Do not turn on DSP until branded search is actually defended.
By gabriel-caceros-528077bb · September 9, 2026 · Curated by George's Blog
Do not turn on DSP until branded search is actually defended.
Not "has a campaign."
Defended.
A home brand was spending $22K a month in DSP because the mix looked full-funnel.
Branded search impression share was 54%.
Competitors were sitting on the brand name while DSP chased lookalikes around the open web.
They were paying to introduce the brand, then leaving the front door open.
DSP cannot fix that.
It can only make the leak more expensive.
We sequenced it:
Exact brand terms always on, with a floor share of 80%
Sponsored Brands headline on the brand plus category term
Product targeting on our own hero ASINs
Only then DSP, with purchasers from the last 30 days excluded
DSP frequency capped at 4, because awareness was not the shortage
30 days after the sequence:
→ Branded impression share 81%
→ Competitor presence on brand terms down
→ DSP new-to-brand quality improved because returning searchers were not the whole audience
→ Total media spend down 7%
→ Contribution up $13K
DSP spend did not need to rise to look "full funnel."
It needed a cleaner door to send people through.
We also found $4,100 a month in DSP hitting households that had searched the brand in the last 7 days and then seen a competitor on the brand term.
That is paying for awareness you already owned, then losing the click at search.
The funnel slide is a pretty order of operations.
The real order is: protect the name, then rent attention you do not already own.
If branded search is weak, DSP is a billboard for a store with a competitor in the doorway.
Fix the doorway first.
#amazon #amazonads #amazonppc #amazonseller #ctr