Dynamite Tool Company, a Bohemia, NY tool retailer, had three Performance Max campaigns producing 2 to 3 online sales a month; after we restructured them, the account peaked at 57 sales in a week and a $60K+ monthly revenue run rate.
Data covers February through April 2026, verified in Google Ads reporting.
Dynamite Tool Company is a professional tool retailer in Bohemia, NY, selling Milwaukee, DeWalt, Lennox, Stabila, Makita and dozens of other brands online. The Google Ads account ran three Performance Max campaigns and produced 2 to 3 online sales a month.
Budgets were set by guesswork, search terms had never been audited, and the Merchant Center feed kept getting products disapproved. The data to fix all of it was sitting in the account. Nobody was reading it.
Milwaukee was the proven performer and was short on budget. DeWalt and Lennox had been paused, restarted and mismanaged. Full Catalog spent freely with no view of what converted.
Competitor names, RC hobby searches, apparel queries and one-word terms all ate budget. "milwaukee" on its own burned $44 in one week with no sales.
"dynamite tools" searches cost $50 to $142 a week across all three campaigns, with conversion rates swinging from 2.9% to 17% week to week. It was the main driver of the ROAS swings.
Nobody had worked out that a 30% margin means a 333% ROAS floor. Without it, every conversation about whether a campaign worked was opinion.
The Merchant Center feed had recurring issues. In April, a shipping mismatch left the feed 92% disapproved, and the account dropped overnight.
Three Performance Max campaigns, real ad spend going out every day, and a handful of online orders to show for it. That is the number every result below is measured against.
No big swings. Every change was backed by the week's data and made in small steps, the way we run every Google Ads management account.
Milwaukee went from $90 a day to $95. DeWalt and Lennox restarted at $25 a day and scaled to $35 over four weeks, in $5 steps and never more than 20% at a time. Full Catalog was held steady on a set evaluation timeline.
20,000 to 28,000 terms a week. We blocked competitor retailer names, RC hobby searches ("dynamite rc tools"), apparel ("milwaukee long sleeve t shirt") and one-word terms burning $44 a week with no sales.
"dynamite tools" cost $50 to $142 a week with conversion rates from 2.9% to 17%. That tracking explained Full Catalog's volatility and produced the pause-or-keep rules.
The search terms surfaced a DeWalt construction jack cluster converting at $2 to $4 per sale. We made sure the DeWalt campaign had the budget to catch that demand.
With a confirmed 30% margin, every campaign became pass or fail. When Full Catalog posted 265% and 179% ROAS in two separate weeks, the data supported pausing it.
Milwaukee returned 893% to 1,422% ROAS week after week at $95 a day. Earlier testing showed a demand ceiling around $95 to $100 a day where ROAS collapses, so we held it there.
ROAS fell from 883% to 296% overnight. The cause was the feed: a shipping mismatch had left it 92% disapproved. We held every campaign setting, found the issue and directed the fix instead of cutting budgets in a panic.
These are peak figures from February through April 2026, labeled as peaks. The account was not at these numbers every week, and April shows what a feed failure does to any account.
For every $1 spent on ads, the client got back $9 to $13 in revenue. At a 30% margin, that is $2.70 to $3.90 in gross profit per ad dollar. February brought in about $46,000 in revenue on about $5,200 in ad spend. March was tracking toward $60,000+, so treat that figure as a pace, not a closed month.
| Metric | Before | After (peak) |
|---|---|---|
| Monthly sales | 2 to 3 | About 220 |
| Weekly revenue | Minimal | $17,236 |
| Monthly revenue | Minimal | $60,000+ run rate |
| ROAS | Below break-even | 900% to 1,289% |
| Cost per purchase | Unknown | $21.43 |
| Search term audits | Never | Weekly, up to 28K terms |
Every campaign measured against the same 333% break-even line. Pick one to see its numbers and what we did with it.
Milwaukee and DeWalt + Lennox peaked at more than four times the break-even line. Full Catalog peaked below it.
Moved from $90 to $95 a day and then held. The temptation every week was to push it higher, but earlier testing showed ROAS collapsing past $95 to $100 a day.
Restarted at $25 a day and raised $5 at a time to $35 over four weeks. The construction jack cluster did much of the lifting.
Branded "dynamite tools" spend drove most of its swings. After weeks at 265% and 179% ROAS, both under the 333% floor, the data supported a recommendation to pause it rather than keep guessing.
Weekly samples of the spend the Friday audits caught and blocked.
Tens of thousands of search terms a week show exactly what is selling and what is bleeding money. Someone has to read them, every week. For an online store, the Merchant Center feed deserves the same weekly look, which is a big part of our ecommerce marketing work.
$5 a day at a time, one-week review windows, never more than 20%. That is how an account scales without crashing.
A 30% margin means 333%. Once that number exists, every campaign decision becomes a yes or a no.
The hardest change to make is none. Milwaukee stayed at $95 a day because the data said that was its ceiling.
When the whole account drops overnight, check the feed before touching the campaigns. Diagnose first, hold settings, then fix the cause.
More results from other industries: all success stories.
We will look at your search terms, your feed and your break-even number, and tell you plainly what we would change. Or email john@designmemarketing.com.
