And the whole CRM sits behind all of it, which is where a search visitor becomes a worked lead rather than a session in an analytics report.

One thing worth stating because it is unusual: an inventory feed is not required to operate the platform at all. Inventory Link is there if you have one. Brokers and individual salespeople run without it. See the website platform page for the fuller version of where we do and do not fit.

If you change one thing on your results page, change the filters. Here is what tends to work, based on how shoppers describe cars to salespeople rather than how software categorizes them.

Lead with body style, not make. A large share of used shoppers are not brand loyal. They want a midsize SUV under a number. Make first forces a decision they have not made yet and hides half your lot.

Put payment next to price. Most buyers are financing and think in monthly terms. A payment filter is closer to how the decision is actually made, provided every displayed figure is clearly labeled as an estimate with the assumptions stated.

Kill the empty result. A zero result page is where sessions end. When filters return nothing, show the nearest matches and say so plainly instead of showing a blank page with a sad message.

Show counts on every option. If a filter value would return three cars, say three next to it before they click. Nothing burns goodwill faster than four clicks into a dead end.

Keep the filters visible on mobile. Filters behind a modal that resets on back navigation is the single most common defect we see, and most traffic is mobile.

Default sort is a business decision disguised as a technical one, and most stores never make it deliberately. Newest first favors your freshest stock. Price ascending favors the shopper and tends to increase engagement while pulling attention to your thinnest units. Aged units first moves the metal you need to move and can feel like a bait and switch if the top results are the worst cars on the lot.

There is no universal right answer, but there is a wrong process, which is letting the vendor's default stand for three years because nobody looked. Pick one, then check it against actual behavior: how deep into results do shoppers go, which sort do they switch to, and which vehicles produce contact rather than views.

The related trap is hidden pricing. A results page with call for price on a meaningful share of units trains shoppers to leave, because they assume they are being handled rather than informed. If your policy requires it on certain units, fine, but understand the cost and keep it rare. Our page on vehicle merchandising software goes further into how listings should be built.

Search results are judged on images before anything else. A row of listings where three cars have real photographs and the rest have stock images or a placeholder does not read as an inventory shortage to the shopper. It reads as a site that is not maintained, and it damages the listings that are complete.

Speed is the other half. Results pages are heavy: dozens of images, filter logic, and usually a stack of third party tags. Every one of those tags was added by someone with a good reason, and collectively they are why your page takes six seconds on a phone with two bars in a parking lot. Audit them annually and remove what nobody reads reports from.

Two practical rules. Load images at the size they will display rather than shipping a full resolution photograph and letting the browser shrink it. And reserve the space each image will occupy so the page does not shift under a thumb that is already reaching for a listing. Layout shift on a results page produces accidental taps, which produce immediate exits. Core web vitals for dealer websites covers the measurement side.

Suppose your search works perfectly. A shopper filters, finds three candidates, opens two, and leaves without filling out anything. That is the normal outcome. Most people are not ready to hand over a phone number the first time they see a car, and no amount of search polish changes that.

So the question is not only how good the search is, it is what you offer someone mid search who is not ready to be a lead yet. A single form that says get more information is a weak offer. Better offers are specific: a payment estimate on this exact vehicle, an availability check, a trade estimate request, a question answered by a human right now.

Our free live chat widget covers the last one and drops onto the site you already run, so a shopper three filters deep can ask whether the silver one has a third row without abandoning their session. Live chat for car dealers covers it properly. Lead pages covers the rest: campaign specific landing pages with their own URL settings, built for one vehicle or one offer, which convert far better than pointing paid traffic at a generic results page. And every conversation, whether it started in chat, on a lead page or on the phone, lands in the same inbox with the same follow up rules.

This is the part that tends to surprise people, so it is worth describing concretely. My Inventory Link can publish a per salesperson inventory site from a template, generated from your feed through a build and publish pipeline. Your salesperson gets a real site showing your real inventory, under their own name.

Why that matters: shoppers increasingly search for a person rather than a store, especially in used cars and especially through social media. A salesperson who can send a link to their own inventory page, with their own contact details on it, gets replies that a corporate results page URL does not. It also gives your people something to post that is theirs, which is the difference between a team that markets and a team that intends to.

The feed side is standard: ingest, storage, a data manager, and a public inventory API with key provisioning if you need to feed something else. If you already have a website vendor you are happy with, this runs alongside it rather than replacing it. More on salesperson websites.

Since we are not selling you a site, here are the questions we would ask if we were buying one. What is the measured mobile load time of a results page with fifty vehicles, on a real device, not a lab score. Can we change the filter set ourselves, or is that a support ticket with a turnaround time. What happens on a zero result search. Can we control default sort per inventory type. Who owns the site and the content if we leave, and what exactly do we get back. What third party scripts are on the page by default and can we remove them.

Get the answers in writing before signature, because a support ticket six months in has none of the persuasive power of an unsigned contract. The website RFP template puts the whole list in order.

On our side: an inventory feed is optional, no integration with the dealer management system you run is required, and pricing is month to month with no long term contract. CRM Only starts at $199 a month, and plans that include exclusive local leads start at $799. Everything is on the pricing page, or contact us and we will tell you straight whether the piece you are missing is one we build.

Do you build dealership websites?

No. We do not sell a full dealer website platform and we are not an OEM certified provider. We provide inventory feed ingest, salesperson inventory sites, lead pages and a free live chat widget that runs on the site you already have.

What does My Inventory Link actually do?

It ingests and stores your inventory feed, exposes a public inventory API with key provisioning, runs a data manager, and publishes per salesperson inventory sites from a template through a build and publish pipeline.

Do we need an inventory feed to use LeadLocate?

No. A feed is not required and neither is an integration with the dealer management system you run, which is why brokers and individual salespeople operate on the platform. If you do have a feed, Inventory Link can use it.

What is the most common inventory search mistake?

Filters behind a mobile modal that resets on back navigation, closely followed by zero result pages with no nearest match fallback. Both end sessions, and most of your search traffic is on a phone.

Should we show payments in search results?

It matches how most buyers think, provided every figure is clearly labeled an estimate with the assumptions shown. Our desking engine produces the numbers on the deal side, so what the customer later sees stays consistent with what the desk is using.

Will better search UX produce more leads?

We cannot guarantee lead volume or conversion, and no vendor honestly can. What we can say is that filter design, zero result handling and mobile speed are the cheapest things to fix and the easiest to measure before and after.

See the chat widget, lead pages and salesperson inventory sites feeding one inbox with real follow up behind it. Month to month, no long term contract.