Monday Amazon Ads review: 10 prompts for Claude or ChatGPT

What should you ask your Amazon Ads data on a Monday morning?

Ask ten questions, in this order: how the week went, why anything moved, where spend leaked, how it is split, and who touched what. Xnurta MCP connects Claude, ChatGPT and other MCP clients to your Xnurta data, so each question runs against your ad accounts instead of an export you pulled on Friday. Copy a prompt, replace the [placeholders] with your own ad account, dates and thresholds, and paste it into your assistant. If you have not connected yet, Connect Claude or ChatGPT to Xnurta MCP takes a few steps.

Do these prompts change anything in your account?

No. All ten are read only. Queries on ad performance, entity settings and the operation log never modify your account, so you need no write access to run any prompt here. Where a change would be tempting, such as raising a budget or negating a term, the prompt says to list only.

Writing comes in three kinds, each with its own permission scope. When a Monday answer does tell you to act, anything involving money, bids or archiving shows a preview first: which objects, and from which value to which value. Nothing is submitted until you confirm. Archiving cannot be undone, so pause a campaign if you only want it to stop.

What goes into a good Monday prompt?

Four things: the ad account, a time window, a comparison and the shape of the answer. Leave one out and the assistant guesses. Name the ad account in the prompt, and if you are unsure what you have, start by asking it to list the ad accounts you are authorized to access. You see only the accounts your own Xnurta login carries, main accounts and sub-accounts included.

How did last week go across your ad accounts?

Start wide, then narrow. These three prompts give you the week at three zoom levels: every ad account, one account over 30 days, and one account in a fixed report format.

Prompt 3 uses the Weekly Ads Report skill, which produces week-over-week KPIs, 7-day trends, anomalies, AI and automation activity, top movers and action suggestions. Skills teach the assistant the routine for a task, and you install them in a separate step after connecting.

List the ad accounts I am authorized to access. For each one, compare last week (Monday to Sunday) with the week before: spend, sales, ACOS and ROAS. Return a table with one row per ad account, then name the five issues that need my attention first and back each one with the numbers.

For ad account [account name], analyze the last 30 days: trends in spend, sales, ACOS, ROAS, CTR and CPC. Point out the key turning points and the likely cause of each. Return a table by week, then three sentences on what changed.

Use the Weekly Ads Report skill for ad account [account name]. Cover last week (Monday to Sunday) against the week before: KPI changes, the 7-day trend, anomalies, AI and automation activity, top movers and suggested actions for this week. Keep it to one page.

Why did a campaign's ACOS move, and what hit its budget?

Point the assistant at the campaign that looks wrong and ask for causes across budget, bids, targeting and structure. A vague "why is ACOS up" gets a vague answer. A named campaign, a seven-day window and the before and after values get a ranked one.

Prompt 5 lists the campaigns that ran into their daily budget cap and stops there. Raising a budget is a separate decision, and a separate step.

In ad account [account name], campaign [campaign name or ID] saw ACOS move from [x]% to [y]% over the last 7 days. Check budget, bids, targeting and structure for the cause. Return the three most likely causes, ranked, each with the data behind it. Do not change anything.

In ad account [account name], which campaigns hit their daily budget cap in the last 7 days? Return a table with campaign name, spend, sales and ACOS, sorted by sales. List only. Do not change any budget.

Which search terms need a decision this week?

Three prompts sort your search terms into push, stop and double-check. The first finds winners and losers. The second finds terms that spent without a single order and still have no negative. The third catches the opposite mistake: a term that earned orders and got negated anyway.

Every one of them returns a list. You decide what happens to each term, and you can ask the assistant for a follow-up on any single row before you touch it.

For ad account [account name], which search terms performed well over the last 30 days? Show CTR, CVR, ACOS and orders. Which terms deserve more investment, and which high-ACOS, low-conversion terms look like negation candidates? Return two tables, winners and candidates. Make no changes.

In ad account [account name], over the last [14] days, list search terms that spent more than [amount] with no orders and have not been negated. Return a table with term, campaign, spend and clicks, biggest spender first. List only.

In ad account [account name], over the last [30] days, find search terms that spent between [low] and [high], generated orders, and were negated. These may be incorrect negatives. Return a table with term, campaign, orders, spend and ACOS. List only.

Where is spend going, and who changed what?

Compare spend share and ACOS by ad type and match type, then read the operation log for the week. The first shows where budget sits. The second shows who moved it, which matters on a Monday when a number shifted over the weekend and nobody owns up.

The operation log holds human and AI actions on campaigns, ad groups, targets, bids, budgets and managed groups, and you can filter by operator: AI, human, automation rule or one person's email. Prompt 10 splits the week by operator. For a single campaign, Who changed my Amazon campaign? goes deeper.

For ad account [account name], compare spend share and ACOS across SP, SB and SD, and across match types, for the last 30 days. Return one table by ad type and one by match type, then say where budget looks misplaced. Recommendation only.

In ad account [account name], list every bid change, budget change and negative-targeting action from the last 7 days, split by operator: AI, human and automation rule. Return a table with operator, action type, campaign and time, newest first, then a two-line summary per operator.

How do you turn the review into a habit?

Save the prompts you reuse and ask follow-ups in the same chat. Start broad, then ask why, then drill into one object. That beats cramming everything into one giant question. Xnurta MCP lets you keep frequent prompts as templates or turn one into a weekly routine that runs every Monday.

Two more habits help. When a product is sliding, the Product Diagnosis skill ranks ASINs and parent ASINs into tiers and flags declining products. And mind the clock: data follows the Xnurta platform's update cadence and is not real time, so a Monday run shows data as of the platform's last update. Daily data is the default, with hourly available for SP, SB and SD campaigns and SP keywords. For more everyday prompts beyond this set, see Top 10 Xnurta MCP prompts and the data each needs. To decide how much access to switch on later, read Read-only vs read-write MCP for Amazon advertisers.