From Two
Failed Agencies
to 200% ROAS

A Southern California golf cart dealer came to Fast Hippo Media after two agencies and a stretch of DIY management left their Google Ads account a mess. Budget was leaking on clicks from outside the service area, negative keywords were going unmanaged, and the website wasn’t built to convert paid traffic. We rebuilt the account structure, added the missing assets, launched a dynamic landing page built specifically for SEM conversions, and put daily optimization in place. The repair work was done in 10 days. Over Q1 2026 the account delivered a 56% drop in cost per conversion, a 64% lift in conversion rate, and a 200% increase in ROAS. Cart sales rose 73% and rentals jumped 212%.

The client is an established golf cart dealer in Southern California (kept anonymous at their request). They both sell and rent new carts, carrying Icon, Club Car, Denago, and Epic. Their market is competitive and price driven, with buyers and renters who tend to search, compare, and act fast.

They had real inventory, a solid local reputation, and a Google Ads budget they were willing to spend. What they didn’t have was an account that turned that budget into customers.

By the time the dealer reached us, they’d already been through two agencies and a period of trying to run the campaigns themselves. The account showed it, and it needed a full rebuild.

Put those together and you get the pattern we see all the time with inherited accounts: plenty of spend, plenty of clicks, very few customers.

We treated the first stretch as triage. Three people went on the account from day one: a Project Manager to run the timeline, a Client Success Manager to keep the dealer in the loop, and a Google certified SEM Specialist to do the technical rebuild.

Before we touched a bid, we spent the first days on research. You can’t fix an account you don’t understand, and rushing to optimize a broken structure only moves the waste around.

The starting point was the search terms report. It showed exactly where the money was going: which queries triggered ads, which ones converted, and which ones burned budget every day. From there we sorted demand into three buckets, buy intent, rent intent, and research only browsing, and built out brand level terms for Icon, Club Car, Denago, and Epic. A shopper typing “Denago carts for sale” is worth more than someone asking how fast a golf cart goes, and the account needed to tell them apart.

That same report seeded the first negative keyword list. Rather than paste in a generic block list, we mined the real data for the junk this account was paying for: parts and repair queries, job searches, free and cheap modifiers, and models the dealer doesn’t carry. We also mapped where paying customers actually come from against where the clicks were landing, then set a realistic drivable radius for sales and a tighter catchment for rentals.

The last piece was context. Unlike most of the country, golf cart demand here isn’t driven by the weather. It moves with dealer pricing cycles and, for rentals, the local event calendar, so we studied those patterns before rebuilding: how competing dealers structured their ads, what offers they ran, and when demand actually shifts. Underneath all of it we checked conversion tracking first, since an account optimizing toward the wrong signal just gets more efficient at wasting money. Only then did we start rebuilding.

The campaigns were repaired and optimized within 10 days: structurally sound, properly targeted, and finally pointed at the right audience. Full maturity took the rest of the quarter. Stopping the waste took 10 days. The bleeding stopped first, growth came after.

Once the account was stable, the work split into a few clear moves.

The biggest structural fix was off the ad platform entirely. The client’s existing website wasn’t built to convert paid search traffic, so we built a dynamic landing page designed around SEM intent.

It handled both sides of the business. A visitor searching to buy a new cart saw inventory focused messaging and a clear path to inquire. A visitor searching to rent saw rental availability and a booking focused call to action. The page adapted to what the searcher actually wanted, which is exactly what a generic homepage can’t do.

Dynamic landing pages work because of message match. The promise in the ad has to be the first thing a visitor sees when the page loads. When the headline, the offer, and the photos line up with what someone just searched, they relax and keep reading. When the page makes them hunt for what the ad promised, they leave, and you’ve paid for the click either way.

The “dynamic” part is literal. We used dynamic text replacement so the page adjusted to the search behind each click. Someone searching “Denago golf carts for sale” saw Denago in the headline; someone searching for rentals near their town saw rental language and their city. Technically, that is keyword and campaign data passed through the URL and swapped into the page as it loads, so one template can serve dozens of tightly matched versions without us building dozens of separate pages.

Getting there took more than a template. Each intent got its own above the old treatment, its own copy, and a single clear action, a call, an inquiry form, or a rental booking, with no competing buttons pulling attention away. We added the trust signals golf cart buyers actually look for: the brands carried, financing options, reviews, and real inventory photos instead of stock images.

Then came the technical work most people skip. Paid visitors are impatient, and Google folds landing page experience into Quality Score, so speed is not optional. We compressed images, deferred non-critical scripts, and built the page mobile first, since most golf cart searches happen on a phone. Call tracking and form tracking wired every lead back into Google Ads, so the account could bid toward real inquiries instead of guessing.

This is also where a chunk of the ad savings came from. Landing page experience is one of three inputs into Quality Score, alongside expected click through rate and ad relevance. A fast, relevant page raises that score, and a higher score lowers what you pay per click. The 28% CPC drop was not all smarter bidding. A good share of it came from finally sending high intent clicks to a page built for them instead of a homepage built for everyone.

Here’s what the account delivered over Q1 2026:

The rental number is the one that jumps out. A 212% increase tells us the dedicated rental campaign and the rental specific landing page experience tapped into demand the old setup was missing completely.

The efficiency gains matter just as much. With cost per conversion down 56% and CPC down 28%, every dollar of the same budget started reaching more of the right people and converting more of them. That’s the combination that produced the 200% ROAS lift: cheaper clicks, a better converting destination, and tighter targeting all pulling in the same direction.

A few lessons from this account apply to any golf cart business running paid search.

Inherited accounts usually need to be rebuilt from the ground up. This client had already spent money with two agencies and on their own before us, and the fix came down to structure, negatives, geo targeting, and a landing page that converts. Fundamentals, done properly, beat clever bidding almost every time.

Sales and rentals behave like two different businesses inside one account. Splitting them let each campaign chase its own kind of customer, which is where the 73% sales lift and the 212% rental jump came from.

Where the click lands decides whether the budget works at all. The dynamic landing page moved conversion rate as much as anything we changed inside Google Ads. You can buy the perfect click and still lose the customer on a page that wasn’t built for them.

Speed mattered too. Repairing the account in 10 days stopped the waste early and left the rest of the quarter for building on the gains instead of cleaning up.