Purpose

Scales bid adjustments proportionally based on distance from target ACOS using adaptive timing controls.

This Blueprint is for:

✓ Accounts wanting intelligent bid optimization that adjusts more aggressively when performance is far from target and more cautiously when close
✓ Keywords and targets with varying data reliability where some have daily orders while others need weeks to generate meaningful signals
✓ Sellers who want proportional responses rather than binary "increase or decrease" logic that treats all performance gaps the same
✓ Accounts ready for sophisticated automation that adapts cooldown periods based on data freshness and volume

Key Features

• Proportional scaling system adjusts bid changes from minimum to maximum based on performance gap magnitude, not fixed percentages across all scenarios
• Adaptive cooldown periods automatically extend wait times when using older data (2 days for fresh weekly data, up to 7 days for extended 60-day lookbacks)
• Four-tier data cascade evaluates 7-day, 14-day, 30-day, and 60-day periods to find the freshest reliable data for each keyword
• Buffer zone protection creates a no-change zone around target ACOS to prevent constant micro-adjustments on borderline performers

Configuration Guidance

• Buffer and Threshold Relationship: The buffer (tacosbufferminchange=10t_acos_buffer_min_change = 10%) creates a no-change zone. Outside this buffer, adjustments scale linearly up to the thresholds (ta​cosb​ufferm​inc​hange=10t_acos_threshold_max_increase = 50%, $t_acos_threshold_max_decrease = 200%). Tighten the buffer to 5% for more frequent adjustments, or widen to 15% for more stability.
• Data Reliability Standards: The $orders_min = 2 setting determines when ACOS is considered reliable. Increase to 3-4 for higher confidence requirements (fewer but more certain adjustments), or keep at 2 for maximum coverage on lower-volume keywords.
• Adjustment Range Tuning: Default 3-10% range ($bid_min_change to $bid_max_change) is conservative. Increase maximum to 15% if you're comfortable with more aggressive changes, or reduce minimum to 2% if you want gentler adjustments on borderline cases.
• CPC-Based Limits: The $bid_limit_cpc_multiplier = 2.0 prevents bids from exceeding 2× your average CPC. This is a crucial safeguard - don't disable it. If you're hitting this limit frequently and performance justifies higher bids, gradually increase to 2.5 or 3.0 rather than removing the limit entirely.

Recommended Workflow

Daily | Automation: Change Bid: Set ($) using $target_bid

Dynamic Bid Management - Technical Details

How the Algorithm Works

The blueprint evaluates performance across four cascading time periods to find the freshest data with sufficient orders, then calculates how far current ACOS is from target ACOS. This distance determines both the magnitude of bid adjustment (closer = smaller changes, farther = larger changes) and the cooldown period before the next change (fresher data = shorter cooldowns). The proportional scaling system creates smooth, graduated responses rather than binary reactions, while buffer zones prevent adjustment churn on keywords performing near target.

Key Technical Concepts

Proportional Performance Gap Calculation
The blueprint doesn't treat a keyword at 15% ACOS the same as one at 25% when target is 30%. It calculates exactly how far performance is from the relevant threshold, then scales the adjustment percentage proportionally within your configured min/max range. A keyword barely outside the buffer zone gets close to $bid_min_change (3%), while a keyword at or beyond the max threshold gets the full $bid_max_change (10%). This creates nuanced responses that match the severity of the performance gap.

Four-Tier Data Cascade with Adaptive Cooldowns
The blueprint evaluates four periods in order: 7-day → 14-day → 30-day → 60-day, stopping at the first period with sufficient orders ($orders_min = 2). But it doesn't just use different data - it also adapts the cooldown period based on which tier is used. Fresh 7-day data gets a 2-day cooldown, while old 60-day data gets a 7-day cooldown. This prevents over-optimization when forced to use stale data due to low order volume.

Dual-Threshold Asymmetric System
The blueprint uses different threshold distances for increases vs decreases:

• Max increase at ACOS ≤ 50% of target (e.g., 15% when target is 30%)
• Max decrease at ACOS ≥ 200% of target (e.g., 60% when target is 30%)

This asymmetry reflects different risk profiles - being very efficient (50% of target) justifies aggressive scaling, while being very inefficient (200% of target) requires aggressive cuts. The buffer zone (90-110% of target by default) sits in the middle, creating a stability zone where no changes occur.

Buffer Zone vs Threshold Logic
Understanding the three zones is critical:

1. Below buffer (ACOS < 90% of target): Linear scaling from min to max increase based on distance to 50% threshold
2. Within buffer (90-110% of target): Zero adjustment - performance is acceptable
3. Above buffer (ACOS > 110% of target): Linear scaling from min to max decrease based on distance to 200% threshold

This creates five distinct response regions: max increase, scaled increase, no change, scaled decrease, max decrease.

Configuration Deep Dive

Buffer and Threshold Interaction
The buffer (tacosbufferminchange)andthresholds(t_acos_buffer_min_change) and thresholds (ta​cosb​ufferm​inc​hange)andthresholds(t_acos_threshold_max_increase, $t_acos_threshold_max_decrease) work together to create the adjustment curve. If you widen the buffer to 15%, you're creating a 85-115% no-change zone, which means adjustments only start when performance is more significantly different from target. This is paired with threshold distances - if your max increase threshold is 50%, the scaling region runs from 85% down to 50% of target ACOS. Narrower buffers + distant thresholds = more gradual scaling. Wider buffers + close thresholds = more aggressive scaling once outside the buffer.

Data Reliability vs Coverage Tradeoff
Setting $orders_min higher (3-4) means you'll rely on longer lookback periods more often, which increases cooldowns and reduces adjustment frequency but improves confidence in each decision. Setting it lower (2 or even 1) means you'll use fresher data more often, enabling more frequent adjustments but with higher risk of reacting to statistical noise. For accounts with high order volume, consider $orders_min = 3-4. For lower volume accounts where most keywords rarely hit 2 orders in 7 days, keeping it at 2 maximizes coverage.

Adjustment Range Philosophy
The min/max range creates bounds for the proportional scaling. A 3-10% range means:

• Barely outside buffer = ~3% change
• Halfway between buffer and threshold = ~6.5% change
• At or beyond threshold = 10% change

Narrower ranges (2-7%) create more predictable, conservative automation. Wider ranges (5-15%) create more dramatic responses to performance extremes. Avoid ranges where min is too close to max (like 8-10%) as this eliminates the benefits of proportional scaling.

CPC Limit Period Selection
The $bid_limit_cpc_period = 90d uses a long period to smooth out CPC volatility and establish a stable benchmark. Shortening this to 30-60 days makes the limit more responsive to recent CPC changes (useful if your category sees seasonal CPC shifts), but also makes it more volatile. Lengthening to 180+ days creates extreme stability but may lag behind market changes. The 90-day default balances stability with reasonable responsiveness.

Diagnostic Property Interpretation

$data_quality Values:

• "High Volume: Short Lookback": Using 7-day data with 2+ orders - most confident scenario, 2-day cooldown active
• "Med Volume: Medium Lookback": Using 14-day data with 2+ orders - good confidence, 3-day cooldown active
• "Low Volume: Long Lookback": Using 30-day data with 2+ orders - moderate confidence, 5-day cooldown active
• "Very Low Volume: Extended Lookback": Using 60-day data with 2+ orders - lowest confidence, 7-day cooldown active
• "insufficient": No period has enough orders - no action taken

These tags appear in the $reason diagnostic to show both why action is/isn't being taken AND which data quality tier is driving decisions.

$reason Patterns:

• "[Data Quality Tag] Within target buffer zone": ACOS is in the acceptable 90-110% range, no adjustment needed
• "[Data Quality Tag] Recently changed - cooldown active": Cooldown period for that data quality tier hasn't elapsed yet
• "[Data Quality Tag] ACOS below/above target threshold": Performance is outside buffer, bid adjustment calculated
• "Limited by CPC-based maximum": Calculated bid exceeds average CPC × 2.0, capped at limit
• "No target ACOS set": Can't evaluate without target benchmark
• "Insufficient order data across all lookback periods": Even 60-day period doesn't have 2 orders

$result Magnitude Categories:

• "Bid Increase/Decrease: Max": At or beyond threshold, getting full $bid_max_change (10%)
• "Bid Increase/Decrease: Upper Range": 50-99% of max change (5-9.9%)
• "Bid Increase/Decrease: Lower Range": Just above min change up to 50% of max (3-5%)

Common Performance Patterns

High-Volume Keywords with Short Lookback
Keywords consistently showing "High Volume: Short Lookback" are your most actively managed terms. They generate enough orders weekly to use fresh 7-day data with 2-day cooldowns, meaning they can be adjusted multiple times per week if performance shifts. These should show the most responsive bid optimization and are where the proportional scaling system shines most - small performance deviations get small adjustments, large deviations get large adjustments, all happening quickly.

Low-Volume Keywords Stuck on Extended Lookback
Keywords consistently showing "Very Low Volume: Extended Lookback" generate fewer than 2 orders in 30 days, forcing the system to use 60-day data with 7-day cooldowns. These will adjust infrequently (maximum once per week), and when they do adjust, the changes are based on very old data. This is working as designed - low-order keywords should change cautiously. If you're seeing too many keywords stuck here and want more frequent adjustments, lower $orders_min to 1 (but expect more noise).

Buffer Zone Stability vs Constant Adjustments
Keywords bouncing between "Within target buffer zone" and "ACOS below/above target threshold" indicate you're near the buffer boundaries. If you're seeing excessive adjustment churn (adjust up, then down, then up again), widen the buffer to 15% to create more stability. If you're seeing too many keywords sitting in "Within target buffer zone" while you think they should be adjusted, narrow the buffer to 5% to reduce the no-change zone.

CPC Limit Constraints
If many keywords show "Limited by CPC-based maximum" frequently, your bid calculations are trying to exceed 2× average CPC regularly. This suggests either: (1) Your ACOS targets are very aggressive relative to current CPC reality, requiring unrealistic bids to achieve them, or (2) The multiplier is too conservative for your category. Review actual performance of keywords hitting this limit - if they're delivering good ACOS even at the capped bid, consider increasing the multiplier to 2.5 or 3.0.

Cooldown Blocking High-Opportunity Keywords
Keywords showing "[High Volume: Short Lookback] Recently changed - cooldown active" with excellent ACOS (far below target) are in a scenario where you'd like to increase bids more aggressively but the adaptive cooldown is holding you back. This is intentional safety - even with fresh data, some cooldown prevents over-reaction to short-term spikes. However, if you're consistently seeing opportunities blocked by cooldowns, you can shorten the tier-specific cooldown periods ($cooldown_short, $cooldown_medium, etc.) to enable more frequent adjustments.