Check KPIs on the Dashboard

Prompts for daily and weekly ad performance reviews. Use them to compare TACoS / ACoS / ROAS against a prior period, aggregate KPIs by label, or measure the Before/After impact of a measure.

When to use this category

“I want to compare this week’s ad KPIs with last week’s”
“I want to see KPIs by brand / use case”
“I want to measure the effect of last week’s bid changes”
“Show the US account dashboard in USD”

Prompt 1: Dashboard KPIs (with prior-period comparison)

Goal

Get a consolidated view of TACoS / ACoS / ROAS / ad spend / sales / orders for the most recent period, compared against the same number of days in the prior period.

When to use

You want to compare this week’s ad KPIs against the prior period
You want a consolidated view of recent ad spend, sales, and orders
You want to see TACoS / ACoS / ROAS in one list

Prompt

Fetch the ad dashboard for the last {{PERIOD}}.
Use daily frequency (frequency=daily) with prior-period comparison (include_previous=true),
and return TACoS / ACoS / ROAS / ad spend / sales / orders.

Determine the currency automatically from the account name:
- US / USA / .com → USD
- DE / EU → EUR
- UK → GBP
- JP / Japan → JPY

Placeholders:

Placeholder	Description	Default
`{{PERIOD}}`	Aggregation period (e.g., last 7 days, this week, last 28 days)	Last 7 days if unspecified

What the AI does

Fixes the period — Automatically computes the specified period and a prior period of the same length
- Why same length: comparing different-length periods breaks the delta math, so day counts are always aligned
Auto-detects currency — USD if the account name contains US or .com, EUR for EU / DE, GBP for UK, otherwise JPY
- Why auto-detect: showing foreign accounts in JPY by mistake produces a 100x order-of-magnitude error
Aggregates KPIs — Rolls up ad spend, sales, orders, TACoS, ACoS, and ROAS at daily granularity, with prior-period deltas
Returns a table — Three-column layout: current period, prior period, delta (absolute and %)

Example Output

Ad Dashboard (last 7 days vs. previous 7 days / JPY)

Metric Current Prior Delta
Ad spend ¥1,248,300 ¥1,196,700 +4.3%
Sales ¥9,820,400 ¥8,950,100 +9.7%
Orders 1,284 1,178 +9.0%
ACoS 12.7% 13.4% -0.7pt
TACoS 6.2% 6.5% -0.3pt
ROAS 7.87 7.48 +5.2%

Sales grew +9.7% while ad spend only grew +4.3%, improving ACoS by 0.7pt. Reply with “show the daily chart” if you’d like to drill into the daily breakdown.

Metric	Current	Prior	Delta
Ad spend	¥1,248,300	¥1,196,700	+4.3%
Sales	¥9,820,400	¥8,950,100	+9.7%
Orders	1,284	1,178	+9.0%
ACoS	12.7%	13.4%	-0.7pt
TACoS	6.2%	6.5%	-0.3pt
ROAS	7.87	7.48	+5.2%

Safeguards

Current and prior periods are always aligned to the same number of days (if you provide mismatched periods manually, the AI asks for confirmation)
Currency detection is an inference based on the account name. The AI confirms with you before running if the account name is ambiguous
Dashboard view only — no changes are made to bids, budgets, or campaigns
If the period includes data that hasn’t finalized yet (the last 1–2 days), the AI notes that explicitly

Execution mode

Phase	State	Conditions
Phase 1 (now)	One-shot	Available with just a Picaro connection
Phase 2 (Q3 2026)	Daily-ready	Save as a “Saved Prompt” for one-click morning runs
Phase 3 (Q4 2026)	Auto-delivery	Auto-posted to Slack every morning at 8:00
Phase 4 (2027)	Auto + anomaly detection	Alerts only on KPI spikes; summary only during normal periods

What to do next

Anomaly detection → bid optimization — if ACoS is worsening, move on to bid adjustments
Log measures and assess impact — cross-reference KPI changes with your measure log
Set monthly KPI targets — register target ACoS / TACoS so they appear on the dashboard

FAQ

Q. If I say “last week,” is the prior period the week before that? A. Yes. If the current period is 7 days, the prior period is the same 7 days with the same day-of-week structure. For month boundaries or holiday-skew concerns, specify dates explicitly.

Q. Can I show JPY and USD side by side on one screen? A. One currency per request. To compare multiple accounts, use Prompt 4 or run separately per account.

Prompt 2: KPI Aggregation by Label Classification

Goal

Use a pre-designed Taxonomy to slice KPIs by classification axes such as brand, use case, or competitor, and compare them.

When to use

You want to see KPIs by brand / use case / competitor
You want to compare your own brand against competitors
You want a Before/After comparison by label

Prompt

Aggregate KPIs by taxonomy for the last {{PERIOD}}.
Filter to category {{TAXONOMY}},
and output Impressions / clicks / ad spend / sales / orders / CTR / CVR / ACoS / ROAS.

If you want a prior-period comparison, also pass include_previous=true.

Placeholders:

Placeholder	Description	Default
`{{PERIOD}}`	Aggregation period (e.g., last 4 weeks)	Last 28 days if unspecified
`{{TAXONOMY}}`	Taxonomy to aggregate (e.g., brand / use / competitor)	All registered taxonomies if unspecified

What the AI does

References the Taxonomy — Loads the taxonomy registered in Design label classifications and maps each Search term and ASIN to a classification axis
- Why a Taxonomy is required: without registered axes, everything falls into “uncategorized” and no meaningful comparison is possible
Aggregates metrics — Rolls up metrics by axis, with prior-period comparison alongside
Handles Auto Campaigns — Search terms in Auto Campaigns are anonymized, so the AI falls back to keyword-string-based inference
- Why fall back: since the anonymized raw data isn’t usable, inference at least preserves directional trends per axis

Example Output

Taxonomy KPIs (category: brand / last 28 days / JPY)

Label Ad spend Sales ACoS ROAS Prior ACoS
Own brand ¥2,180,400 ¥21,948,000 9.9% 10.07 10.4%
Competitor A ¥684,200 ¥3,041,800 22.5% 4.44 24.1%
Competitor B ¥412,800 ¥1,587,300 26.0% 3.84 27.8%
Generic ¥1,824,500 ¥10,488,000 17.4% 5.75 18.2%

Own-brand ACoS is 9.9% (down 0.5pt vs. prior period), while every competitor slot is above 20% with ROAS hovering around 4x.

Label	Ad spend	Sales	ACoS	ROAS	Prior ACoS
Own brand	¥2,180,400	¥21,948,000	9.9%	10.07	10.4%
Competitor A	¥684,200	¥3,041,800	22.5%	4.44	24.1%
Competitor B	¥412,800	¥1,587,300	26.0%	3.84	27.8%
Generic	¥1,824,500	¥10,488,000	17.4%	5.75	18.2%

Safeguards

Items without a registered taxonomy are grouped under “uncategorized,” with a note at the top of the output
Auto Campaign rows are flagged as inferred values in the output
Dashboard view only — label definitions and campaign settings are not changed
If the classification axis is omitted, the AI iterates through every registered axis (it shows the count up front since output volume can grow)

Execution mode

Phase	State	Conditions
Phase 1 (now)	One-shot	Available once a taxonomy is registered
Phase 2 (Q3 2026)	Daily-ready	Save as a “Saved Prompt”
Phase 3 (Q4 2026)	Weekly auto-delivery	Posted to Slack on Monday mornings
Phase 4 (2027)	Auto + anomaly detection	Alerts only on label-level ACoS spikes

What to do next

Design label classifications — start here if your taxonomy isn’t registered yet
Analyze sales contribution with N-grams — drill into competitor slots at the Search term level
Adjust bids — tighten bids on high-ACoS competitor labels

Prompt 3: Before/After Impact Measurement for a Measure

Goal

For a completed measure, aggregate the Before vs. After deltas in ad spend, sales, ACoS, and ROAS across two periods, then classify as improved / neutral / degraded.

When to use

You want to measure the effect of last week’s bid changes
You want a verdict on ACoS / ROAS movement
You want to pull the comparison periods from a past measure log by ID

Prompt

Show a Before/After comparison for a measure.
- Before period: {{BEFORE_START}} – {{BEFORE_END}}
- After period: {{AFTER_START}} – {{AFTER_END}}
- Event ID: {{EVENT_ID}} (optional; the ID registered in the measure log)

Aggregate the deltas in ad spend, sales, ACoS, and ROAS,
and classify as improved / neutral / degraded.

Placeholders:

Placeholder	Description	Default
`{{BEFORE_START}}` / `{{BEFORE_END}}`	Period before the measure	Auto-computed from the measure log when an Event ID is supplied
`{{AFTER_START}}` / `{{AFTER_END}}`	Period after the measure	Same as above
`{{EVENT_ID}}`	ID registered in the measure log	Optional; if omitted, only the supplied periods are used

What the AI does

Fixes the periods — Aligns the Before and After period lengths and excludes weeks where data hasn’t finalized
- Why align period lengths: mismatched lengths break the absolute-value sales comparison
- Why exclude unfinalized weeks: mixing in still-aggregating data destabilizes the verdict
Aggregates the deltas — Computes absolute and % deltas for ad spend, sales, ACoS, and ROAS
Classifies — Returns improved / neutral / degraded based on the ACoS improvement and statistical noise
- Why three-way classification: to avoid misreading tiny deltas as “improved,” the noise band is treated as neutral
Links to the measure log — When an Event ID is supplied, the AI cross-references the log from Log measures and assess impact and includes the measure description

Example Output

Measure Before/After Comparison (Event ID: evt_2026_04_28_bid_down / JPY)

Measure: Lowered bids on 18 high-ACoS keywords by an average of -22%

Metric Before (4/14–4/27) After (4/28–5/11) Delta Verdict
Ad spend ¥412,800 ¥318,400 -22.9% —
Sales ¥1,892,000 ¥1,748,500 -7.6% —
ACoS 21.8% 18.2% -3.6pt improved
ROAS 4.58 5.49 +19.9% improved

Overall verdict: improved (ACoS -3.6pt, ROAS +19.9%, sales decline smaller than the ad spend cut)

Metric	Before (4/14–4/27)	After (4/28–5/11)	Delta	Verdict
Ad spend	¥412,800	¥318,400	-22.9%	—
Sales	¥1,892,000	¥1,748,500	-7.6%	—
ACoS	21.8%	18.2%	-3.6pt	improved
ROAS	4.58	5.49	+19.9%	improved

Safeguards

Before and After period lengths are always aligned to the same number of days (the AI asks for confirmation if they differ)
The verdict is a mechanical classification based on delta aggregation only; it doesn’t account for external factors (seasonality, stock-outs, price changes). Final judgment is up to you
If the Event ID isn’t found in the measure log, the AI says so and runs the period aggregation only
Dashboard view only — no changes are made to bids, budgets, or campaigns

Execution mode

Phase	State	Conditions
Phase 1 (now)	One-shot	Available with just a Picaro connection
Phase 2 (Q3 2026)	Stays one-shot	Periods still need human judgment per measure
Phase 3 (Q4 2026)	Daily auto	Before/After auto-computed 7 and 14 days after each measure runs
Phase 4 (2027)	Auto-delivery	Results auto-posted to Slack with an external-factors comment field

What to do next

Log measures and assess impact — the prerequisite for mechanizing Before/After
Generate reports — comprehensive reports that include impact measurement
Adjust bids — roll back or readjust if the verdict is degraded

Prompt 4: Currency-Specific Dashboard for Foreign Accounts

Goal

Fetch the dashboard for foreign accounts (US / EU / UK / JP, etc.) displayed in local currency.

When to use

You want the US account dashboard in USD
You want European account KPIs in EUR
You want to compare accounts across multiple currencies

Prompt

Fetch the dashboard for account_id {{ACCOUNT_ID}} (account name: {{ACCOUNT_NAME}}).
Infer currency from the account name:
- US / USA / .com → currency=USD
- DE / EU → currency=EUR
- UK → currency=GBP
- JP / Japan → currency=JPY

Run with period {{PERIOD}}, daily frequency, and prior-period comparison.

Placeholders:

Placeholder	Description	Default
`{{ACCOUNT_ID}}`	Target account ID	Picked from connected Picaro accounts
`{{ACCOUNT_NAME}}`	Account name (used for currency detection)	Same as above, fetched automatically
`{{PERIOD}}`	Aggregation period	Last 7 days if unspecified

What the AI does

Detects currency — Resolves to JPY / USD / EUR / GBP based on keywords in the account name
- Why these four: they’re the currencies Picaro currently supports. Other accounts are converted to JPY with a note
Confirms ambiguous account names — For names like JP US Test that match multiple candidates, the AI confirms with you before running
- Why confirm: to prevent 100x order-of-magnitude errors
Fetches the dashboard — Same as Prompt 1, with prior-period comparison
Attaches currency symbols — Renders each number with the appropriate symbol (¥ / $ / € / £)

Example Output

Ad Dashboard (last 7 days vs. previous 7 days / USD)

Account: AMZN-US-Brand (account_id: acct_xxxx) / Currency: USD

Metric Current Prior Delta
Ad spend $8,420 $7,980 +5.5%
Sales $62,140 $57,310 +8.4%
Orders 482 451 +6.9%
ACoS 13.6% 13.9% -0.3pt
TACoS 7.1% 7.4% -0.3pt
ROAS 7.38 7.18 +2.8%

Healthy trend in the current period: sales +8.4% and ACoS -0.3pt.

Metric	Current	Prior	Delta
Ad spend	$8,420	$7,980	+5.5%
Sales	$62,140	$57,310	+8.4%
Orders	482	451	+6.9%
ACoS	13.6%	13.9%	-0.3pt
TACoS	7.1%	7.4%	-0.3pt
ROAS	7.38	7.18	+2.8%

Safeguards

Supported currencies are JPY / USD / EUR / GBP. Other accounts are converted to JPY with a note at the top
The AI confirms with you before running if the account name matches multiple currency candidates
Currency conversion is display-only; no changes are made to the Amazon Ads data
When comparing across multiple accounts, run separately per account (mixed-currency display tends to cause order-of-magnitude errors)

Execution mode

Phase	State	Conditions
Phase 1 (now)	One-shot	Available for foreign accounts connected to Picaro
Phase 2 (Q3 2026)	Daily-ready	Save per-account “Saved Prompts”
Phase 3 (Q4 2026)	Auto-delivery	Slack delivery every morning to per-account channels
Phase 4 (2027)	Auto + anomaly detection	Learns per-currency anomaly thresholds, alerts only on spikes

What to do next

Set monthly KPI targets — register per-account target ACoS / TACoS
Generate reports — produce monthly reports across multiple accounts
Connect an Amazon Ads account — add any unconnected accounts here

Prompt Quick Reference

What you want	Prompt to use
Standard Dashboard KPIs	Prompt 1
Aggregation by brand / use case	Prompt 2
Before/After 比較 of a measure	Prompt 3
Currency-specific foreign account	Prompt 4

Design label classifications — prerequisite for Prompt 2
Log measures and assess impact — pairs with Prompt 3
Generate reports — comprehensive reports including Dashboard KPIs
Adjust bids — move here once an anomaly is detected

Check KPIs on the Dashboard

When to use this category

Prompt 1: Dashboard KPIs (with prior-period comparison)

Goal

When to use

Prompt

What the AI does

Example Output

Safeguards

Execution mode

What to do next

FAQ

Prompt 2: KPI Aggregation by Label Classification

Goal

When to use

Prompt

What the AI does

Example Output

Safeguards

Execution mode

What to do next

Prompt 3: Before/After Impact Measurement for a Measure

Goal

When to use

Prompt

What the AI does

Example Output

Safeguards

Execution mode

What to do next

Prompt 4: Currency-Specific Dashboard for Foreign Accounts

Goal

When to use

Prompt

What the AI does

Example Output

Safeguards

Execution mode

What to do next

Prompt Quick Reference

Related categories