The Branded Keyword Trap

Tune Up Fitness had been running Amazon ads for about two years before connecting Astra. Three basic campaigns, modest spend, and on paper the results looked acceptable. Some months delivered 4 to 8x ROAS on branded searches.

Here is what that number was actually measuring. Those shoppers were already searching for the brand by name. They knew the product and intended to buy. The ads were intercepting sales that would have happened organically, not generating new customer acquisition. The ROAS looked strong because the targeting was easy.

Amazon's algorithm does what it should: optimize for what converts. Branded searches always convert at higher rates than category searches, so budget flows there, which trains the algorithm further. After 12 to 24 months without active intervention, most sellers end up almost entirely branded with no visibility on what they are missing.

By late 2023, the campaigns had stalled entirely. ROAS dropped sharply, spend fell near zero, and the brand was running on organic traffic alone. On every high-volume category search, "massage ball," "therapy ball," "massage ball for back," Tune Up Fitness was invisible. Competitors owned those results. Any shopper discovering the category for the first time had no path to this brand.

Before vs After

Before Astra (2 Years Self-Managed)

Campaign structure: 3 branded campaigns
Keyword reach: Brand name searches only
Category presence: None
Ad orders: Baseline
Status: Stalled by late 2023

With Astra (22 Months)

Campaign structure: 23 targeted portfolios
Keyword reach: Category, branded, ranking, video
Category presence: Ranking for 'massage ball,' 'therapy ball'
Ad orders: 6x the prior 2-year total
Status: Compounding at elevated baseline

Why Escaping It Is Harder Than It Looks

Most sellers know they should advertise on category keywords. Almost all have tried. The failure pattern is consistent: category keywords cost more per click than brand terms, conversion rates are lower because the shopper does not know the brand, and after 30 days of mediocre performance the spend gets cut. The seller retreats to branded and concludes category keywords are not worth it.

The problem is not the keywords. It is the management method. Making category keyword campaigns profitable requires continuous optimization: testing hundreds of search terms, migrating winners to exact match, cutting losers daily, adjusting bids as competition shifts seasonally. No human process can sustain this at scale across a multi-SKU catalog. The branded trap is not a keyword problem. It is an infrastructure problem.

How Astra Rebuilt It

Astra launched in June 2024. First month: 8.1x ROAS as the system began learning which category keywords convert for this catalog. By November 2024, five months in, total sales had climbed 74% above the pre-Astra baseline. The build phase was complete.

Full campaign architecture from scratch

Three branded campaigns became 23 distinct portfolios organized by strategy type:

Before Astra

Tune Up Fitness Branded 1

Tune Up Fitness Branded 2

Tune Up Fitness Branded 3

With Astra: 23 Portfolios

OG Therapy Balls · Category CAT

Alpha Twin · Category CAT

Coregeous · Category CAT

TB Plus, Therapy Stretch Strap · Category CAT

OG, Alpha, Coregeous · Brand Defense DEF

Ranking · "massage ball," "massage ball for back" RANK

Storefront Video Campaigns VIDEO

Brand defense portfolios run at 12 to 18x ROAS, protecting branded traffic from competitor poaching at minimal cost. Ranking campaigns push organic visibility on the highest-volume category terms. Each portfolio has a distinct job.

Keyword migration at scale

Every keyword starts in broad or auto-discovery campaigns. Astra monitors performance, migrates keywords that convert at acceptable cost to exact match for scaling, and cuts the ones that do not. Tune Up Fitness now ranks for terms it was previously invisible on:

massage balltherapy ballmassage ball for backfeet massage ballTune Up Fitness OG BallCoregeous BallAlpha Twin

Purple: category keywords. These shoppers have never heard of Tune Up Fitness. They are searching for what the brand sells, not who sells it.

Scale spend without degrading quality

Ad spend scaled 5x above pre-Astra levels. Normally scaling this fast tanks ROAS as the new spend reaches lower-quality audiences. Here it did not:

0.33% to 0.59%CTR nearly doubled despite targeting 16x more keywords

Flat CPCAmazon rewarded the relevance signals with favorable auction placement

23.3%Conversion rate held flat at scale, quality did not degrade

The Results

The headline metric is the one most sellers say is not possible: the same ROAS on competitive category keywords as on branded searches, where the shopper was already looking for them by name.

5.3xLegacy ROAS
Branded only

Same ROAS. Completely different keywords.

The legacy 5.3x came from shoppers already searching for the brand name. The Astra 5.3x includes competitive category terms like "massage ball" where the shopper has never heard of Tune Up Fitness. Matching branded efficiency on category keywords is what makes the revenue growth sustainable.

5.3xAstra ROAS
Category + branded mix

+105%Total revenue above pre-Astra baseline

6xMore ad orders than prior 2 years of self-managed PPC combined

+26%Q4 2025 total sales vs Q4 2024, compounding year over year

Metric	Pre-Astra	With Astra	Change
ROAS	5.3x (branded only)	5.3x (category + branded mix)	Matched on harder keywords
Campaign Portfolios	3 branded	23 targeted portfolios	Full rebuild
CTR	0.33%	0.59%	+79%, nearly doubled
Conversion Rate	24.5%	23.3%	Held flat at scale
Ad Orders	Baseline (2 years)	6x that baseline	New-customer acquisition
Brand Defense ROAS	None	12 to 18x	Branded traffic protected cheaply
Organic Revenue Share	~100% (ads stalled)	59% of total revenue	Organic grew ~20% during Astra period
Q4 Year over Year	Q4 2024 baseline	+26% in Q4 2025	Compounding seasonal momentum

Organic revenue grew roughly 20% during the Astra period. Scaling paid 5x did not cannibalize organic. Category keyword ads drive ranking signals that lift organic visibility on the same terms. They compound each other.

"Sellrbox is gold standard. They are responsive and understand all the nuances of the major etailer platforms to help ensure success and grow your brand."

Robert

Tune Up Fitness

Trustpilot ★ 5.0

The Takeaway

If your top-performing Amazon keywords are all your own brand name, your ads are not acquiring customers. They are serving existing ones. The ROAS looks fine because you are showing up to people who were already looking for you. The branded trap is invisible from inside it.

Tune Up Fitness spent two years in it. The fix was not bidding harder on category keywords. It was building the infrastructure that makes them profitable at scale: 23 portfolios, a continuous keyword migration system, and daily optimization no manual process can sustain. The result was the same ROAS on "massage ball" as on their own brand name. Revenue doubled. And every month it compounds further.

How SpitJack Proved That "Fine" Was Leaving $9,100 a Year on the Table

The Branded Keyword Trap

Before vs After

Why Escaping It Is Harder Than It Looks