AI Pilot

Your rules. Your logic. Running on their own.

AI Pilot is the manual cockpit for sellers and agencies who’d rather build the strategy than hand it over. Compose the exact rules you want, arrange the order they run, and let them work — while the AI stays a helper you switch on, never a driver you can’t switch off.

Rule automation

Build a rule for anything you can measure

Every rule you build tests real Amazon metrics and takes a real action — no presets, no black box deciding for you. Stack conditions together, set the lookback window that matters to you, arrange your rules in the order they fire, and set each to run as often as every hour.

You decide what runs first.

Rules apply in the order you arrange them — raises, cuts, negatives, pauses, whatever sequence your strategy calls for. When two rules could touch the same target, precedence is yours to set, not ours to guess. That’s the difference between automation you supervise and automation that surprises you.

A rule can test

ACoS

ROAS

CPA

CVR

CTR

CPC

Current bid

Clicks

Impressions

Spend

Sales

PPC orders

Units sold

Top-of-search IS

Top-of-search modifier

Match type

New-to-brand orders

New-to-brand sales

New-to-brand rate

vs. your target

Keyword & target actions

Set bid to $

Raise / lower bid $

Raise / lower bid %

Set to CPC

Revenue per click

Bid to campaign target metric

Enable / pause targeting

Add negative

Campaign actions

Adjust placement

Set daily budget

Raise / lower budget $

Raise / lower budget %

Enable / pause campaign

Stack conditions with AND / OR, set your own lookback window per condition, and arrange rules in the order they fire. You compose the logic — nothing is a preset.

New Automation Rule
Define when and what action to take automatically
Rule Settings
RULE NAME
High CVR, High ACOS
APPLY TO
Keywords
Hit counter rules run continuously after each sync — no schedule.
Trigger
Threshold
Metric crosses a value
Hit Counter
Accumulates since last run
HIT COUNTER — RUN WHEN TOTAL REACHES
Clicksreaches10since last run
Hourly data (Marketing Stream) — always on for Keywords & PATs
Reacts same-day using intraday data. Falls back to daily reports.
Click Gate
Click Gate
Rule only fires if minimum clicks reached — prevents acting on low-data keywords
IF ConditionsClear all
IFACoS (%)<=25%Lookback ?7 daysExclude ?Today only (default)
AND
ANDCVR (%)>18%Lookback ?4 daysExclude ?Today + 1 day
OR
IFACoS (%)<=25%Lookback ?14 daysExclude ?Today + 1 day
AND
ANDClicks>25Lookback ?All timeExclude ?Today + 2 days
AND
ANDCVR (%)<22%Lookback ?60 daysExclude ?Today + 3 days
AND
ANDNTB Orders>3Lookback ?All timeExclude ?Today only (default)
+ OR (additional condition)
THEN
THEN Take Action+ Add Action
Increase Bid (%)25%Ceiling $2.75
Apply To Campaigns
APPLY TO CAMPAIGNS
No campaigns selected — pick campaigns/tags, or check 'Apply to all campaigns' below
Apply To Campaign Tags
APPLY TO CAMPAIGN TAGS
None
No tag filter — applies to all campaigns above
Apply to all campaigns in this account
Rule is Active CancelCreate Rule
New Automation Rule
Define when and what action to take automatically
Rule Settings
RULE NAME
High CVR, High ACOS
APPLY TO
Keywords
Trigger
Threshold
Metric crosses a value
Hit Counter
Accumulates since last run
HIT COUNTER — RUN WHEN TOTAL REACHES
Clicksreaches10
Hourly data (Marketing Stream)
Reacts same-day using intraday data.
Click Gate
Click Gate
Rule only fires if minimum clicks reached
IF ConditionsClear all
IFACoS (%)<=25%Lookback7 days
AND
ANDCVR (%)>18%Lookback4 days
OR
IFACoS (%)<=25%Lookback14 days
AND
ANDClicks>25LookbackAll time
AND
ANDCVR (%)<22%Lookback60 days
AND
ANDNTB Orders>3LookbackAll time
+ OR (additional condition)
THEN
THEN Take Action+ Add Action
Increase Bid (%)25%Ceiling $2.75
Apply To Campaigns
APPLY TO CAMPAIGNS
No campaigns selected — pick campaigns/tags, or check 'Apply to all campaigns' below
Apply To Campaign Tags
APPLY TO CAMPAIGN TAGS
None
Apply to all campaigns in this account
Rule is Active CancelCreate Rule
New Automation Rule
Define when and what action to take automatically
IF Conditions
IFACoS (%)
<=25%
Lookback 7 days · Exclude Today only
ANDCVR (%)
>18%
Lookback 4 days · Exclude Today + 1 day
OR
IFACoS (%)
<=25%
Lookback 14 days · Exclude Today + 1 day
ANDClicks
>25
Lookback All time · Exclude Today + 2 days
ANDCVR (%)
<22%
Lookback 60 days · Exclude Today + 3 days
ANDNTB Orders
>3
Lookback All time · Exclude Today only
THEN
THEN Take Action
Increase Bid (%)
25%Ceiling $2.75
Trigger
Threshold
Metric crosses a value
Hit Counter
Accumulates since last run
HIT COUNTER — RUN WHEN TOTAL REACHES
Clicksreaches10
Hourly data (Marketing Stream)
Reacts same-day using intraday data.
Click Gate
Click Gate
Rule only fires if minimum clicks reached

The real builder — compose conditions, pick actions, set your order.

Run on the cadence you choose

Set each rule to run hourly, daily, every few days, weekly, or monthly — as often or as rarely as its job calls for.

Fresh metrics every hour

We pull your data from Amazon Marketing Stream every hour — so hourly rules act on what’s happening now, not yesterday’s report.

Two ways to trigger

Fire whenever your chosen threshold is crossed — ACoS, clicks, impressions, and more — or on a running tally of clicks, spend, or orders.

A click gate for thin data

Set a minimum number of clicks and the rule leaves a keyword alone until there’s enough data to trust. No acting on two unlucky clicks.

Every condition reads its own window

Judge ACoS over the last few days while another condition weighs orders over a longer stretch. Per-condition lookbacks, not one window for the whole rule.

Test against your own target

Compare a metric to each campaign’s goal instead of a fixed number — one rule adapts across campaigns with different targets.

Stack multiple actions

One rule can take several actions in a single pass — lower the bid and add the wasting term as a negative, together.

Compose the logic, not a preset

Stack conditions with AND inside a group, OR across groups. Build the exact logic your strategy calls for, as deep as you need.

Keyword manager

Every keyword, every metric, one screen

Your whole account’s keyword performance in one sortable table — 30 columns from spend and ACoS to new-to-brand and rank position. Filter by campaign, tag, match type, or any metric range. Edit a bid inline, or select a batch and adjust, pause, negate, or tag in a single move.

Performance TrendsSpendSalesOrdersClicksImpressionsACoS %ROASCPC
SPEND
$1,011.16
SALES
$1,453.20
ACOS
69.6%
ROAS
1.44x
ORDERS
79
CLICKS
395
IMPR
135.9K
CPC
$2.56
CTR
0.3%
CVR
20.0%
CPA
$12.80
SEARCH
Search keywords…
CAMPAIGN
None
AD GROUP
None
CAMPAIGN TAG
None
KEYWORD TAG
None
CAMPAIGN TYPE
All (4)
MATCH TYPE
All (3)
STATUS
All (3)
MANAGED
None
KPI FILTERS
IMPRESSIONS
MinMax
CLICKS
MinMax
CTR (%)
Min %Max %
SPEND ($)
$ Min$ Max
SALES ($)
$ Min$ Max
CVR (%)
Min %Max %
CPC ($)
$ Min$ Max
ACOS (%)
Min %Max %
ROAS
MinMax
CPA ($)
$ Min$ Max
BID ($)
$ Min$ Max
Select keywords for bulk actionsToday7d14d30d60d90dAll Time
KEYWORDMATCHBIDIMPRCLICKSCTRSPENDSALESORDACOSROASCPACPCCVRCAMPAIGNRANK
airtight pantry containersAPSPEXACT$2.5313.6K250.2%$53.32$67.96478.5%1.27$13.33$2.1316.0%Norlend GC4 — Research Exact
glass canister set with lidsAPSPBROAD$4.343.6K180.5%$40.46$50.97379.4%1.26$13.49$2.2516.7%Norlend GC4 — Seeding
kitchen storage jarsCPSPBROAD$3.842.1K140.7%$39.98$33.982117.7%0.85$19.99$2.8614.3%Norlend BL6 — Seeding88
flour storage containerCPSPEXACT$4.163.0K110.4%$33.15$50.97365.0%1.54$11.05$3.0127.3%Norlend BL6 — Research Exact44
pantry organization setAPSPBROAD$4.343.2K80.3%$27.44$16.991161.5%0.62$27.44$3.4312.5%Norlend GC4 — Seeding
cereal storage containerAPSPBROAD$4.345.6K100.2%$25.27$33.98174.4%1.34$25.27$2.5310.0%Norlend GC4 — Seeding73
stackable storage canistersCPSPBROAD$3.862.5K90.4%$24.66$84.95529.0%3.44$4.93$2.7455.6%Norlend BL6 — Seeding
airtight food storage setCPSPEXACT$4.162.5K80.3%$23.26$33.98268.5%1.46$11.63$2.9125.0%Norlend BL6 — Research Exact
Performance TrendsSpendSalesOrdersACoS %
SPEND
$1,011
SALES
$1,453
ACOS
69.6%
ROAS
1.44x
ORDERS
79
CLICKS
395
CPC
$2.56
CVR
20.0%
KPI FILTERS
IMPRESSIONS
MinMax
CLICKS
MinMax
CTR (%)
MinMax
SPEND ($)
MinMax
SALES ($)
MinMax
CVR (%)
MinMax
CPC ($)
MinMax
ACOS (%)
MinMax
ROAS
MinMax
CPA ($)
MinMax
BID ($)
MinMax
MATCH TYPE
All (3)
Select keywords for bulk actions7d14d30d60d90d
KEYWORDMATCHBIDCLICKSSPENDSALESACOSROAS
airtight pantry containersEXACT$2.5325$53.32$67.9678.5%1.27
glass canister set with lidsBROAD$4.3418$40.46$50.9779.4%1.26
kitchen storage jarsBROAD$3.8414$39.98$33.98117.7%0.85
flour storage containerEXACT$4.1611$33.15$50.9765.0%1.54
stackable storage canistersBROAD$3.869$24.66$84.9529.0%3.44
Performance Trends
SPEND
$1,011
SALES
$1,453
ACOS
69.6%
ROAS
1.44x
ORDERS
79
CLICKS
395
KPI FILTERS
SPEND ($)
MinMax
ACOS (%)
MinMax
CLICKS
MinMax
ROAS
MinMax
airtight pantry containersEXACT
BID
$2.53
CLICKS
25
SPEND
$53.32
ACOS
78.5%
Norlend GC4 — Research Exact
glass canister set with lidsBROAD
BID
$4.34
CLICKS
18
SPEND
$40.46
ACOS
79.4%
Norlend GC4 — Seeding
kitchen storage jarsBROAD
BID
$3.84
CLICKS
14
SPEND
$39.98
ACOS
117.7%
Norlend BL6 — Seeding
stackable storage canistersBROAD
BID
$3.86
CLICKS
9
SPEND
$24.66
ACOS
29.0%
Norlend BL6 — Seeding

Then drill in — by the hour

Open any keyword and switch daily to hourly

See how a keyword actually moved through the day — impressions, clicks, spend, and sales, hour by hour. It’s built on Amazon Marketing Stream, Amazon’s hour-by-hour feed — the granularity most tools flatten into a single daily total.

DailyHourly 7d14d30d90dAll timeCSV
HOURIMPRCLICKSCTR%SPENDSALESORDACOS%CVR%
Mar 4 12:00 PM4212.4%$2.78$18.99114.6%100%
Mar 4 11:00 AM1180$0.00$0.000
Mar 4 10:00 AM1510$0.00$0.000
Mar 4 9:00 AM1240$0.00$0.000
Mar 4 8:00 AM1330$0.00$0.000
Mar 4 7:00 AM1160$0.00$0.000
Mar 4 6:00 AM720$0.00$0.000
Mar 4 5:00 AM610$0.00$0.000
Recent hours settle within 48h as attribution finalizes
DailyHourly 30d90dAll timeCSV
HOURIMPRCLICKSSPENDSALESACOS%
Mar 4 12:00 PM421$2.78$18.9914.6%
Mar 4 11:00 AM1180$0.00$0.00
Mar 4 10:00 AM1510$0.00$0.00
Mar 4 9:00 AM1240$0.00$0.00
Mar 4 8:00 AM1330$0.00$0.00
Mar 4 7:00 AM1160$0.00$0.00
Mar 4 6:00 AM720$0.00$0.00
Mar 4 5:00 AM610$0.00$0.00
Recent hours settle within 48h as attribution finalizes
DailyHourly
HOURIMPRSPENDSALES
Mar 4 12:00 PM42$2.78$18.99
Mar 4 11:00 AM118$0.00$0.00
Mar 4 10:00 AM151$0.00$0.00
Mar 4 9:00 AM124$0.00$0.00
Mar 4 8:00 AM133$0.00$0.00
Mar 4 7:00 AM116$0.00$0.00
Mar 4 6:00 AM72$0.00$0.00
Mar 4 5:00 AM61$0.00$0.00
Recent hours settle within 48h as attribution finalizes

Filter and sort anything

Search, campaign, tag, match type, status, and metric-range filters. Sort by any column.

Act in bulk

Adjust bids, enable, pause, archive, add negatives, or tag — across a whole selection at once.

Daily or hour by hour

Flip any keyword’s detail from daily to hourly and watch it move through the day.

Outcomes in full view

You see everything the AI does — the results, not a black box

Every keyword and campaign it manages sits right in your Pilot tables, badged so you always know it owns them — same 30 columns, same bid history, same hour-by-hour view. Every bid it moves, every winner it harvests, every point of rank it earns shows up in the reporting you already use. The engine’s ruleset — the logic that decides all of it — stays under the hood, tuned and proprietary. You get the outcomes, fully in view. The formula stays ours.

See where the sale is actually lost.

Every product moves shoppers through four stages — searched, clicked, added to cart, bought. Pulse Pilot shows yours against the market at every stage, so you know whether the problem is your ad, your listing, or your price.

Amazon reports this data for brand-registered products.

Weekly funnel
This counts everything shoppers did in Amazon's search results — whether they found you through an ad or organically.
PRODUCT
Norlend GC4 airtight canister set, 4pk
B0CN7QK4LM
YOU VS THE MARKET
96%
leak
Shoppers who saw it
38,400
ADS + RANKING
You appeared 38,400 times across 196,000 searches — an appearance rate of 19.6%.
Appearance19.6%
GradedLow
1.5% click through rate
4%
leak
Clicked it
576
ADS + LISTING
Out of 31,200 clicks across these searches, 576 came to you.
Yours1.5%
Market avg1.8%
vs market-17%
61.0% add to cart rate
0%
leak
Added it to cart
351
YOUR LISTING
Out of 18,100 cart adds across these searches, 351 came to you.
Yours61.0%
Market avg58.0%
vs market+5%
76.4% buy after carting
0%
leak
Bought it
268
YOUR LISTING
Out of 12,100 purchases, 268 came to you.
Yours76.4%
Market avg66.8%
vs market+14%
OVERALL CONVERSION
of shoppers who clicked, how many bought
Of the shoppers who clicked, 46.5% bought — 19% above the market average on the same terms.
YOURS
46.5%
MARKET AVG
39.0%
DIFFERENCE
+19%
Benchmark: the whole market's performance on the same 31 search terms shoppers used to find you, with 6 adjacent-market terms excluded. Your own totals always include every search term — a real sale is never hidden; relevance only changes the market benchmark.
Weekly funnel
This counts everything shoppers did in Amazon's search results — whether they found you through an ad or organically.
PRODUCT
Norlend GC4 airtight canister set, 4pk
B0CN7QK4LM
YOU VS THE MARKET
96%
leak
Shoppers who saw it
38,400
ADS + RANKING
You appeared 38,400 times across 196,000 searches — an appearance rate of 19.6%.
Appearance19.6%
GradedLow
1.5% click through rate
4%
leak
Clicked it
576
ADS + LISTING
Out of 31,200 clicks across these searches, 576 came to you.
Yours1.5%
Market avg1.8%
vs market-17%
61.0% add to cart rate
0%
leak
Added it to cart
351
YOUR LISTING
Out of 18,100 cart adds across these searches, 351 came to you.
Yours61.0%
Market avg58.0%
vs market+5%
76.4% buy after carting
0%
leak
Bought it
268
YOUR LISTING
Out of 12,100 purchases, 268 came to you.
Yours76.4%
Market avg66.8%
vs market+14%
OVERALL CONVERSION
of shoppers who clicked, how many bought
Of the shoppers who clicked, 46.5% bought — 19% above the market average on the same terms.
YOURS
46.5%
MARKET AVG
39.0%
DIFFERENCE
+19%
Benchmark: the whole market's performance on the same 31 search terms shoppers used to find you, with 6 adjacent-market terms excluded. Your own totals always include every search term — a real sale is never hidden; relevance only changes the market benchmark.
Weekly funnel
This counts everything shoppers did in Amazon’s search results — whether they found you through an ad or organically.
PRODUCT
Norlend GC4 airtight canister set, 4pk
B0CN7QK4LM
YOU VS THE MARKET
96%leak
Shoppers who saw it
38,400
ADS + RANKING
You appeared 38,400 times across 196,000 searches — an appearance rate of 19.6%.
Appearance19.6%
GradedLow
1.5% click through rate
4%leak
Clicked it
576
ADS + LISTING
Out of 31,200 clicks across these searches, 576 came to you.
Yours1.5%
Market avg1.8%
vs market-17%
61.0% add to cart rate
0%leak
Added it to cart
351
YOUR LISTING
Out of 18,100 cart adds across these searches, 351 came to you.
Yours61.0%
Market avg58.0%
vs market+5%
76.4% buy after carting
0%leak
Bought it
268
YOUR LISTING
Out of 12,100 purchases, 268 came to you.
Yours76.4%
Market avg66.8%
vs market+14%
OVERALL CONVERSION
of shoppers who clicked, how many bought
YOURS
46.5%
MARKET AVG
39.0%
DIFF
+19%
Of the shoppers who clicked, 46.5% bought — 19% above the market average on the same terms.
Benchmark: the whole market's performance on the same 31 search terms shoppers used to find you, with 6 adjacent-market terms excluded. Your own totals always include every search term — a real sale is never hidden; relevance only changes the market benchmark.

The stage that’s costing you

Each stage gets graded against what the rest of the market achieves. The weakest one gets called out by name, so you’re not guessing which end of the funnel to fix.

It tells you whose problem it is

Clicks are your ad’s job. Cart-adds are your listing’s. Purchases are your price and reviews. Each stage points at the thing that actually controls it.

Query by query, week over week

Every search term shoppers used to find your product, tracked weekly against the market — so you can see a listing change work, or fail, in the data.

It feeds the engine, not just the report

The same market conversion data setting these grades is what the AI grades your keywords against. One source of truth, one bar.

Search-term harvest

Harvest your winners — and you decide where they go

Your harvest reads the real phrases shoppers type to find you and promotes the proven winners into keywords. Build a harvest rule, pick its destination, and every winner it finds lands exactly where you want it.

Choose your destination

Its own campaign

Graduate a winner into a dedicated single-keyword campaign — its own budget, its own bid, isolated so you can scale a proven term without the noise of the campaign it came from.

A pooled campaign

Route winners into an existing campaign that pools your keywords — keep your account structured exactly the way you already run it.

RULE NAME
Harvest converting search terms
Schedule
FREQUENCY
Daily
TIME
07:00 AM PST
LOOKBACK PERIOD
90 days
If search term conditions
Orders>0
AND
ACoS (%)<20
AND
CVR (%)>15
SOURCE CAMPAIGN
SPNorlend GC4 - Seeding
DESTINATIONS
#1Single keyword / PAT campaign
Norlend GC4 - 2026 - Research - Campaign first -
Exact 0.75 Phrase Broad ASIN
Use src CPC -Neg exact in source -Neg phrase in source -Neg ASIN in source
ALSO ADD AS NEGATIVE IN
+ add another
+ Add Destination
RULE NAME
Harvest converting search terms
Schedule
FREQUENCY
Daily
TIME
07:00 AM PST
LOOKBACK PERIOD
90 days
If search term conditions
Orders>0
AND
ACoS (%)<20
AND
CVR (%)>15
SOURCE CAMPAIGN
SPNorlend GC4 - Seeding
DESTINATIONS
#1Single keyword / PAT campaign
Norlend GC4 - 2026 - Research - Campaign first -
Exact 0.75 Phrase Broad ASIN
Use src CPC -Neg exact in source -Neg phrase in source -Neg ASIN in source
ALSO ADD AS NEGATIVE IN
+ add another
+ Add Destination
RULE NAME
Harvest converting search terms
Schedule
Daily - 07:00 AM PST - 90 days
If search term conditions
Orders>0
AND
ACoS (%)<20
AND
CVR (%)>15
SOURCE CAMPAIGN
SPNorlend GC4 - Seeding
DESTINATIONS
#1Single keyword / PAT campaign
Norlend GC4 - 2026 - Research
Match Exact - 0.75

Winners get promoted. The terms that don’t earn? Your rules handle those — bid them down or pause them, on the conditions you set.

Dayparting

Bid by the hour — on your schedule, not a guess

Paint your own by-hour schedule across the week and pick exactly which keywords and campaigns it governs. Your bids shift up, down, or pause by hour of day — every hour, automatically, on the hours you set. Or let Claude author the schedule from your own hourly data.

Your week, hour by hour

12a

6a

12p

6p

11p

Mon

Tue

Wed

Thu

Fri

Sat

Sun

Bid up

Baseline

Bid down

Pause

Paint your hours

Set every hour of every day — bid up when you convert, down or off when you don’t.

Pick what it governs

Choose exactly which keywords and campaigns the schedule applies to — nothing you didn’t select.

Runs every hour

Once it’s set, it applies your schedule automatically, every hour, on its own.

Every automated change is on the record

Nothing happens in the dark — three logs show exactly what changed, on which keyword or campaign, and when.

Rules Log

Every rule fire and what it changed.

Harvest Log

Every promotion and negative, and where it landed.

Pacing Log

Every budget move, up or down.

COPILOTOPTIONAL AI ASSIST
Hand one campaign to the Autopilot engine.
Copilot is manual by default - every bid, every rule, your call. When you want a hand on a specific campaign, you enroll it. Enrolled campaigns run on the same engine that powers Autopilot, not a lesser version of it. Everything you don't enroll stays exactly as you built it.
Pick where it's headed
You choose a goal, not a formula - launching, growing, holding an ACoS, clearing stock, or winding down.
Five goals across a product's life
Optional target ACoS on top
Goal sets pace and step size
Opening judgment pass on day one
Reconfigure any campaign later
Watch before you commit
Shadow mode shows every bid it would have made without making one. Turn it off when you're convinced.
Full decisions, zero changes
Review it in Account Health
Bids only - budget stays yours
Cuts bids on clicks that never convert
Tunes placement modifiers
Reversible, precisely
Only your bid rules step aside while enrolled. Unenroll and they come back, armed the way you left them.
Your pause rules keep running
Your negation rules keep running
Your budget rules keep running
Only bid authority is handed over
Unenroll any time, no waiting window
Or hand over the hours
On campaigns you haven't enrolled, let the AI write the hour grid instead - a separate opt-in, per rule.
All 168 hours, authored for you
Built from your hourly Stream data
Rewritten every Sunday night
Its reasoning shown to you
Or keep painting the grid yourself
Configure & arm enrollment
These settings apply to the campaign you selected. Each can be reconfigured individually afterwards.
Norlend GC4 - 2026 - Research
Goal *
Launch
Brand-new product with little history. Bids to discover what converts without burning budget.
Expect: ACoS runs high while it learns.
Grow
Proven converter you want to scale. Pushes on winners and expands into new terms.
Expect: spend and sales climb together.
Optimize
Established product held to a specific ACoS. Trims the inefficient, holds the efficient.
Expect: converges to target over weeks.
Liquidate
Clearing stock fast. Bids for velocity, units sold over efficiency.
Expect: ACoS runs hot - that's the trade.
Wind down
Deliberately exiting. Steps bids down and stops chasing new terms.
Expect: a controlled taper, not a cliff.
TARGET ACOS (%) optional
25
Bid management will be AI-controlled. Budget stays under your manual control.
Shadow mode - recommended for testing
Shows you exactly what it would do, and makes no changes.
BackEnroll campaign
The line that matters: the AI never touches a rule you wrote yourself. It manages its own rules on campaigns you enrolled, and leaves yours alone - on enrolled campaigns they're set aside intact, everywhere else they run untouched.
COPILOTOPTIONAL AI ASSIST
Hand one campaign to the Autopilot engine.
Copilot is manual by default - every bid, every rule, your call. When you want a hand on a specific campaign, you enroll it. Enrolled campaigns run on the same engine that powers Autopilot, not a lesser version of it. Everything you don't enroll stays exactly as you built it.
Pick where it's headed
You choose a goal, not a formula - launching, growing, holding an ACoS, clearing stock, or winding down.
Five goals across a product's life
Optional target ACoS on top
Goal sets pace and step size
Opening judgment pass on day one
Reconfigure any campaign later
Watch before you commit
Shadow mode shows every bid it would have made without making one. Turn it off when you're convinced.
Full decisions, zero changes
Review it in Account Health
Bids only - budget stays yours
Cuts bids on clicks that never convert
Tunes placement modifiers
Reversible, precisely
Only your bid rules step aside while enrolled. Unenroll and they come back, armed the way you left them.
Your pause rules keep running
Your negation rules keep running
Your budget rules keep running
Only bid authority is handed over
Unenroll any time, no waiting window
Or hand over the hours
On campaigns you haven't enrolled, let the AI write the hour grid instead - a separate opt-in, per rule.
All 168 hours, authored for you
Built from your hourly Stream data
Rewritten every Sunday night
Its reasoning shown to you
Or keep painting the grid yourself
Configure & arm enrollment
These settings apply to the campaign you selected. Each can be reconfigured individually afterwards.
Norlend GC4 - 2026 - Research
Goal *
Launch
Brand-new product with little history. Bids to discover what converts without burning budget.
Expect: ACoS runs high while it learns.
Grow
Proven converter you want to scale. Pushes on winners and expands into new terms.
Expect: spend and sales climb together.
Optimize
Established product held to a specific ACoS. Trims the inefficient, holds the efficient.
Expect: converges to target over weeks.
Liquidate
Clearing stock fast. Bids for velocity, units sold over efficiency.
Expect: ACoS runs hot - that's the trade.
Wind down
Deliberately exiting. Steps bids down and stops chasing new terms.
Expect: a controlled taper, not a cliff.
TARGET ACOS (%) optional
25
Bid management will be AI-controlled. Budget stays under your manual control.
Shadow mode - recommended for testing
Shows you exactly what it would do, and makes no changes.
BackEnroll campaign
The line that matters: the AI never touches a rule you wrote yourself. It manages its own rules on campaigns you enrolled, and leaves yours alone - on enrolled campaigns they're set aside intact, everywhere else they run untouched.
COPILOTOPTIONAL AI ASSIST
Hand one campaign to the Autopilot engine.
Copilot is manual by default - every bid, every rule, your call. When you want a hand on a specific campaign, you enroll it. Enrolled campaigns run on the same engine that powers Autopilot, not a lesser version of it. Everything you don't enroll stays exactly as you built it.
Pick where it's headed
You choose a goal, not a formula - launching, growing, holding an ACoS, clearing stock, or winding down.
Five goals across a product's life
Optional target ACoS on top
Goal sets pace and step size
Opening judgment pass on day one
Reconfigure any campaign later
Reversible, precisely
Only your bid rules step aside while enrolled. Unenroll and they come back, armed the way you left them.
Your pause rules keep running
Your negation rules keep running
Your budget rules keep running
Only bid authority is handed over
Unenroll any time, no waiting window
Configure & arm enrollment
These settings apply to the campaign you selected. Each can be reconfigured individually afterwards.
Norlend GC4 - 2026 - Research
Goal *
Launch
Brand-new product with little history. Bids to discover what converts without burning budget.
Expect: ACoS runs high while it learns.
Grow
Proven converter you want to scale. Pushes on winners and expands into new terms.
Expect: spend and sales climb together.
Optimize
Established product held to a specific ACoS. Trims the inefficient, holds the efficient.
Expect: converges to target over weeks.
Liquidate
Clearing stock fast. Bids for velocity, units sold over efficiency.
Expect: ACoS runs hot - that's the trade.
Wind down
Deliberately exiting. Steps bids down and stops chasing new terms.
Expect: a controlled taper, not a cliff.
The line that matters: the AI never touches a rule you wrote yourself. It manages its own rules on campaigns you enrolled, and leaves yours alone - on enrolled campaigns they're set aside intact, everywhere else they run untouched.

Daily brief

Every morning, a jargon-free account summary

Copilot’s daily brief is written by Claude: a plain-English account-health summary — what moved, what’s working, what needs an eye — in words you’ll actually read, on your dashboard. It’s a read on your account, not a hand on the controls.

Copilot Brief MAR 4, 2026 View activity >
MARKET POSITION - MANAGED PRODUCTS
BELOW MARKET
Norlend GC4 converts at 17.1% where its market converts at 37.6% - measured over 7 weeks of search data.
The drop-off is on the listing page: 29% of your clicks add to cart, against 56% for the market. Once shoppers do add to cart you close at 58%, against 68%. Price is not the lever here - you sell at $18.99 against a market median of $27.50.
17 of your 99 clicks in this window fell on searches for a different market, mostly "coffee bean storage" and "coffee canister". Those are left out of the comparison above.
Amazon favours better-converting listings. Converting below your market makes organic rank harder to earn, and it means every paid click has to work harder to turn into a sale.
3 products can't be judged yet - 3 have no market data.
Yesterday, your 20 managed ad campaigns spent $24.09 and brought in $0 in sales - no orders came through, compared to a recent daily average of about $12 in sales and roughly one order per day. Spend was close to normal, but every dollar returned nothing, so it was an off day worth keeping an eye on.
Each click cost $4.01 against a recent norm of $3.63, and with only 6 clicks across 3,196 impressions there wasn't enough traffic to expect consistent sales. The automation made 621 adjustments - mostly scaling moves on placements that have performed well, plus two protective actions. No new terms harvested, no campaigns blocked, no budget changes needed.
Collapse brief >
Copilot Brief MAR 4, 2026 View activity >
MARKET POSITION - MANAGED PRODUCTS
BELOW MARKET
Norlend GC4 converts at 17.1% where its market converts at 37.6% - measured over 7 weeks of search data.
The drop-off is on the listing page: 29% of your clicks add to cart, against 56% for the market. Once shoppers do add to cart you close at 58%, against 68%. Price is not the lever here - you sell at $18.99 against a market median of $27.50.
17 of your 99 clicks in this window fell on searches for a different market, mostly "coffee bean storage" and "coffee canister". Those are left out of the comparison above.
Amazon favours better-converting listings. Converting below your market makes organic rank harder to earn, and it means every paid click has to work harder to turn into a sale.
3 products can't be judged yet - 3 have no market data.
Yesterday, your 20 managed ad campaigns spent $24.09 and brought in $0 in sales - no orders came through, compared to a recent daily average of about $12 in sales and roughly one order per day. Spend was close to normal, but every dollar returned nothing, so it was an off day worth keeping an eye on.
Each click cost $4.01 against a recent norm of $3.63, and with only 6 clicks across 3,196 impressions there wasn't enough traffic to expect consistent sales. The automation made 621 adjustments - mostly scaling moves on placements that have performed well, plus two protective actions. No new terms harvested, no campaigns blocked, no budget changes needed.
Collapse brief >
Copilot Brief MAR 4, 2026 View activity >
MARKET POSITION - MANAGED PRODUCTS
BELOW MARKET
Norlend GC4 converts at 17.1% where its market converts at 37.6% - measured over 7 weeks of search data.
The drop-off is on the listing page: 29% of your clicks add to cart, against 56% for the market. Once shoppers do add to cart you close at 58%, against 68%. Price is not the lever here - you sell at $18.99 against a market median of $27.50.
17 of your 99 clicks in this window fell on searches for a different market, mostly "coffee bean storage" and "coffee canister". Those are left out of the comparison above.
Amazon favours better-converting listings. Converting below your market makes organic rank harder to earn, and it means every paid click has to work harder to turn into a sale.

Bulk in a spreadsheet

Manage at scale in a spreadsheet

For agencies and power users, the biggest jobs don’t happen one row at a time. Download a template, work in the spreadsheet you already know, and upload to apply.

Build campaigns in Excel

Download the template, fill it in, and upload to create Sponsored Products campaigns, ad groups, keywords, targets, negatives, and tags in one shot.

Export, edit, re-upload

Pull your campaigns into a spreadsheet, change bids or add keywords in bulk, and upload to apply the changes.

Map campaigns for harvest

Define your source-to-destination harvest mappings in bulk from a spreadsheet, instead of wiring them one by one.

Uploads run in review-then-apply — you see the parsed result before anything goes live.

Manual control

Everything’s editable by hand, too

Rules are the engine, but you’re never forced through them. Every lever is one you can pull directly:

Bids — set, raise, or lower any keyword or target, one at a time or in bulk, with your own min/max clamps.

Keywords — add exact, phrase, or broad, at the starting bid you choose.

Negatives — block a wasting term in one campaign or fan it across many at once.

Placement — tune top-of-search, product-page, and rest-of-search modifiers by hand.

Search terms — promote a winner to a keyword or negate a loser, right from the report.

Rules for the work you want automated. Direct control for everything else.

Rank tracking

Watch your rank move, organic and sponsored

Track your keywords on any ASIN and see daily rank as a heatmap — organic and sponsored, side by side. Expand a tracked keyword to see which of your real customer search terms are feeding it. It’s the proof that your rule strategy is actually moving position, not just spending budget.

Rank by day — lower is better

Organic

Sponsored

Jul 11

Jul 11

Jul 12

Jul 12

Jul 13

Jul 13

Jul 14

Jul 14

Jul 15

Jul 15

Jul 16

Jul 16

Jul 17

Jul 17

stainless water bottle

stainless water bottle

41

33

27

19

14

9

6

insulated tumbler

insulated tumbler

52

44

31

23

16

11

7

glass food containers

glass food containers

38

29

22

17

12

8

4

bamboo cutting board

bamboo cutting board

47

40

34

26

20

13

9

reusable straws

reusable straws

33

26

21

15

10

7

5

silicone baking mat

silicone baking mat

58

49

37

28

18

12

8

Needs work

Climbing

Top ranked

Behind every rank

See every search building it — and the target winning it

Expand any tracked keyword to see the real customer searches building its rank — every query that contains every word of your keyword, in any order. For each one, you see which of your targets served the ad, the match type, and how it performed. Because exact matches aren’t the only sales that move rank: phrase, broad, and auto all count.

PPC Search Terms · Last 14 Days

toddler swim vest

Current rank

#6

Customer queries containing every word of this keyword, and the ads serving them.

Customer search term

Advertised target

Campaign

Match

Spend

Sales

ACoS

CVR

toddler

swim

vest

toddler swim vest

SP · Swim Vest

Exact

$4.80

$142

3.4%

83%

toddler

swim

vest

float

toddler swim vest

SP · Swim Vest

Phrase

$1.30

$54

2.4%

67%

best

toddler

swim

vest

2-3

toddler swim vest

SP · Swim Vest

Phrase

$0.90

$41

2.2%

50%

swim

vest

for

toddler

swim vest

SP Auto · Research

Broad

$2.10

$38

5.5%

33%

toddler

puddle

jumper

swim

vest

loose-match

SP Auto · Research

Auto

$0.75

$0.00

Matched words are highlighted. Every match type that contains your keyword’s words counts toward its rank — not just exact.

Deep enough to run client accounts

If you’re managing accounts for other people, Pilot gives you what a black box can’t: rules you can explain, an order you can defend, and rank movement you can screenshot into a report. Full manual control, optional AI help exactly where you want it, and nothing happening you didn’t set up.

Amazon PPC on autopilot. Every decision explained.

Amazon PPC on autopilot. Every decision explained.

Pulse Pilot automates your Amazon advertising — bids, dayparting, search-term harvesting, and budgets — and shows you the receipt for every dollar it moves.