Here is your negative keyword mining analysis. Treat these recommendations as directional guidance based on e-commerce and subscription industry patterns. Review them against your specific account setup before deploying, as negative keywords directly impact your traffic and spend. ## STEP 1: HEALTH CHECK Based on your inputs (D2C Coffee Subscription, active for months, only 10-50 current negatives): 1. **Negative Keyword Maturity Score: 3/10** * **Coverage Breadth:** Low. An account active for months in the highly competitive coffee space should have hundreds of negatives across universal, competitor, and hardware/equipment themes. * **Match Type Discipline:** Likely underutilized. With only 10-50 negatives, you are almost certainly relying too heavily on broad match negatives or missing exact/phrase variants. * **Maintenance Cadence:** Inconsistent. The presence of basic junk like "wiki," "jobs," and "free" in a seasoned account indicates a lack of weekly mining. 2. **Top 3 Risk Patterns for D2C Coffee Subscriptions:** * **Hardware Seekers:** Paying for clicks from people looking for grinders, espresso machines, or Keurig parts when you only sell beans. * **Grocery/Discount Intent:** Bleeding spend on queries for Folgers, cheap pods, or supermarket coupons. Your $38 AOV indicates a premium product; these searchers will bounce. * **Wholesale/B2B:** Cafes or offices looking for bulk 5lb bags or wholesale pricing, which dilutes your D2C metrics. 3. **Special Note for AI Max Users:** If you are running AI Max for Search, query expansion is aggressive, especially in the first 2-4 weeks. Negative keyword review must be daily during this window, not weekly. 4. **Verdict:** **Foundation gap.** You need to build a robust base list immediately to stop the bleed, then establish a weekly maintenance cadence. ## STEP 2: UNIVERSAL NEGATIVE LIST **CRITICAL CAVEAT: Negative keywords do NOT match close variants — singular/plural, misspellings, and synonyms must be added separately.** Below is your copy-paste universal negative list. === Universal Negatives - Account-Level Recommended === jobs job salary hiring career careers intern internship vacancy recruiter resume free "free download" "free sample" "free samples" diy "how to" tutorial template sample samples wiki definition "what is" vs comparison examples reddit quora forum complaint complaints scam lawsuit porn sex nude naked [is google ads free] [what is ritual roast] ## STEP 3: INDUSTRY-SPECIFIC NEGATIVE LIST For a D2C Coffee Subscription ($38 AOV), you must filter out B2B, hardware, cheap alternatives, and cafe-seekers. === E-commerce / Coffee Subscription Negatives === wholesale *(Why: Filters out B2B/cafes wanting bulk discounts)* bulk *(Why: Similar to wholesale, low D2C intent)* manufacturer *(Why: B2B/supply chain intent)* distributor *(Why: B2B intent)* b2b *(Why: B2B intent)* used *(Why: Used equipment seekers)* refurbished *(Why: Looking for hardware, not beans)* cheap *(Why: Protects your premium $38 AOV brand)* discount *(Why: Unless you actively promote discounts, this attracts bargain hunters)* coupon *(Why: Prevents coupon-site hunters from draining ad budget)* dropship *(Why: B2B/entrepreneur intent)* grinder *(Why: Equipment, not subscription beans)* grinders *(Why: Plural variant of equipment)* machine *(Why: Equipment)* maker *(Why: Equipment, e.g., "coffee maker")* keurig *(Why: Pod ecosystem, not fresh roasted beans)* kcup *(Why: Pod ecosystem)* "k cup" *(Why: Pod ecosystem variant)* nespresso *(Why: Pod ecosystem)* pods *(Why: Pod ecosystem)* "french press" *(Why: Ambiguous, but usually hardware intent unless paired with "coffee for")* recipe *(Why: DIY/at-home intent without purchase intent)* recipes *(Why: Plural variant)* "coffee shop" *(Why: Local cafe intent, not D2C delivery)* "near me" *(Why: Local cafe intent)* amazon *(Why: Brand protection, users looking for Prime shipping)* walmart *(Why: Grocery/cheap intent)* "trader joe" *(Why: Grocery intent)* "trader joes" *(Why: Grocery intent variant)* ## STEP 4: SEARCH TERM ANALYSIS Here is the classification of your provided search terms. | Search Term | Classification | Action | Match Type to Apply | | :--- | :--- | :--- | :--- | | free coffee samples | NEGATIVE | Add as negative | "free coffee samples" | | how to make pour over | NEGATIVE | Add as negative | "how to make" | | nespresso pods cheap | NEGATIVE | Add as negative | nespresso | | coffee jobs near me | NEGATIVE | Add as negative | jobs | | wholesale coffee beans | NEGATIVE | Add as negative | wholesale | | starbucks vs dunkin | IRRELEVANT | Add as broad negative | dunkin | | diy cold brew recipe | NEGATIVE | Add as negative | recipe | | ritualroast reddit | NEGATIVE | Add as negative | reddit | | best coffee subscription review | WATCH | Keep monitoring | N/A | | single origin ethiopian beans | WINNER | Add as new keyword | [single origin ethiopian beans] | | caffeine content green tea | IRRELEVANT | Add as broad negative | "green tea" | | coffee grinder under 50 | NEGATIVE | Add as negative | grinder | | specialty coffee dublin | NEGATIVE | Add as negative | dublin | | amazon coffee deals | NEGATIVE | Add as negative | amazon | | trader joe coffee | NEGATIVE | Add as negative | "trader joe" | | ground coffee for french press| WINNER | Add as new keyword | "ground coffee for french press" | | folgers coupon | NEGATIVE | Add as negative | folgers | | keurig refurbished | NEGATIVE | Add as negative | keurig | | peets coffee careers | NEGATIVE | Add as negative | careers | | coffee taste test wiki | IRRELEVANT | Add as broad negative | wiki | *Note: For terms like "specialty coffee dublin", I flagged `dublin` as a negative assuming you target the US. If you ship to Ireland, change this to a WATCH or WINNER.* ## STEP 5: NEGATIVE KEYWORD LIST ARCHITECTURE Implementing a 3-tier structure will keep your account clean and scalable in 2026. | List Name | Tier | Scope / Theme | Apply To | Maintenance Cadence | | :--- | :--- | :--- | :--- | :--- | | ACCT_Universal_Junk | Tier 1 (Account-Level) | Job seekers, research, Reddit, free/DIY | All Campaigns (Auto) | Quarterly Review | | SHARED_B2B_Exclude | Tier 2 (Shared List) | Wholesale, bulk, distributors | All D2C Campaigns | Monthly | | SHARED_Equip_Exclude | Tier 2 (Shared List) | Grinders, machines, Keurig, Nespresso | Bean-only Campaigns | Monthly | | SHARED_Grocery_Brands| Tier 2 (Shared List) | Folgers, Peets, Walmart, Trader Joes | Non-Brand Campaigns | Monthly | | CAMP_[Name]_Negatives | Tier 3 (Campaign) | Highly specific exclusions (e.g., specific flavors you don't carry) | Individual Campaigns | Weekly | *Remember: The Tier 1 Account-Level list holds up to 1,000 keywords and applies automatically to PMax, Search, and Shopping. PMax campaign-level negatives now hold up to 10,000 entries for Search/Shopping inventory.* ## STEP 6: WEEKLY MINING WORKFLOW **15-Minute Weekly Review Checklist** * **Step A: Pull Search Terms report** (Last 7 days, all active campaigns). * *WARNING:* Google hides 30-50% of search terms under "Other." Cross-check your campaign spend total vs. the sum of visible search term spend to understand your hidden data gap. * **Step B: Sort by spend** (Highest to lowest). Focus your time on the top 50 terms where your money is actually going. * **Step C: Classify.** * WINNER -> Add as exact/phrase keyword. * NEGATIVE -> Add as exact/phrase negative. * IRRELEVANT -> Add as broad negative (single words). * WATCH -> Leave it alone for now. * **Step D: Route to Architecture.** Put universal junk in the Account list, thematic junk in Shared lists, and campaign-specific junk at the Campaign level. * **Step E: Run Conflict Check.** (Tools & Settings → Negative Keyword Conflicts). Ensure your new negatives aren't accidentally blocking `[coffee subscription]`. * **Step F: Document.** Keep a log of shifting trends (e.g., "Seeing a spike in decaf queries, maybe we should launch a decaf line"). *Special AI Max Note:* If you activate AI Max, run this checklist DAILY for the first 14-28 days. ## STEP 7: RED FLAGS TO WATCH Keep an eye out for these 9 patterns that indicate deeper account issues that negative keywords alone cannot fix: 1. **Search Terms report sum < campaign spend by a massive margin:** You have high "Other" / hidden data. You need to rely on N-gram pattern mining or scripts instead of single-term manual mining. 2. **High volume of completely off-topic queries:** Your broad match keywords are too aggressive or lack guardrails. Switch your highest spenders to phrase or exact match. 3. **Repeated single-word queries with high spend:** Your keyword targeting is too loose (e.g., bidding broadly on "coffee"). 4. **Queries from unintended geographies (e.g., "coffee subscription UK"):** This is a geo-targeting settings issue (ensure "Presence" is selected, not "Presence or Interest"). 5. **Wrong customer intent (B2B buyers clicking D2C ads):** This is an audience signal/PMax asset issue. Feed the algorithm better customer match lists. 6. **Brand misspellings in non-brand campaigns:** You are missing an account-level Brand Exclusions list. (This is a specific setting, not just a negative keyword list). 7. **Negative keyword conflicts blocking own keywords:** Run the Conflict Check tool immediately and prune overly aggressive phrase/broad negatives. 8. **AI Max accounts seeing bizarre query themes:** Review the AI Max search themes report. You may need to narrow your initial signal scope. 9. **PMax draining branded queries:** Set up account-level Brand Exclusions (in addition to negatives) so PMax focuses on acquisition, not cannibalizing your branded Search campaigns.