Shopping campaigns for equipment dealers: feed setup that wins the click

Your feed was built by a plugin, and that is the whole problem
I have opened a lot of dealer Merchant Centre accounts. Nine times out of ten the feed was generated by whatever plugin the web developer installed three years ago, approved, and then nobody looked at it again.
The symptom is always the same. Product titles that read like a stock system export. Something like "MF-7726S-DYNA6". A farmer searching for a 270hp tractor has no idea what that is, and neither does Google. The title is the single strongest relevance signal in the feed, and dealers routinely hand Google a part number.
Fix the titles first. Before you touch bids, before you restructure campaigns, before you argue about budget. In my experience it is the change that moves click-through rate the most, and it costs you nothing but an afternoon.
The title formula I use for agricultural machinery
Front-load the words rural buyers actually type. Google truncates the title in the ad, so the first 40 characters do most of the work.
For new gear I build titles in this order: Brand, Model, Category, headline spec, condition. So "John Deere 6155M Tractor 155HP New" instead of the model code on its own. For used, the year matters more than almost anything else, so it goes early: "2018 Case IH Magnum 340 Tractor 3,400 Hours".
Hours and year are the two fields used-machinery buyers scan for before they scan for price. Putting them in the title means your listing answers the question before the click, which sounds like it would cost you traffic. It does. It costs you the traffic you were never going to convert, and that is the point.
- New: Brand, Model, Category, key spec, condition
- Used: Year, Brand, Model, Category, hours
- Parts: Brand, part number, what it fits, such as fits 6R series
- Under 150 characters, but the first 40 are the real estate that counts
- No ALL CAPS
- No BEST PRICE or promotional text, Google will disapprove it
Used machinery breaks Shopping unless you set it up properly
Every used machine is a one-off. One unit, one photo set, one price, and when it sells it is gone. That is a genuinely awkward fit for a channel designed around restockable ecommerce products, and it is where most dealer feeds fall apart.
Three things matter. Give every machine a unique, stable ID that does not get recycled when a new unit lands. Set availability honestly and update it daily, because a Shopping click on a machine that sold last Tuesday is a wasted click and an annoyed buyer. And put a real price in the feed. "POA" or "Contact us" means the item gets disapproved or, worse, gets approved with a nonsense value and burns budget.
I know the objection. Dealers do not like publishing used prices because they want the conversation. I understand it, but the market has moved. Buyers are comparing three dealers on their phone in the ute before they ring anyone. If you are not in that comparison you are not in the deal, and hiding the price does not protect your margin, it just removes you from the shortlist.
Photograph the actual machine. Real photos taken in the yard, tyres and hours meter included, will beat a manufacturer render every single time on click-through. Rural buyers can smell a stock image.
Parts and attachments are what keep the account alive between seasons
Machinery sales are lumpy and seasonal. Enquiries lift ahead of seeding from about February through April, then again around the spraying window, and header interest builds from late winter into a harvest that runs November into January depending on where you are. Between those peaks the phone goes quiet.
Parts, filters, attachments and consumables do not care about any of that. They are searched all year, they convert fast, and they are the reason I always push dealers to get their parts catalogue into the feed properly rather than treating Shopping as a machinery-only channel.
The economics are very different too. A parts click might cost you a dollar or two and convert same session. A tractor enquiry in a competitive category can run a cost per lead in the $150 to $400 band, sometimes higher for high-horsepower gear, and then sit in the pipeline for three to six months. Running both in the same account, with the same budget pool and the same target, is how dealers end up concluding "Shopping does not work for us". Separate them.
Custom labels are how you actually control spend
Shopping does not give you keywords, so your control comes from how you slice the feed. Custom labels are the lever, and most dealer accounts leave all five of them empty.
Here is what I put in them for an equipment dealer. It is not the only way, but it has held up across a lot of accounts.
Once those are populated you can do the things that matter. Push more budget at pre-season into the categories that sell in that window. Bid harder on machines that have been sitting in the yard past 120 days. Split branch stock so a Wagga buyer is not clicking on a machine parked in Toowoomba. None of that is possible when every product sits in one undifferentiated bucket.
If you are not tracking to the sale, you are optimising for noise
This is the part most dealers skip, and it is the part that decides whether the account gets better or just gets busier.
A form fill is not a result. In agriculture marketing the gap between an enquiry and a sales-qualified lead is enormous, because a fair share of farm equipment leads are tyre kickers, parts questions that landed on the wrong form, or someone comparing against a machine they have already committed to. If you feed raw form fills back into Google as your conversion, the algorithm will get very good at finding you more of exactly that.
What I do instead: capture the click ID on the form, push it into the CRM against the lead, and import the stage changes back into Google Ads as offline conversions. Qualified enquiry gets one value. Quoted gets a higher one. Machine sold gets the real gross profit. Now the bidding is chasing revenue rather than form submissions, and after 60 to 90 days of that data the difference in lead quality is obvious without anyone squinting at a spreadsheet.
Long sales cycles make this harder, not optional. A header sold in December might trace back to a click in August. If your reporting window is 30 days you will never see it, and you will switch off the campaign that was working. Set your conversion windows out to 90 days minimum on machinery, keep them short on parts, and keep the two in separate campaigns so they are not fighting each other.
Get the feed right and Shopping becomes the cheapest high-intent traffic an equipment dealer can buy. Leave it to the plugin and you are paying Google to show model codes to people who have no idea what they mean.
Two questions before you go
Answer them and you will know whether your feed is working or just approved.
-
1. You have one afternoon to improve a dealer Shopping account. What do you do with it?
Rewrite the titles. The title is the strongest relevance signal in the feed, Google truncates it in the ad so the first 40 characters do most of the work, and most dealer feeds hand Google a stock code like MF-7726S-DYNA6. A farmer searching for a 270hp tractor has no idea what that is. In my experience it moves click-through rate more than anything else you can do in an afternoon.
-
2. Parts convert same session for a couple of dollars a click. Machinery takes three to six months at $150 to $400 a lead. How do you run them?
Separate them. Running both in the same budget pool against the same target is exactly how dealers end up saying Shopping does not work for us. Parts want short conversion windows, machinery wants 90 days minimum, and a header sold in December might trace back to an August click. Split them and both work.
Want this system built for your business?
Harvest Lead builds lead and sales systems for agriculture and machinery businesses across Australia. Book a call and we will map your search demand.
Book a strategy call