How the Logic Works

The template splits your search term history into two periods: recent (default 7 days) and historical (everything before that). It identifies search terms with at least one order in the recent period AND zero orders historically, representing genuinely new converting queries. For terms with 3+ recent orders, it validates ACOS is within 150% of target to ensure quality. Results are categorized by order volume (Emerging/Solid/Strong) to help prioritize which new winners warrant immediate keyword addition versus continued monitoring.

Key Technical Concepts

True New Winner Logic

The critical distinction is zero historical orders, not just "no recent orders before this period." A search term that sold 20 times last year but had a dry spell isn't a new winner - it's an established term that just converted again. By requiring historical orders = 0, the template only surfaces search terms that represent genuinely new traffic patterns, seasonal shifts, or emerging market trends worth investigating.

Time Period Boundary Management

The recent and historical periods must not overlap or gap. With default settings (recent: 0d..7d, historical: 8d..), day 7 is the most recent full day in "recent," and day 8 starts "historical." This creates a clean split of your entire search term history. If you change recent to 0d..14d, you must change historical to 15d.. to maintain the boundary. Overlapping periods would incorrectly classify terms, while gaps would miss data entirely.

ACOS Validation Strategy

For terms with 1-2 orders, ACOS is highly variable and unreliable - a single high-AOV order can show 15% ACOS while a single low-AOV order shows 60% ACOS from the same underlying profitability. The template skips ACOS validation below 3 orders (configurable) to avoid rejecting potentially good terms due to small sample noise. Above this threshold, ACOS must be within 150% of target to qualify, ensuring you don't expand into consistently unprofitable queries.

Signal Strength Categorization

Strong New Winner (5+ orders): High confidence - multiple conversions in short window indicate real demand and likely profitability. Add as exact or phrase match keywords immediately.

Solid New Winner (3-4 orders): Medium confidence - meaningful conversion signal but less certain. Good candidates for exact match testing or continued auto campaign exposure.

Emerging New Winner (1-2 orders): Lower confidence - could be random variance or genuine new trend. Monitor for additional conversions before keyword addition, or add as broad match if very relevant to product.

Configuration Deep Dive

Time Window Strategy

Shorter recent periods (3-5 days) catch trends faster but surface more noise - you'll see many single-order terms that may not repeat. Longer periods (14-21 days) provide more reliable signals but you discover opportunities later, potentially missing early trend adoption advantages. The 7-day default balances these tradeoffs for most accounts. High-velocity businesses (daily deal sites, trending products) benefit from shorter windows. Stable catalog businesses benefit from longer windows.

ACOS Threshold Interaction with Business Model

The 150% threshold creates different selection pressures based on your target ACOS. With aggressive 20% target ACOS, 150% threshold = 30% acceptable - you're only seeing very efficient new terms. With conservative 40% target ACOS, 150% threshold = 60% acceptable - you'll discover more terms but need to validate their long-term viability. Match threshold tightness to your profit margins and willingness to test marginal terms.

Minimum Orders for ACOS Interaction

This setting determines when ACOS filtering activates. At 3 orders (default), terms with 1-2 orders bypass ACOS check entirely - you see all low-volume new winners regardless of efficiency. Lowering to 2 orders means ACOS filtering starts at 2-order terms, reducing false positives from lucky single high-margin orders. Raising to 5 orders means you see all new winners with 1-4 orders, only filtering those with 5+ orders - useful when you want to manually evaluate all emerging trends.

Historical Period Definition

Using "8d.." (day 8 through earliest available data) means the template looks at your entire search term history before the recent window. This is intentional - even if a term sold once 18 months ago, it's not a "new" winner today. If you only want to flag terms that are new within a specific lookback (e.g., terms that didn't sell in last 90 days), you could adjust to "8d..97d" but this misses the core use case of finding genuinely novel search patterns.

Diagnostic Property Interpretation

Understanding $result Categories

Strong New Winner (5+ orders) - Highest priority for keyword addition. Multiple conversions in recent window represent reliable demand signal. Action: Add as exact match keywords immediately in manual campaigns. Consider phrase match if search term shows clear intent variations (e.g., "red running shoes" could become phrase "red running shoes").

Solid New Winner (3-4 orders) - Good candidates for testing. Meaningful conversion signal but not yet proven at scale. Action: Add as exact match in dedicated testing campaign with controlled budget, or continue monitoring in auto campaigns for one more period. If ACOS is excellent (<50% of target), promote to immediate keyword addition.

Emerging New Winner (1-2 orders) - Early signals requiring validation. Could be random variance or genuine new trend. Action: Monitor for one more review period. If term appears again with additional orders, promote to keyword addition. If highly relevant to product (exact product name variation, clear purchase intent), consider immediate broad match addition for exposure control.

New Sales - Poor ACOS - Converting but unprofitable. These passed the new winner test but failed ACOS validation. Action: Review for relevance - if search term is off-target (wrong product category, informational intent), add as negative keyword. If relevant but expensive, monitor to see if efficiency improves with scale, or accept that this query isn't profitable for your product.

Common $reason Patterns

"New winner identified" - Qualified term meeting all criteria. Most common scenario during active discovery periods. Indicates healthy auto campaign performance generating new converting traffic.

"Already has historical sales" - Term converted in recent period but has prior order history. This is normal behavior for established terms and not a concern - these are filtered out to focus your attention on genuinely new opportunities.

"Recent ACOS too high" - Term is converting for first time but inefficiently. Evaluate relevance: if term is off-target, negate it. If relevant, the product may not be competitive for this query (check organic rank, competitor pricing, review quality).

"Insufficient recent orders" - Very rare with default 1-order minimum. Only appears if you've raised the minimum recent orders setting above 1. Indicates you have strict discovery criteria and are waiting for stronger signals.

Common Performance Patterns

Seasonal Discovery Cycles

Expect higher volumes of new winners at season starts (back-to-school in July, holiday in October, summer in April). These represent seasonal search pattern shifts as customers begin researching category needs. Process these batches systematically - don't add all terms at once. Prioritize Strong/Solid categories first, monitor Emerging terms for follow-up conversions before committing budget.

Product Launch Discovery

New product listings typically show 10-30 new converting search terms in first 2-4 weeks as Amazon's algorithm tests different placements and customers find various search paths to your product. This is healthy and expected. Focus on terms that appear in Strong/Solid categories - these represent your product's core search relevance. Emerging terms may be exploratory placements that won't sustain.

Competitive Displacement Patterns

Sudden clusters of new winners on established products may indicate competitor stock-outs or price increases driving their traffic to your listings. These opportunities can be temporary - if you see 5-10 new winners in a single week on a mature product, check top competitors' availability and pricing. Add the most relevant terms quickly to capture shifted demand before competitors return to market.

Auto Campaign Optimization

Consistent weekly discovery of 2-5 new winners indicates healthy auto campaign performance. Zero new winners for multiple consecutive weeks suggests: auto campaign budgets too constrained (not getting enough exposure for discovery), bids too low (not winning auctions on new terms), or market saturation (you've already captured all relevant search terms for your niche).

Review Cadence Impact

Daily reviews will show many Emerging (1-2 order) terms that may not repeat. Weekly reviews show fewer total terms but higher proportion of Solid/Strong signals. Monthly reviews miss fast-moving trends but show only the most sustained new traffic patterns. Match review frequency to your business velocity and available time for keyword management.

Expected Volume by Account Size

Small accounts ($5-10k/month spend): 2-5 new winners per week is healthy. More suggests very broad targeting, fewer suggests over-reliance on manual campaigns.

Medium accounts ($10-30k/month spend): 5-15 new winners per week typical. Should include mix of Strong/Solid/Emerging.

Large accounts ($30k+/month spend): 15-30+ new winners per week common, especially with diverse product catalogs. Requires systematic processing workflow to avoid overwhelming keyword management.

‍