Skip to content
Picaro.AI Picaro.AI Prompt Library

Customer-Lens Ad Analysis with AMC

Prompts that go beyond ACoS / ROAS to answer “how much is each customer worth?”, “do the customers we acquired through ads come back?”, and “which campaigns are saturating?” using Amazon Marketing Cloud (AMC) data. The lens is customer-centric, cutting across SP / SB / SD / DSP, all the way to ASIN portfolio, campaign judgment, and budget allocation.

Prompt quick reference

Section titled “Prompt quick reference”

Brand-wide snapshot

Section titled “Brand-wide snapshot”
What you want to doPrompt
Calculate LTV / CAC / true ROAS (including brand halo) per customerPrompt 1
Decompose halo contribution (clicked ASINs vs brand-wide)Prompt 2
Monthly brand acquisition health check (NTB rate / RPR / LTV:CAC trend)Prompt 3

ASIN portfolio decisions

Section titled “ASIN portfolio decisions”
What you want to doPrompt
See NTB acquisition strength per ASINPrompt 4
Sort ASINs into “acquisition / retention / sunset” bucketsPrompt 5

Cohort & retention

Section titled “Cohort & retention”
What you want to doPrompt
See monthly cohort retention and LTV growthPrompt 6
Post-purchase retention funnel (M+1 / M+3 / M+6 return rates)Prompt 7

Ad-type comparison

Section titled “Ad-type comparison”
What you want to doPrompt
Compare ad-type (SP / SB / SD / DSP) NTB power and revenuePrompt 8
Budget-shift proposal toward the lowest NTB CAC ad typePrompt 9
Compare M+1 return rate (customer quality) by ad typePrompt 10

Campaign trends & signals

Section titled “Campaign trends & signals”
What you want to doPrompt
Track campaign-level attribution trends month over monthPrompt 11
Early-detect saturation / diminishing-return campaignsPrompt 12
Identify campaigns where customer count and revenue diverge (basket-size shift)Prompt 13

For deeper specs, see Safety / Automation phases / Troubleshooting.

Prompt 1: Customer LTV / CAC / true ROAS

Section titled “Prompt 1: Customer LTV / CAC / true ROAS”

Calculate LTV (lifetime value), CAC (customer acquisition cost), repeat purchase rate, and halo-inclusive ROAS for ad-attributed customers. Returns two figures — one limited to ad-clicked ASINs, and one including the customer’s brand-wide purchases (halo effect).

Using AMC data for the past {{PERIOD}}, calculate
LTV / CAC / repeat purchase rate / true ROAS (including halo)
for ad-attributed customers.
Show both the figures for ad-clicked ASINs only (no halo) and brand-wide (with halo)
side by side, and judge whether the LTV : CAC ratio is healthy.

Placeholders: {{PERIOD}} (e.g. last 90 days, 2026-Q1. 30+ days recommended for trend reliability).

Returns: Customer count × total revenue × promoted-ASIN revenue × halo-inclusive revenue × CAC × LTV (with / without halo) × RPR × ROAS × LTV:CAC ratio. Includes a health judgment against the typical 3× benchmark.

What the AI will do

  1. Aggregate AMC data over {{PERIOD}} and extract ad-touched customers
  2. Compute per-customer revenue under “promoted ASINs only” and “brand-wide (with halo)”
  3. Divide same-period ad spend by customer count for CAC
  4. Calculate LTV ÷ CAC and flag health (3× or higher is healthy)
  5. Segments with fewer than 100 unique customers are excluded per AMC privacy rules

Interaction pattern: Immediate display (read-only, no ad-setting changes)

Prompt 2: Decompose halo effect contribution

Section titled “Prompt 2: Decompose halo effect contribution”

Surface how much of revenue / LTV comes from ad-clicked ASINs vs the halo (other brand ASINs the same customer bought). Reveals the structure where “ad budget looks efficient, but actually another ASIN is being sold in the background.”

Using AMC data for the past {{PERIOD}}, decompose ad-attributed customers' purchases
into "clicked ASINs" and "halo (other brand ASINs)."
Return revenue share × LTV share × top halo-beneficiary ASINs.
When halo share is high, identify which ASINs are the halo beneficiaries.

Placeholders: {{PERIOD}} (e.g. last 90 days).

Returns: Revenue composition (clicked ASINs vs halo) × LTV composition × ranked halo-beneficiary ASINs. Includes a judgment against the typical 30–50% halo range.

What the AI will do

  1. Extract ad-touched customers and classify each customer’s purchases into “clicked ASINs” and “other ASINs”
  2. Compute revenue and LTV composition for each
  3. Rank halo-beneficiary ASINs by revenue
  4. If halo share is unexpectedly high / low, interpret the meaning (brand loyalty vs one-off purchases)
  5. Segments below 100 unique customers are excluded per AMC privacy rules

Interaction pattern: Immediate display (read-only, no ad-setting changes)

Prompt 3: Monthly brand acquisition health check

Section titled “Prompt 3: Monthly brand acquisition health check”

A one-pager dashboard for “is the brand healthy this month?” Lines up monthly trends for NTB rate / RPR / LTV:CAC and flags the one worst-performing indicator.

Using AMC data for the past {{PERIOD}}, return a monthly brand-wide health check.
Month × NTB rate (unique-customer basis) × RPR × CAC × LTV × LTV:CAC ratio.
Judge each indicator as improving / flat / declining,
and call out the single most concerning indicator.

Placeholders: {{PERIOD}} (e.g. last 6 months, last 12 months. 3+ months recommended for trend judgment).

Returns: Month × NTB rate × RPR × CAC × LTV × LTV:CAC ratio + trend judgment + the single worst-performing indicator.

What the AI will do

  1. Aggregate AMC data by month
  2. Compute NTB rate / RPR / CAC / LTV per month
  3. Trend-judge each indicator
  4. Use the LTV:CAC trend for overall health
  5. Surface the single worst-deteriorating indicator
  6. Cells below 100 unique customers per month are excluded per AMC privacy rules

Interaction pattern: Immediate display (read-only, no ad-setting changes)

Prompt 4: NTB acquisition strength by ASIN

Section titled “Prompt 4: NTB acquisition strength by ASIN”

Break down “how many new-to-brand customers each ASIN brought in” and “whether the ad spend was worth it” at the ASIN level. Helps you decide which ASINs to push for acquisition vs which to treat as retention plays.

Using AMC data for the past {{PERIOD}}, surface NTB (New-to-Brand) performance
for SP / SB / SD ads at the ASIN level.
Return a table of ASIN × total purchases × NTB purchases × NTB rate ×
attributed revenue × ad spend × ACoS.
Separately list ASINs with poor NTB ROI (scale-down candidates)
and the top 10 ASINs by NTB rate with room to scale.

Placeholders: {{PERIOD}} (e.g. last 30 days, last month).

Returns: ASIN × total purchases × NTB purchases × unique customers × NTB unique customers × NTB rate × attributed revenue × ad spend × ACoS, with “scale down” / “scale up” labels.

What the AI will do

  1. Extract per-ASIN purchase counts and new-customer counts from AMC data
  2. Compute NTB rate (NTB customers ÷ total customers) and CAC per NTB customer
  3. Rank ASINs by NTB acquisition efficiency (NTB ÷ ad spend)
  4. Split into top 10 (scale up) and bottom (scale down)
  5. ASINs with fewer than 100 unique customers are excluded per AMC privacy rules

Interaction pattern: Immediate display (read-only, no ad-setting changes)

Prompt 5: Sort ASINs into acquisition / retention / sunset

Section titled “Prompt 5: Sort ASINs into acquisition / retention / sunset”

Auto-classify each ASIN as acquisition-strong, retention-driver, or sunset candidate. Used to clarify the role of each ASIN in the portfolio.

Using AMC data for the past {{PERIOD}}, sort ASINs currently running SP / SB / SD ads
into "acquisition / retention / sunset" buckets.
State the classification criteria (NTB rate / ad ROI / repeat purchases) explicitly.
Show the top 5 ASINs of each bucket with their metrics,
and propose a budget reallocation plan.

Placeholders: {{PERIOD}} (e.g. last 90 days. 90+ days recommended to absorb seasonality).

Returns: ASIN × bucket × NTB rate × ad ROI × repeat purchases × recommended action. Top 5 ASINs per bucket + budget reallocation plan.

What the AI will do

  1. Extract per-ASIN NTB rate / ad ROI / repeat purchases
  2. Auto-sort into 3 buckets by 3-axis scoring (criteria stated first)
  3. Extract the top 5 ASINs of each bucket
  4. Propose which ASIN group should get more budget
  5. ASINs below 100 unique customers are excluded per AMC privacy rules

Interaction pattern: Immediate display (read-only, no ad-setting changes)

Prompt 6: Monthly cohort retention and LTV

Section titled “Prompt 6: Monthly cohort retention and LTV”

Group customers by acquisition month (cohort) and measure how much they continue to buy at M+1 / M+3 / M+6. Surfaces whether newer cohorts are stickier than older ones and when churn typically happens.

Using AMC data for the past {{COHORT_RANGE}}, return monthly cohort retention and LTV.
Acquisition month × cohort size × M+1 revenue × M+3 revenue × M+6 revenue × cohort LTV.
Judge whether newer cohorts are retaining better than older ones — call out the trend.

Placeholders: {{COHORT_RANGE}} (e.g. last 12 months, 2025-04 to 2026-03. 6+ months recommended to see cohort growth).

Returns: Acquisition month × cohort size × M+1 / M+3 / M+6 revenue × cohort LTV × monthly trend judgment. Months with cohort size under 100 unique customers are suppressed by AMC.

What the AI will do

  1. Build a cohort for each acquisition month from ad-touched first-time buyers
  2. Aggregate revenue each cohort produced at M+1 / M+3 / M+6
  3. Trend cohort LTV across months (improving / declining / flat)
  4. Cohorts below 100 unique customers are excluded per AMC privacy rules

Interaction pattern: Immediate display (read-only, no ad-setting changes)

Prompt 7: Post-purchase retention funnel

Section titled “Prompt 7: Post-purchase retention funnel”

Aggregate “what percent of customers came back N months after their first purchase” by time slot (M+1 / M+3 / M+6). Splitting by acquisition campaign reveals which campaigns bring back customers that actually stick.

Using AMC data for the past {{PERIOD}}, return the post-purchase retention funnel.
Slot (M+1 / M+3 / M+6) × surviving customers × retention rate × average repeat interval.
Split by {{BY_CAMPAIGN}}, and judge which campaigns bring in customers who stick
and which do not.

Placeholders: {{PERIOD}} (e.g. last 6 months, since 2025-10), {{BY_CAMPAIGN}} (e.g. by acquisition campaign or overall only).

Returns: Slot × surviving customers × retention rate × average repeat interval. When campaign breakdown is requested, top and bottom campaigns are labeled “stickiest” / “weakest retention.”

What the AI will do

  1. Extract first-time buyers from AMC data
  2. Aggregate whether each slot (M+1 / M+3 / M+6, etc.) saw a repeat purchase
  3. Visualize the retention curve and find drop-off points
  4. With campaign breakdown, rank acquisition paths by retention rate
  5. Segments below 100 unique customers are excluded per AMC privacy rules

Interaction pattern: Immediate display (read-only, no ad-setting changes)

Prompt 8: Ad-type NTB power and revenue comparison

Section titled “Prompt 8: Ad-type NTB power and revenue comparison”

Compare buyers, attributed revenue, NTB share, and NTB ROAS across SP / SB / SD / DSP. Used to decide where to shift budget between ad types.

Using AMC data for the past {{PERIOD}}, return by ad type (SP / SB / SD / DSP):
buyers × attributed revenue × NTB rate × NTB CAC × NTB ROAS.
Recommend where to shift budget, judged by NTB acquisition efficiency.

Placeholders: {{PERIOD}} (e.g. last 30 days, last month).

Returns: Ad type × buyers × attributed revenue × NTB rate × NTB CAC × NTB ROAS, plus a budget shift recommendation. Accounts without DSP delivery return SP / SB / SD only.

What the AI will do

  1. Aggregate AMC data split by ad type for buyers and revenue
  2. Compute NTB rate and NTB CAC
  3. Rank ad types by NTB ROAS (NTB revenue ÷ ad spend)
  4. Flag if budget is overly concentrated in one ad type
  5. Ad types below 100 unique customers are excluded per AMC privacy rules

Interaction pattern: Immediate display (read-only, no ad-setting changes)

Prompt 9: Budget shift toward lowest NTB CAC ad type

Section titled “Prompt 9: Budget shift toward lowest NTB CAC ad type”

Compare NTB acquisition cost across ad types and propose a concrete shift plan to the most efficient type — source / target / projected NTB lift / risks.

Using AMC data for the past {{PERIOD}}, compare NTB CAC across SP / SB / SD / DSP.
Show projected NTB lift if {{BUDGET_SHIFT}} is moved to the lowest-CAC type,
and the amount that should be pulled from the highest-CAC type.
Include risks (scale ceiling / saturation signals).

Placeholders: {{PERIOD}} (e.g. last 30 days, last 60 days), {{BUDGET_SHIFT}} (e.g. $5,000 or 10% of current).

Returns: Ad type × NTB CAC × source / target × projected NTB lift × risk notes. Accounts without DSP delivery compare and shift across SP / SB / SD only.

What the AI will do

  1. Compute NTB CAC per ad type
  2. Identify the lowest- and highest-CAC types
  3. Project NTB lift from shifting {{BUDGET_SHIFT}} to the lowest-CAC type, using current CAC linearly
  4. Flag risks of the shift (scale ceiling / saturation / seasonality)
  5. Ad types below 100 unique customers are excluded per AMC privacy rules

Interaction pattern: Immediate display (read-only, no ad-setting changes)

Prompt 10: Compare M+1 return rate by ad type

Section titled “Prompt 10: Compare M+1 return rate by ad type”

Compare which ad type brings back “sticky” customers using the M+1 return rate. Measures customer quality, not just NTB volume.

Using AMC data for the past {{PERIOD}}, compare M+1 return rate of NTB customers
acquired via SP / SB / SD / DSP.
Ad type × NTB acquired × M+1 returners × return rate × quality KPI.
Judge which ad type acquires the stickiest customers.

Placeholders: {{PERIOD}} (e.g. last 90 days. At least 2 months needed to observe M+1).

Returns: Ad type × NTB acquired × M+1 returners × return rate + “sticky” / “flash” classification. Accounts without DSP delivery return SP / SB / SD only.

What the AI will do

  1. Extract NTB customers by ad type
  2. Determine if each NTB made a second purchase by M+1
  3. Rank ad types by return rate
  4. Classify as “sticky (high return)” or “flash (low return)”
  5. Ad types below 100 unique customers are excluded per AMC privacy rules

Interaction pattern: Immediate display (read-only, no ad-setting changes)

Section titled “Prompt 11: Campaign-level attribution trends”

Track monthly buyer count / revenue / NTB rate at the campaign level. Sorts campaigns into “scale up,” “saturated,” and “sunset” buckets.

Using AMC data for the past {{PERIOD}}, return campaign × month
buyer count × attributed revenue × NTB rate trend.
Bucket campaigns into scale up / saturated / sunset,
and list the top campaigns by NTB-to-spend ratio.

Placeholders: {{PERIOD}} (e.g. last 6 months, since 2025-12. 3+ months recommended for trend reliability).

Returns: Campaign × month × buyers × attributed revenue × NTB rate, plus “scale up / saturated / sunset” labels and the top NTB-to-spend ranking.

What the AI will do

  1. Aggregate AMC data at campaign × month granularity
  2. Judge trend from month-over-month growth in buyers and revenue
  3. Use NTB rate trend to detect “NTB saturation”
  4. Bucket each campaign as “scale up / saturated / sunset”
  5. Campaign × month cells below 100 unique customers are excluded per AMC privacy rules

Interaction pattern: Immediate display (read-only, no ad-setting changes)

Prompt 12: Early-detect saturating campaigns

Section titled “Prompt 12: Early-detect saturating campaigns”

Detect “saturation signals” — falling NTB rate, rising CPA, flat revenue — monthly and use them as a signal to tighten budget or add fresh keywords. Where Prompt 11 is exhaustive, this is the alert-only version that surfaces only campaigns with worsening signs.

Using AMC data for the past {{PERIOD}}, early-detect saturating / diminishing-return campaigns.
Check monthly: NTB rate decline / CPA rise / revenue plateau.
List campaigns with "2 or more saturation signals" as alerts.
Include action proposals (tighten budget / add new keywords / expand targets).

Placeholders: {{PERIOD}} (e.g. last 6 months. Minimum 3 months required for trend judgment).

Returns: Campaign × saturation-signal count × main deteriorating metric × monthly trend × recommended action. Campaigns with less than 3 months of history go into a separate “trend judgment deferred” section.

What the AI will do

  1. Compute monthly change of NTB rate / CPA / revenue at campaign × month granularity
  2. Set “worsening flag” for each, extract campaigns with 2+ flags
  3. Identify the main deteriorating metric per campaign
  4. Auto-suggest actions matched to the deteriorating metric (NTB decline → new KWs / CPA rise → tighter bids / revenue plateau → target expansion)
  5. Campaigns with less than 3 months of monthly data are excluded as “trend judgment deferred”
  6. Campaign × month cells below 100 unique customers are excluded per AMC privacy rules

Interaction pattern: Immediate display (read-only, no ad-setting changes)

Prompt 13: Identify customer-count vs revenue divergence

Section titled “Prompt 13: Identify customer-count vs revenue divergence”

Identify campaigns where customer count is growing but revenue is shrinking (basket-size shrinking) or customer count is shrinking but revenue holds (price up). Structural shifts that headline revenue alone hides.

Using AMC data for the past {{PERIOD}}, identify campaigns where
customer count and revenue are diverging.
Campaign × customer-count change % × revenue change % × basket-size change × likely cause.
Split into "more customers, less revenue (price down)"
and "fewer customers, revenue held (price up)" patterns.

Placeholders: {{PERIOD}} (e.g. last 3 months vs prior 3 months. Needs a comparison-pair period).

Returns: Diverging campaigns × customer / revenue / basket-size change % × pattern classification × likely-cause hypothesis.

What the AI will do

  1. Compute customer-count and revenue change % at campaign × month granularity
  2. Extract the “more customers, less revenue” and “fewer customers, revenue held” patterns
  3. Include the basket-size (revenue ÷ customers) change
  4. Hypothesize likely causes (promo discount / price hike / product mix shift / sale period)
  5. Campaigns below 100 unique customers are excluded per AMC privacy rules

Interaction pattern: Immediate display (read-only, no ad-setting changes)

Section titled “Related categories”