Auto vs Manual Targeting: Complete Transition Guide with Harvest Loop Framework

Introduction

"Should I just leave Auto running?" "What's the exact criteria for moving to Manual?" — these are universal pain points in Amazon advertising operations.

The answer: Auto is your "discovery engine", Manual is your "monetization engine". Connect them with a weekly harvest loop that systematically applies Promote / Hold / Exclude decisions. That's the complete framework.

This article assumes standardized KPI definitions and attribution windows. If not yet established, start with:

• /en/blog/acos-roas-tacos-complete-guide
• /en/blog/attribution-window-analysis
• /en/blog/custom-dashboard-design

---

Table of Contents

1. Executive Summary (Conclusion First)
2. Role Division: Auto=Discovery / Manual=Monetization
3. Transition Thresholds (Keywords/ASINs)
4. Harvest Loop Implementation Template (Weekly Operations)
5. Structure Template: Campaigns/Bids/Negatives
6. Measurement Pitfalls and Dashboard Integration
7. Case Studies (2 Examples)
8. Frequently Asked Questions (FAQ)
9. Checklist and Next Actions

---

Executive Summary (Conclusion First)

• Never pause Auto: Keep it discovering new queries/ASINs continuously
• Promote to Manual: Copy winning search terms/ASINs to Manual and block them in Auto with negatives
• Weekly loop: Run harvest, promotion, and exclusion 1–2 times per week systematically
• Thresholds: Adjust by category, but start with standard thresholds below
• Bidding: Base on #27 reverse-engineering formula, then adjust ±10–15% based on performance

---

Role Division: Auto=Discovery / Manual=Monetization

Role	Purpose	Strengths	Weaknesses	Common Failure
Auto	New query/ASIN discovery	Scalability, low operational overhead	Lower precision, wasted clicks	Neglect → waste accumulation
Manual	Profitability and placement control	Precise bid/match/placement control	Limited discovery	Slow promotion → missed opportunities

Principle: Never pause Auto. Always grow Manual. Related: /en/blog/long-tail-keyword-mining, /en/blog/negative-keyword-design

---

Transition Thresholds (Keywords/ASINs)

Keywords (Search Terms → Keyword Promotion) Guidelines

Start with Exact match (add Phrase/Broad separately if needed)

• Promote (to Manual-Exact)
  • Past 14–30 days: Clicks ≥ 20 AND Orders ≥ 2
  • OR ACoS ≤ Target+5pp AND Spend ≥ 5× Avg CPC
• Hold
  • Clicks 10–19, zero orders: Consider SBV/creative improvements first
• Exclude (Negative)
  • Clicks ≥ 25 AND Orders = 0 (add Exact negative to Auto)
  • Impressions ≥ 1,500 AND CTR ≤ 0.15% (Phrase negative candidate)

ASINs (Product Targeting) Guidelines

• Promote (to Manual-Product Targeting)
  • Clicks ≥ 15 AND Orders ≥ 1, OR ACoS ≤ Target+5pp
• Exclude (Negative ASIN)
  • Clicks ≥ 20 AND Orders = 0, OR ACoS ≥ Target+15pp

Note: Adjust by category/price point. Loosen thresholds for seasonal gifting, tighten for niche high-ticket items. Related: Event response /en/blog/event-period-bid-tracking

---

Harvest Loop Implementation Template (Weekly Operations)

Step A: Data Collection

• Pull Search Term Report (STR) and Product Targeting performance for past 30 days (also use 7-day granularity)

Step B: Classification (4-Box Framework)

1. Promote: Exceeds thresholds → copy to Manual
2. Hold: Close to thresholds → fix creative/pricing/inventory first
3. Exclude: Wasteful → add negatives to Auto/reduce bids
4. Learn: Insufficient data → continue discovery at low bids

Step C: Implementation (Apply Changes)

• Add to Manual-Exact/ASIN campaigns
• Add Exact/Phrase negatives to Auto (prevent cannibalization)
• Set bids: Initialize with #27 formula → adjust ±10–15% after 1 week
• Log changes: Record rationale in action history (#42)

-- Example: Threshold view for harvesting (BigQuery/Redshift)
create or replace view v_str_candidates as
select
  query,
  sum(clicks) as clicks,
  sum(orders) as orders,
  safe_divide(sum(cost), nullif(sum(sales),0)) as acos,
  avg(cpc) as avg_cpc,
  sum(impressions) as imps
from str_last_30d
group by query
qualify (
  (clicks >= 20 and orders >= 2)
  or
  (acos <= :target_acos + 0.05 and sum(cost) >= 5 * avg_cpc)
);

---

Structure Template: Campaigns/Bids/Negatives

Campaign Hierarchy

1. SP Auto (4-way split): Close / Loose / Substitutes / Complements
2. SP Manual-Keyword: Exact core, separate budget for Phrase/Broad if needed
3. SP Manual-Product Targeting: Competitor ASINs / Complementary ASINs

Bidding and Budget

• Initial bid: Reverse-engineer from target ACoS using /en/blog/bid-calculation-target-acos
• Adjustments: ±10–15%/week based on performance, advance threshold pre-events (#31)
• Budget: Prioritize Manual (monetization). Keep Auto sufficiently funded for discovery.

Negative Design (Cannibalization Prevention)

• Add Exact negatives to Auto for promoted terms, cross-negative to Phrase/Broad
• Add Phrase negatives to Auto for low-quality broad terms
• Block ASINs similarly with Negative ASIN settings

Also address CTR→CVR bottlenecks via PDP/image/title improvements: /en/blog/brand-registry-advertising-benefits, /en/blog/brand-registry-advertising-benefits

---

Measurement Pitfalls and Dashboard Integration

• Post-promotion evaluation period: Allow 7–14 days as learning phase, avoid hasty judgments
• Attribution window misalignment: Fix at account level, display permanently on screen (#41)
• TACoS for holistic optimization: Fold SB/SD NTB/view-through into TACoS, don't judge SP alone
• Anomaly detection: Alert on click surges, CVR drops, impression share declines (#42)

---

Case Studies (2 Examples)

Case 1: Kitchen Products (Mid-Price Tier)

• Actions: Bi-weekly harvest loop, promotion to Exact core, cross-negatives to Auto
• Results (6 weeks): ACoS 23% → 17%, sales +21%, wasted clicks -28%
• Lesson: Exact focus + Auto cleanup drives immediate efficiency gains

Case 2: Gift Food (Seasonal)

• Actions: Pre-event threshold loosening → accelerated promotion, price/stock × bid sync (#32)
• Results (8 weeks): CTR +18%, CVR +12%, TACoS -2.1pp
• Lesson: Seasonality is a promotion speed game. Auto neglect = missed opportunities

---

Frequently Asked Questions (FAQ)

Q1. Should I pause Auto and only run Manual? A. No. Discovery stops and growth stagnates. Keep Auto as a narrow discovery channel.

Q2. Is Exact enough for Manual? A. Exact is the monetization core. However, using Phrase/Broad at low bid/budget as "semi-discovery" is effective.

Q3. How to handle high-click, zero-conversion terms? A. Default to negative (Exact). If retrying after creative/PDP improvements, consider Phrase negative only.

Q4. Where to start with ASIN targeting? A. Target competitors' weaknesses (price/reviews/inventory) and complementary products. Use looser promotion thresholds for clicks/orders.

---

Checklist and Next Actions

Pre-Launch Checklist

• Auto (4-way) campaigns active
• Weekly 1–2× harvest loop defined
• Promoted terms blocked in Auto with Exact negatives
• Exact core and ASIN core separated
• Bids initialized with target ACoS reverse formula
• Attribution window/definitions permanently displayed on dashboard

Immediate Action Items

1. Pull 30-day STR and classify using this article's thresholds into Promote/Hold/Exclude.
2. Copy Promote to Manual-Exact/ASIN, implement negatives in Auto.
3. Initialize bids with #27 formula → adjust ±10–15% after 1 week.
4. Record and visualize action history per #42.

---

Related Resources

• /en/blog/acos-roas-tacos-complete-guide
• /en/blog/cpc-vs-bid-amazon-auction
• /en/blog/long-tail-keyword-mining
• /en/blog/negative-keyword-design
• /en/blog/bid-calculation-target-acos
• /en/blog/event-period-bid-tracking
• /en/blog/price-stock-bid-sync
• /en/blog/attribution-window-analysis
• /en/blog/custom-dashboard-design

---

Author: Arctavia Product Team

Related Arctavia resources

Use these supporting pages to compare Amazon PPC operating models and implementation choices.

• tools / amazon-listing-score
• tools
• tools / acos-calculator
• tools / amazon-ai-visibility-score
• pricing
• compare