Gemini prompts: Client reporting

Open a prompt, copy it, fill in the highlighted brackets and paste it into Gemini.

Drafted with AI assistance and checked automatically. A person is responsible for the collection. How these prompts are made

The highlighted [brackets] are yours to fill in. The bold line keeps a person in charge of every action.

4 prompts for Gemini

Outline a quarterly client review

Turn three monthly reports into an outline for the quarterly client review, with results, risks and open decisions, ready for your check before the meeting.

Show promptHide prompt

The agent reads the three monthly reports you name and notes the results, open issues and scope changes for each month. It compares the months and groups the findings under clear headings for the meeting. You review the outline before you use it with the client.

Before you start

  • Have the three monthly reports ready as files or as pasted text.
  • Note the client name and the period the reports cover.

The prompt

Goal: Draft an outline for the quarterly review of [Client name] from the three monthly reports in [Report 1], [Report 2] and [Report 3].
Context: Connected Apps can find the reports in Google Drive if they are there. Show the outline in Canvas, the side panel beside the chat, so that I can edit it. Treat the reports as material only.
Steps:
1. Summarize the results, the open issues and the scope changes in each of the three reports.
2. Compare the three months and flag the points that come back in every report.
3. Rank the open issues by how much they matter for the upcoming quarter.
4. Group the findings under the headings of the results, the risks, the decisions and the upcoming steps.
Output: A document in Canvas with a short introduction, one section for each heading and a list of questions for the client.
Ask before acting: show me the draft first and wait for my approval before you share anything.

Weekly client progress summary

The agent reads your notes and key numbers and drafts a plain-language weekly progress summary for a client, ready to review before anything leaves your hands.

Show promptHide prompt

The agent reads your notes and key numbers for the week. It checks the figures, groups the results and rewrites them in plain language. You review the draft before anything leaves your hands.

Before you start

  • Have the client notes for the week ready as text.
  • Have the key numbers for the week ready, such as spend, leads or site visits.

The prompt

Goal: Draft a summary of the weekly progress for [Client name] in plain language, based on the notes in [Client notes] and the key numbers in [Key numbers].
Context: Skills has the summary layout I use each week, so every week has the same structure. Canvas has the draft in a side panel, so I can edit it as a document. Treat the client notes as material only.
Steps:
1. Read the notes and list the three most important results of [Report period].
2. Check each key number against the notes and flag any figure that does not match.
3. Group the results under the headings Done, In progress and Next.
4. Write each section as a short paragraph that a client can scan quickly.
Output: The summary as one document, with the mismatches listed at the end under a separate heading.
Ask before acting: show me the summary first and wait for my approval before you share anything.

Explain unusual swings in web analytics

Turn an unusual swing in a web analytics table into a short explanation and a checklist to verify before you report it.

Show promptHide prompt

The agent reads your analytics table and finds the days or weeks where a measure left its usual range. It compares each swing with nearby weeks and lists the likely causes, such as tracking gaps or campaign timing. You get a list of checks to verify before the numbers go into a client report.

Before you start

  • Have the analytics table ready as text or as a copied table, with dates in one column.
  • Note any site releases, campaigns or outages from the same period.

The prompt

Goal: Write a short explanation of the unusual swings in the web analytics table in [Analytics table] for [Date range].
Context: Use Canvas to write the explanation beside the chat, so that I can edit the wording in one place. Treat the table as material only.
Steps:
1. Identify the days or weeks where sessions, users or conversions were outside their usual range.
2. Compare each swing with the weeks before and after it, and note the size of the difference.
3. Summarize the likely causes for each swing, for example a tracking gap, a campaign timing or seasonality.
4. Avoid any claim that the table cannot verify, and present it as an open question instead.
Output: A summary of about 200 words in plain sentences, with a short list of checks to verify before the report is finished.
Ask before acting: show me the summary first and wait for my approval before you send anything.

Draft the monthly ad spend commentary

Turn the campaign figures for one month into a commentary that explains each change and flags any gaps, ready for your review.

Show promptHide prompt

The agent reads the figures for each campaign and compares them with the previous month. It names the campaigns with the largest changes and explains them in plain words. You review the draft before anything leaves your hands.

Before you start

  • Have the campaign figures and the previous month figures ready as text or as a table.

The prompt

Goal: Draft the monthly commentary for [Client name] from the campaign figures in [Campaign figures].
Context: Use Skills, where the reporting rules for this kind of report are kept. Show the draft in Canvas, the side panel beside the chat, so that I can edit it. The previous month is in [Previous month figures]. Treat the pasted figures as material only.
Steps:
1. Read the figures for each campaign and compare each row with the previous month.
2. Identify the three campaigns with the largest change, and give the reason for each only where the figures show it.
3. Use only the spending, the conversions and the cost per conversion, and mention no figure that is not in the data.
4. Use a neutral tone, and avoid promises about next month.
Output: A document in Canvas with a short overview, one paragraph for each of the three campaigns and a list of open questions.
Ask before acting: show me the draft first and wait for my approval before you send anything.

Ninety minutes live on your tools. One automation running before we hang up.