After reverse-ASIN keyword research in Helium 10 Cerebro (roughly 200–300 filtered keywords), the results are split into four separate lists rather than dumped into one pile:
Splitting into four lists lets each category be tracked independently in PPC testing, so performance of e.g. misspelled-keyword traffic can be isolated from main-keyword traffic.
Rather than guessing typos manually, ChatGPT is prompted to generate common Amazon-search misspellings of the master keywords, producing the dedicated misspelled-keyword list.
Each of the four lists is built as its own Helium 10 "folder" populated from Cerebro Filtering Protocol (Organic / Competitor Rank / Number of Competitors) output (via Reverse ASIN Lookup (Helium 10 Cerebro)) plus ChatGPT-Assisted Misspelling Generation for the misspelled bucket. Keeping the lists separate, rather than one undifferentiated dump, lets Main/Related/Misspelled be tested and bid on independently in PPC — the analogy given is a food truck that logs timing, best-sellers, and promo effects versus one that just serves whoever walks up. The combined master list then feeds Helium 10 Listing Builder, Helium 10 Index Checker, and Helium 10 Keyword Tracker.
The four-list split (main, related, misspelled, master) exists specifically so that different match types and bidding strategies later produce cleanly separated, analyzable PPC data — not as arbitrary organization.
The master list is built by combining two Helium 10 tools rather than starting from scratch: Helium 10 Cerebro reverse-engineers the keywords top-performing competitor ASINs already rank for, and Helium 10 Magnet (Seed Keyword Expansion) expands outward from a seed keyword to catch adjacent terms competitors' listings may not surface. "Instead of having to start from scratch, we're going to steal all of the top performing keywords from the top sellers." The combined master list is then split into the main, relevant, and misspelled sub-lists using the prioritization criteria below, rather than being used as one undifferentiated bucket.
Each of the four lists (Master, Main, Related, Misspelled) is kept separate specifically so it can be targeted and tested independently in its own campaign. Mixing them into one undifferentiated keyword dump — dumping every keyword a tool spits out into a single campaign and hoping Amazon sorts it out — is the "backwards" approach that makes it impossible to see which segment is actually working. See Automatic Campaign Four-Way Split (Close Match, Loose Match, Substitutes, Complements) for how this segmentation carries into campaign structure.
The 'Master' list is the deduplicated output of the Cerebro Filtering Protocol (Organic / Competitor Rank / Number of Competitors) pull (plus Keyword List Seed-and-Steal Assembly (Dedupe & Manual Verification) seed-expansion results). Sellers pass through the master list multiple times, each pass tagging a different subset into its own saved Helium 10 folder: highly-relevant terms into Main, broader/adjacent terms into Related, and natural typos into Misspelled (ChatGPT-Assisted Misspelling Generation). Each list then becomes its own PPC campaign so match-type and relevance data stay clean per campaign.