Rule of thumb from a live PPC audit: auto-targeting campaigns should receive roughly 10% of ad budget, used purely for keyword harvesting (discovering new search terms via Amazon Search Term Report) rather than as a primary sales driver. The rest of budget should go to manual exact/phrase-match campaigns built from vetted, winnable keywords (via Cerebro Filtering Protocol (Organic / Competitor Rank / Number of Competitors)) plus Manual Product Targeting (Category vs. Individual ASIN). The diagnosis: an agency running mostly on auto was burning spend on head terms it surfaced that were dominated by category giants (see "Right to Win" Keyword Selection Framework) and on non-incremental branded-keyword clicks (see Branded Keyword Ad Spend Non-Incrementality).
A five-year-old candy brand plateaued at ~$5K/month partly because its PPC spend sat in low-ROI buckets: broad auto and its own branded keyword. The fix capped auto spend at roughly 10% of budget, used purely for keyword harvesting, and shifted the rest into manual keyword and product targeting selected via Helium 10 Cerebro and Switchers Tactic (Product Targeting by Competitor Star-Rating Gap). See "Right to Win" Keyword Selection Framework for how the manual keyword list was chosen.
Broad match campaigns serve the same keyword-discovery and organic-ranking-research function as auto campaigns — neither is meant to be the account's long-term profit driver. Left un-optimized, both get expensive fast, which is why winning terms need to be proactively harvested out into exact match — see Search Term Harvesting into Exact Match.
Don't under-fund an auto campaign meant for keyword discovery — budget at least £10/day so the campaign gathers a reasonable amount of search-term data within the first day or two, rather than starving it and getting too few clicks to act on.
Из тем: PPC Campaign Structure & Bidding