Rental Marketing Pros
Explore Free Tools Book Discovery Call
Call Now
Roll-off dumpster on a job site
Full Case Study · Dumpster Rental

A&A Waste Solutions:
The Whole Story.

Ninety days, one dumpster operator, five counties. The baseline we inherited, what we built, what the numbers did, and the part that has not worked yet.

Roughly a 12 minute read · Published 2026-08-03

Before and After

The Whole Thing On One Screen.

Baseline measured at engagement in April 2026. Results measured May 1 to July 31, 2026.

Metric April 2026 July 2026 Change
Cost per lead Baseline 65% lower Across all Google ads
Paid leads per month 3.6 30.9 8.6x
Service area pages 0 107 +107
Local citations 2 45 +43
Mobile PageSpeed 41 / 100 99 to 100 / 100 +58 points
Site audit score 42 / 100 84 / 100 +42 points
Business details for Google None Every page Added
Website analytics None Full tracking Added
Organic sessions per month Not tracked 197 First baseline

Advertising budget roughly tripled across the same period, so these are efficiency gains on a larger budget rather than the same budget stretched further. Cost per lead fell by roughly two thirds over the same period. We publish that change and not the figures behind it. What A&A pay for a lead is their business, not something we hand to their competitors. No organic baseline exists because the previous site had no analytics at all. The site audit score is from our own tool, free and public at seo.rentalmarketingpros.com, run the same way on the same site before and after.

01

The Business

A&A Waste Solutions is a roll-off dumpster company based in New Bern, North Carolina. They run 12 and 18 yard roll-off containers, which is a deliberately narrow fleet, and they serve a 50 mile radius from New Bern covering Craven, Carteret, Jones, Onslow and Pamlico counties.

Their trucks could reach five counties, but online they existed in one town. That gap is the whole problem. A&A were not a struggling operator with a service problem, they were a good business sitting on a broken foundation. Five stars on Google and the number 2 local ranking in New Bern, in a market with roughly 24 competitors.

They engaged us on April 27, 2026. The replacement site went live on May 26, 2026. Everything measured in this study covers May 1 to July 31, 2026.

02

The Baseline We Inherited

Before writing a strategy we audit what exists. What came back was not one problem with a website. It was five, sitting in different systems, and the reason they had gone unfixed for years is that each one was holding the others back. Fix one and the other four keep failing.

Google did not know what the business was

A&A were running two competing Google Business Profiles: one storefront, one service area business. The second was never needed. A single listing can show a physical address and cover a service area at the same time, which is how a yard serving a 50 mile radius should be set up. Between them the two profiles split the reviews and the ranking strength, so neither pulled its weight the way one whole profile would have. Running two also breaks Google's rules, and when they enforce that the listing comes down and everything on it goes.

Google had the main profile filed as a Waste Management Service rather than a Dumpster Rental, and carried no business description at all.

That category error alone is enough to cap a business. Google uses the primary category to decide which searches a listing is even eligible for. A&A held second place for "dumpster rental" in New Bern on the strength of proximity and reviews, while sitting outside the top 20 for "dumpster rental near me", which is how people actually search.

The web disagreed about where they were

Search engines confirm a business is real by cross-checking it against third-party directory records. The industry calls these citations. Get enough of them agreeing and the business becomes a known quantity.

We found 9 listings for A&A across the directories search engines actually check, and only 2 of them were claimed, correct and consistent. The rest carried 11 errors between them in the name, address and phone data, including one major aggregator that had the business in the wrong town entirely and another with the wrong postcode.

So only 2 citations were of any use, against 24 for the top-ranked local competitor. If the directories do not agree on your address, Google cannot confirm you are where you say you are, so it stops showing you to people in the next town over.

The website could not be found or fixed

A&A's site ran on the GoDaddy Airo website platform. That is the root of this one. Airo hands you a set of templates and very little control underneath them, so there is a ceiling on what the site can ever do.

What that looked like in practice: no pages at all for the towns outside New Bern, and none for the services either. Everything they rent was covered by a couple of lines on the homepage saying they do it. Google ranks pages, not businesses, so with no page about dumpster rental there was nothing for it to put in front of someone searching for one. The same page title repeated across 3 pages, and 21 of 36 images with no description attached. Nothing handed Google the business details directly either, so it was left to work them out from the homepage text. Google can guess at that. It just does not rank a guess the way it ranks something it is certain about.

Someone searching for a dumpster in Havelock, fifteen minutes down the road, got nothing. With no page for a town, Google has no way to know you serve it, so it falls back on distance. The further out someone searched, the less likely A&A were to show up. Ads kept paying for those clicks anyway, and every one landed on a general homepage with a mobile speed score of 41 out of 100. A contractor on a job site with one bar of signal does not wait for that to load. He hits back and calls the next result, and A&A paid for that click either way.

None of it was fixable on Airo. The platform will not let you build the pages, control how fast they load, or give Google the details it needs. So the site was replaced rather than patched.

The website itself was unmeasured

Nobody had ever put Google Analytics or Search Console on the site, so we set both up from scratch. Google Ads was counting conversions from calls placed on the ads, but nobody could see what visitors to the site were doing: how many arrived, which towns they came from, what they looked at before leaving, or whether a visit ever turned into a call.

Tracking is not reporting for its own sake. It tells you which towns are calling, which pages turn a visit into a quote request, and where the ad budget is going to waste. Without it, every decision about the site or the campaigns is a guess. Now they come from what people actually did.

Paid media was paying for the gaps

Across the four months before we started, winning a single job through the ads ate a meaningful share of what that job was worth.

Pull the account back further and it had not moved in over a year. The monthly budget never changed once across thirteen months, and the cost of a lead over that stretch was slightly worse than in the four months we used as the baseline, so we were not picking an unusually bad patch. Same budget, same settings, month after month.

Nobody had run that account badly. They had run it without support. Bidding on a town when your site has no page for that town is expensive no matter how well the ads are built: you pay full price for the click and then waste it.

03

The Strategy

Four phases, and what matters is the order they went in. Instinct says finish the site first and turn the ads on when it is done. We did it the other way round.

Getting the listing, the site, the ads and the tracking to line up with each other mattered just as much. A clean Google listing does nothing if the website contradicts it. Build pages for every town and they sit unused unless the ads point at them. And without tracking you never find out which part did the work, which is why that went in too. Handled as separate jobs, each one helps a little on its own and nothing builds on anything else. Connected, every fix makes the next one worth more.

Landing pages are built to convert, and they do that well for a buyer who already knows what they want. What they cannot do is catch someone still working out what size they need, or searching for their own town. That is what the full site is for, and for a few weeks it did not exist. We took that deliberately, because a month with no leads costs an operator more than a few weeks of narrower coverage.

1

Fix the Entity

Weeks 1 to 2 · Apr 27 to May 10

Until Google is certain who this business is and where it operates, nothing else you do sticks. Everything after this point is built on it, so it came first.

  • Merged two competing Google Business Profiles down to one
  • Rewrote the name, categories, description, services and service area on the profile we kept
  • Built the first 35 consistent local citations, up from 2, correcting name, address and phone data everywhere it appeared
  • Started posting weekly, with a real reason to call in every post
2

Capture Demand Early

Weeks 2 to 5 · May 4 to May 31

A 116 page site takes weeks to build. The phone does not wait for it. So we put up a handful of landing pages in weeks 2 and 3 and pointed the ads at those, which meant the ads were earning while the real site was still being built.

  • Built dedicated landing pages on a separate subdomain across weeks 2 and 3, with sizes and a sticky quote form, live just before the ads relaunch
  • Relaunched their Google Ads Search campaigns in week 4, rebuilt from the account structure up rather than tuned in place
  • Launched Google Local Services Ads in week 5, which sit above the map and charge per lead instead of per click
3

Land the Real Site

Week 5 · Live May 26

A&A's old site ran on GoDaddy Airo, their website builder. Airo cannot be fully optimised for SEO, AEO or GEO, and gives no real control over how the site behaves or how fast it loads on a phone. At 41 out of 100 on mobile, it was losing people before they read a word. So we replaced it rather than patched it.

  • 116 pages live, of which 107 are service-area pages structured as county hubs, city pages and city plus service pages
  • Pages served pre-built from servers close to the visitor, scoring 99 to 100 on mobile speed
  • Every page now hands Google the business details directly, which the old site never did
  • Every town page sits inside its county page, so the whole group builds each other up instead of competing
4

Expand and Refine

Week 8 onward · From mid June

With the ads running and the new site indexed on Google, the back half of the quarter went on finding more ways in and sharpening what was already up.

  • Launched direct B2B outreach in week 8, targeting local businesses pulling construction permits
  • Sharpened the pages in weeks 8 and 10, once there was real search data showing what people were actually typing
  • Put call tracking on the whole website, not just the landing pages, so calls from Google listings and search results get counted too
  • Added a further 10 local citations in July, bringing the total to 45

How the 116 pages break down

Page count on its own means nothing. What matters is how the pages are organised, and the easiest way to picture that is folders. There is a page for each county, and inside it a page for each town in that county, and inside that a page for residential and one for commercial. So a homeowner in Havelock searching for a dumpster lands on a page that is genuinely about Havelock, not a generic page with the town name dropped into it. It also means these pages support each other instead of competing: Google reads them as one organised set covering a region, rather than 107 near-identical pages arguing over the same searches.

5 County hub pages
34 City pages
68 City plus service pages
9 Core service and conversion pages
04

The Results, Channel by Channel

Before we started, the advertising numbers cover December 29, 2025 to April 30, 2026, or 4 months. Afterwards they cover May 3 to July 31, 2026, which is only 3. Comparing raw totals across periods of different lengths would understate the change, so conversions are stated per month throughout.

Google advertising

Leads went from 3.6 a month before we started to 30.9, an 8.6x increase. Three months of it produced more than the previous thirteen months put together: 92.6 against 42.4. Those thirteen months included a summer of their own, so this is not a seasonal bounce. Same account, same platform, across a period where the budget had not moved in over a year.

Cost per lead fell 65 percent over the same stretch, and advertising spend went up, roughly 3.0x a month. So this is efficiency on a bigger budget rather than the same budget stretched further, and it is media spend rather than our fees.

Local search

A&A held their number 2 position in New Bern throughout. The expansion happened around it. They started ranking in surrounding towns too, at 4 in Havelock and 9 in Atlantic Beach, where before they appeared nowhere at all.

Organic and profile

In July, the new site recorded 384 users across 406 sessions, of which 197 came from organic search. Those are modest absolute numbers and we are not going to dress them up. The domain was two months old at that point. What matters is the mix: close to half of all sessions arriving organically that early is the signal that Google is finding and understanding the whole set of county and town pages, not just the homepage.

Their Google listing delivered 1,392 impressions, 32 direct calls and 188 website clicks during July.

AI search visibility

Measured 2026-07-31 on Google's AI answer engine, across 8 tracked brands in this market. One surface, deliberately, and the reasoning matters more than the numbers do.

Opening ChatGPT or Claude is a deliberate act. A contractor who needs a dumpster this afternoon is not doing that. They are Googling it, the way they always have, and the AI answer is simply there at the top of the page whether they asked for one or not. Nobody opts in. That makes Google's answer engine the AI surface that reaches people who never chose to use AI at all, which for a local service business is almost all of them. Google's version also reads the live web before answering, which makes it the only one local search work can move.

Of 35 brand mentions across the whole market, A&A took 18. That is 51 percent and rank 1, against 12 for the next brand and two or fewer for everyone else. A&A is also the only brand cited as a source at all. Google cited them 18 times, and every one of the seven competitors zero. Across those 18 mentions, 8 are positive, 10 neutral and 0 negative.

Those answers repeat details that only exist because of the new pages: the street address, the phone number, the container sizes, with none invented and the 2 to 2.5 ton allowance and the $65 per ton overage. Google is reading the site and the listing and quoting back what it finds there. In April there was nothing to read.

One limit, stated plainly: there is no pre-engagement tracker run, so this is a position rather than a measured change. A&A had no findable AI presence at engagement, which is an operator statement consistent with the audit, not a number we captured. Tracking on a fixed monthly cadence starts now, so the next edition can show movement rather than a snapshot.

Technical

Mobile PageSpeed moved from 41 to 99 to 100. Our internal audit score moved from 42 to 84 out of 100, with technical health at 91 and service area coverage at 91. That audit is our own tool, so the score itself is ours to define. What matters is the change, because both runs used the same tool on the same site. The tool is free and public at seo.rentalmarketingpros.com, so rather than take the score on trust you can run it on your own site and on ours.

05

What It Did to the Business

Everything above this point measures marketing. A&A's invoicing records measure the business, which is the only thing that decides whether the rest of it was worth doing.

Both comparisons below run against the same calendar months of 2025, so the busy season sits on both sides of every figure. Percentages only: what A&A invoice is theirs, and it goes no further than this page.

Before we started

January to April 2026, against the same four months of 2025 ran 11 percent behind on jobs invoiced and 15 percent behind on invoiced revenue. A&A were losing ground against their own prior year. Work began on April 27.

After we started

May to July 2026, against the same three months of 2025 came in 59 percent ahead on jobs invoiced and 13 percent ahead on invoiced revenue. Month by month, jobs invoiced ran 88 percent up in June and better than double in July, and those were the first two full calendar months with the new site live.

A business is more than its marketing, and A&A were running their own quarter over the same weeks. What the records show is four months behind the prior year, then three months ahead of it, with the line between them falling in the week the engagement started. Whether that is cause or coincidence is not something a case study can settle.

More jobs, smaller jobs

Average invoice per job fell 29 percent across the same window. Publishing the job count beside the revenue figure makes that derivable in a single division, so it belongs here rather than in whatever a reader works out afterwards.

A shift toward smaller residential jobs is the obvious reading, and it is what advertising aimed at homeowners tends to do. No job-type data sits behind that, so it stays a reading rather than a finding. Settling it comes first next quarter, and it is why the B2B outreach that started in week 8 matters more than its lead count suggests so far.

06

What Actually Moved the Needle

Four things changed at once, which makes clean attribution impossible. Anyone who tells you otherwise is selling something. What we can do is rank the levers by how much of the change each one plausibly explains, and say why.

  1. Landing page relevance, by a distance

    Cost per lead fell by roughly two thirds on an account running broadly the same campaigns. What changed most was where the traffic landed. A search for a dumpster in Havelock now reaches a Havelock page instead of a general homepage. Google charges less for a click when the page matches the search, and more of those clicks turn into calls. The two compound, which is why the cost fell by two thirds rather than a few percent.

  2. Profile consolidation

    Going from two competing profiles to one listing removes the ambiguity that was capping local visibility. This is unglamorous work with no immediate dashboard payoff, and it is a precondition for everything else. The citation build on top of it (35 in the first fortnight, 10 more in July) is what let the profile start ranking outside its home radius.

  3. Adding a channel that did not exist

    Local Services Ads did not exist on this account before. They sit above the map and charge per lead instead of per click, and they came in cheaper than search. Some of that volume is genuinely new and some is almost certainly taken from the search campaigns, so we would not claim all of it was added on top.

  4. Site speed and technical fixes

    Real, and last on this list on purpose. Going from 41 to 99 on mobile stops losing people who will not wait for a page to load, and makes the site easier for Google to read. But over 90 days it supports the other work rather than driving it. It pays off across the next year, not this quarter.

07

What Is Still Open

A case study that only reports wins is a brochure. Here is the live list at the close of the measurement period.

None of it is finished, and it is not meant to be. Tracking exists so the work can be steered when behaviour changes, not on a schedule. The AI answers section is the clearest case for why that matters: that surface barely existed a year ago, and the businesses showing up in it now are the ones whose foundations were already readable when it arrived.

Not enough reviews

A&A's profile carries a flawless rating, but the review count is thin. Rating quality is not the issue. Volume is. It is the clearest remaining ceiling on both ranking movement and on whether AI answer engines describe the business accurately, because thin review corpora give them very little to work with.

Tap cards are with the field crews and an SMS and email request sequence is the next thing to ship. As of the close of this study, that work is not done, and it is the highest leverage item outstanding.

Invisible to the models that answer from memory

A&A appeared in 0 of 104 mentions generated from training data. Answers that go and look things up are won. The ones coming out of memory are not, and no amount of work on the website fixes that. Those answers only change when the rest of the web starts writing about a business: reviews, press, directory listings, mentions that stick around long enough to be learned from. Slow work, almost all of it off the website, and the same underlying gap as the thin review count above.

Top three positions outside New Bern

Entering the local results at 4 and 9 is progress from nothing, but it is not the top three, and the top three is where most of the clicks go. Getting there is a content and review volume problem more than a technical one, and those two feed each other.

Organic volume is still small

197 organic sessions in a month is an early signal, not a channel. The county and town structure is sound and the pages are indexed, so the honest expectation is compounding over the next two to three quarters rather than anything dramatic next month.

08

Takeaways for Other Rental Operators

Your service radius and your findable radius are different numbers

A&A could reach five counties and could be found in one. Most rental operators we audit have the same gap. The fix is pages that exist for the places you actually serve, structured so they support each other.

Two Google profiles is not twice the visibility

It is a suspension waiting to happen, and the reviews go with it. If you have a duplicate listing, consolidating it is more urgent than anything on your marketing plan.

Cost per lead is usually a landing page problem

Before restructuring bids, check where the click lands. Sending geo-specific intent to a generic homepage is the most expensive common mistake in local paid search.

You cannot see what a broken foundation costs you

Marketing arrives on an invoice every month, so you always know what it costs. The jobs you lose while the foundation stays broken never appear anywhere. They keep leaking for as long as you leave it, and they compound.

Fix the entity before you scale the content

Merging profiles and cleaning up your listings is boring work with nothing to show for it on day one, and everything else you build sits on top of it. Do it first.

Curious What Your Baseline Looks Like?

We run the same audit that produced the April numbers above. Fifteen minutes, your market, no pitch.

Book a Discovery Call

Methodology and Sources

Advertising figures come from two exports. The baseline covers December 29, 2025 to April 30, 2026 (4 months) and the results cover May 3 to July 31, 2026 (3 months). A longer export covering April 2025 to April 2026 supplies the 13 month account history used for the year-on-year comparison. Lead volume is reported per month because the two shorter windows differ in length. Those counts add Google Search conversions, which Google reports as an estimate, to Local Services leads, which it reports one at a time, so the total is close rather than exact. Of the Local Services leads only the 48 that were charged for are counted; Google credited the rest back as invalid, and counting refunded leads would flatter both sides of the ratio. Cost per lead is advertising spend divided by leads and excludes our fees. Page counts were verified against the live sitemap rather than taken from the build brief. Where a figure comes from our own tooling it is flagged as first party below.

  1. Google Ads time series and campaign exports: April 2025 to April 2026 for the thirteen month account history, December 29 2025 to April 30 2026 for the baseline, and May 3 to July 31 2026 for the results
  2. Google Local Services Ads lead inbox export, May 1 to July 31 2026
  3. Local rank tracking and Google Business Profile performance report, May 1 to July 31 2026
  4. Google PageSpeed Insights, mobile, measured before replacement and after launch
  5. Rental Marketing Pros site audit tool, free and public at seo.rentalmarketingpros.com, run against the live site in April 2026 and July 2026 First party. Scored by our own tooling.
  6. Live sitemap and build output for aawastesolutions.com, verified July 31 2026
  7. Google Analytics 4 property for aawastesolutions.com, July 2026
  8. A&A Waste Solutions invoicing records, monthly invoiced job counts and invoiced revenue, January 2025 to July 2026, supplied by the client. Each 2026 period is compared against the same calendar months of 2025, so season sits on both sides of every figure. Percentages only: the underlying dollar amounts are the client's and are not published. Payments collected is excluded, because cash banked in a month is not work won in that month
  9. Rental Marketing Pros AI Visibility Tracker, run July 31 2026 across 8 tracked brands on Google's AI answer engine. No pre-engagement run exists, so these figures describe a position rather than a measured change First party. Scored by our own tooling.

Published 2026-08-03. Figures reflect the period stated and are not a forecast. Four significant changes were made concurrently, so individual attribution is estimated rather than measured, and is described as such in section 05. Results depend on market, fleet, pricing and budget, and vary between operators. The pace described in section 03 reflects the scope of this engagement and our capacity at the time. It is a record of what happened, not a standard delivery timeline.

Read the short version of this case study