The Best Claude Prompts for Campaign Performance Reports

If you've ever spent hours transforming raw campaign data into a readable and compelling report, you know how frustrating it can be. Scattered numbers, metrics to interpret, stakeholders to satisfy — and all on a tight deadline. Claude can completely change this scenario, but only if you know how to speak to it the right way.

Claude is not simply a text generator: it's an extraordinarily capable analysis tool when you provide it with the right context. Unlike other models, Claude excels at structured reasoning and data interpretation, making it particularly suited for transforming campaign numbers into strategic narrative. I've tested hundreds of prompts on this topic and the difference between a generic prompt and a well-constructed one is the difference between a useless report and one your CMO wants to print and hang in the office.

In this article I'll show you 10 tested, copy-ready prompts for every stage of the performance report — from initial data analysis to final strategic recommendations. You'll also learn the prompt engineering principles behind each one, so you can adapt them to your specific situation.

The Best Prompts for Campaign Reports with Claude

Prompt 1: Initial Raw Data Analysis

You are a senior marketing analyst with 10 years of experience. I'm providing you with raw data from my [CAMPAIGN NAME] campaign for the period [DATES]. The total budget was [AMOUNT].

Here's the data:
- Impressions: [X]
- Clicks: [X]
- CTR: [X%]
- Conversions: [X]
- CPA: [X€]
- ROAS: [X]
- Total cost: [X€]

Please:
1. Identify the 3 most important insights from this data
2. Flag any anomalies or unusual trends
3. Compare performance against [INDUSTRY] benchmarks if you know them
4. Provide an overall judgment: did the campaign perform well, average, or below expectations?

Use a professional but clear tone, as if you were explaining to a non-technical marketing director.

Why this prompt works: It assigns Claude a precise role ("senior marketing analyst"), provides structured data, and requests numbered output. The specification of final tone is crucial — it prevents Claude from producing overly technical jargon or, conversely, being too vague. The placeholder [INDUSTRY] activates the model's contextual knowledge about relevant benchmarks.

Expected output: A summary in 4 well-defined sections, with qualitative interpretations of the numbers and a clear final verdict. Claude typically produces approximately 400-600 words of dense but readable analysis.

Personalization tip: If you have data from multiple channels (Meta, Google, LinkedIn), add a line per channel and ask Claude to do cross-channel comparison. The prompt scales very well.

Prompt 2: Creating the Executive Report Structure

I need to present the results of the [CAMPAIGN NAME] campaign to the board/senior executives. The campaign's primary objective was [OBJECTIVE: e.g., lead generation, brand awareness, direct sales].

Create a detailed outline for a maximum 2-page executive report that includes:
- An executive summary of 3-4 sentences
- Key metrics to highlight (choose the most relevant for the stated objective)
- A "What Worked / What Didn't" section
- Recommendations for the next campaign

The audience is not a digital marketing expert, so avoid unexplained acronyms and use concrete analogies when helpful.

Why this prompt works: It specifies the target audience (non-technical executives), constrains length (2 pages), and defines the campaign objective — which radically changes which metrics make sense to highlight. A report for a brand awareness campaign should focus on reach and frequency, not CPA.

Expected output: An outline with headings, subheadings, and guidance notes on what to include in each section. It's a perfect skeleton to fill in with your actual data.

Personalization tip: Replace "board" with your actual audience — it completely changes the register. If you're presenting to your team, you can request a more technical and interactive format.

Prompt 3: Interpreting Negative Results

My [CAMPAIGN NAME] campaign underperformed against objectives. Here are the targets vs. actual results:

- Target CTR: [X%] → Result: [Y%]
- Target CPA: [X€] → Result: [Y€]
- Target conversions: [X] → Result: [Y]
- Budget spent: [X€] out of [Y€] available

Help me to:
1. Identify possible causes of this underperformance (list at least 5 plausible hypotheses based on this data)
2. Distinguish between controllable factors (we can improve) and external factors (market, seasonality, etc.)
3. Suggest how to present these results to stakeholders in an honest but constructive way, without minimizing problems but highlighting improvement opportunities

I don't want you to "narratively fix" the numbers — I want strategic honesty.

Why this prompt works: The final phrase — "I don't want you to narratively fix the numbers" — is essential. Claude tends to be diplomatic by default; this explicit instruction pushes it toward analytical honesty. Separating controllable from external causes is a classic performance marketing framework that Claude applies very well.

Expected output: A list of diagnostic hypotheses (very useful for internal post-mortems) and a guide on how to construct the report narrative even when results aren't stellar.

Personalization tip: Add market context details (e.g., "we were in holiday season" or "a competitor launched an aggressive promotion") for even more precise hypotheses.

Prompt 4: Generating Strategic Recommendations

Based on the following results from the [CAMPAIGN NAME] campaign for the period [DATES]:

[PASTE HERE A SUMMARY OF THE MAIN RESULTS IN 5-10 POINTS]

Generate a recommendation plan for the next campaign, structured as follows:
- **Immediate actions** (to implement in the next campaign, within 2 weeks): maximum 3 points
- **Medium-term optimizations** (to test within 30-60 days): maximum 3 points
- **Strategic hypotheses to test** (recommended A/B tests, new channels to explore): maximum 3 points

For each recommendation specify:
a) What to do exactly
b) Why (reasoning based on data)
c) How to measure success (specific KPI)

Available budget for next campaign: [AMOUNT]. Consider this in the practicality of recommendations.

Why this prompt works: The three-tier temporal structure (immediate / medium-term / hypotheses) is derived from OKR and sprint planning frameworks. Claude responds very well to explicit budget constraints — without them, recommendations tend to be ambitious but unrealistic. Asking for KPIs for each recommendation forces actionable output instead of vague advice.

Expected output: An action plan in 9 points (3 per category), each with logic and success metrics. Often usable almost directly in a presentation.

Personalization tip: If you work at an agency, add "keep in mind that the client is in [INDUSTRY] and has low risk tolerance" to calibrate the boldness of recommendations.

Prompt 5: Writing the Executive Summary

Write an executive summary for the [CAMPAIGN NAME] campaign report. It must be:
- Exactly 150-200 words long
- Written so a senior executive reads it in 60 seconds and immediately understands whether the campaign was successful
- Structured in 3 paragraphs: (1) Context and objectives, (2) Main results, (3) Implications and next steps

Here are the key data points:
- Campaign objective: [DESCRIPTION]
- Budget: [X€]
- Main result: [KEY METRIC AND VALUE]
- ROI/ROAS: [X]
- Comparison to previous campaign: [BETTER/WORSE/SAME] by [Y%]

Use a professional and direct tone. Start with results, not context — executives want to know immediately "did it go well or not".

Why this prompt works: Constraining length to 150-200 words forces Claude to be precise and concise. The instruction "start with results" reflects journalistic technique (the inverted pyramid) that works exceptionally well in business reports too. Often reports start with three pages of context before getting to the point — this prompt deliberately inverts that structure.

Expected output: Text ready to copy directly into the report, requiring virtually no editing. Claude respects word-count constraints very well when expressed numerically.

Personalization tip: If you want a more enthusiastic tone (for good news) or more cautious (for mixed results), add it explicitly. Claude interprets "professional and direct" in a fairly neutral way by default.

Prompt 6: Multi-Campaign Comparative Analysis

I have data from 3 different campaigns that I want to compare to understand which performed best in terms of efficiency. Here's the data:

**Campaign A ([NAME]):**
- Budget: [X€] | Conversions: [X] | CPA: [X€] | ROAS: [X]

**Campaign B ([NAME]):**
- Budget: [X€] | Conversions: [X] | CPA: [X€] | ROAS: [X]

**Campaign C ([NAME]):**
- Budget: [X€] | Conversions: [X] | CPA: [X€] | ROAS: [X]

Please:
1. Create a comparative table with a rating for each metric (★ = underperformance, ★★★ = average, ★★★★★ = excellent)
2. Identify which campaign is most efficient considering cost/result ratio
3. Suggest where to concentrate future budget and why
4. Warn if there are missing data points that could change the analysis

Note: the campaigns had different objectives — A was brand awareness, B was lead gen, C was direct conversion. Factor this into the comparison.

Why this prompt works: Comparing campaigns with different objectives is one of the classic reporting mistakes — this prompt anticipates it explicitly. The star-rating system is a visual heuristic that Claude applies consistently and that makes the report immediately readable. The request to "warn if there are missing data points" activates the model's critical thinking instead of simple completion.

Expected output: A well-formatted markdown table + narrative analysis. Perfect to copy into Notion, Confluence, or directly into a PowerPoint presentation.

Personalization tip: Add the channel (Meta, Google Ads, email) for each campaign if you want cross-channel analysis as well.

Prompt 7: Translating Data into Non-Technical Stakeholder Storytelling

I need to present these results to stakeholders who don't understand digital marketing (e.g., CFO, CEO, legal team):

[PASTE YOUR DATA OR MAIN RESULTS HERE]

My goal is for them to understand:
1. How much value the campaign generated in business terms (not marketing metrics)
2. Whether we're using the budget intelligently
3. What will happen if we invest more (or less)

Transform this data into:
- 3 opening sentences that capture attention without using terms like CTR, CPM, ROAS
- A concrete analogy that explains campaign efficiency (e.g., "for every dollar invested, we got X back")
- A single slide title that summarizes everything in 10 words
- The 2 questions they'll probably ask and recommended answers

Why this prompt works: Asking Claude to anticipate stakeholder questions is an advanced technique that produces extraordinary output — it's like having a presentation coach doing role-play with you. The request to avoid technical acronyms isn't a restriction, it's liberation: it forces Claude to find the core of value instead of hiding it behind jargon.

Expected output: Material nearly ready for a C-suite presentation. Particularly useful for agency professionals who need to "translate" marketing for non-technical clients.

Personalization tip: Specify the company's industry and stakeholders' roles — a startup CFO and a bank CFO respond to very different framing.

Prompt 8: Identifying Mid-Campaign Optimization Opportunities

My [CAMPAIGN NAME] campaign has been running for [X] days out of [Y] total planned days. Here's the current data:

- Budget spent so far: [X€] out of [Y€] total
- Spending pace: [ahead / behind / on track]
- Conversions obtained: [X] against target [Y]
- Best-performing channel: [NAME] with CPA of [X€]

Why this prompt works: This prompt activates Claude's real-time optimization thinking. Rather than waiting for a campaign to end to analyze it, mid-campaign analysis lets you course-correct while there's still budget and time remaining. The structured format makes it easy to extract actionable recommendations.

Expected output: Immediate optimization recommendations that can be implemented within hours or days, with projected impact on remaining budget.

Personalization tip: Include data from your ad platform's forecasting tool if available — Claude can incorporate predicted trends into its recommendations.