What AI Can (and Can't) Do for Marketing Analysts | Jobsolv AI Academy

AI won't replace marketing analysts — but analysts who use AI will replace those who don't. Learn what AI tools actually excel at, where they fall short, and how to think about AI as your analytics coworker.

If you've been hearing about AI tools like ChatGPT, Claude, and Gemini but aren't sure how they fit into your day-to-day analytics work — you're in the right place. This lesson cuts through the hype and shows you exactly what AI can do for marketing analysts today.

What AI Actually Excels At

AI tools are incredibly good at tasks that are repetitive, pattern-based, or involve transforming data from one format to another. Here's where they shine for analysts:

Data cleaning — fixing messy CSVs, standardizing formats, handling missing values
Writing SQL queries — from simple selects to complex joins and window functions
Summarizing large datasets — finding patterns humans would miss in thousands of rows
Building formulas — Excel, Google Sheets, even Python scripts for automation
Explaining code — understanding what existing queries or scripts actually do

Manual Workflow

1. Open CSV in Excel 2. Manually scan for blank cells 3. Find-and-replace inconsistent formats 4. Create VLOOKUP to merge two sheets 5. Build pivot table 6. Format and email to stakeholder Time: ~45 minutes

With AI

1. Upload CSV to Claude 2. Ask: "Clean this data, standardize the date formats, merge with this second file, and summarize the top trends" 3. Review the output 4. Ask for adjustments Time: ~5 minutes

Time saved: 45 min → 5 min

Where AI Falls Short

AI is a powerful tool, but it's not a replacement for your analytical judgment. Here's where it consistently struggles:

Business context — AI doesn't know why your Q3 numbers dropped or that your CEO cares about LTV over CAC
Strategic recommendations — it can analyze data but can't tell you what your team should actually do
Data accuracy verification — it can produce convincing but wrong numbers if you don't check

Watch Out

Never present AI-generated numbers to stakeholders without verifying them first. AI can hallucinate statistics, invent trends, and produce plausible-looking but completely wrong calculations. Always spot-check the output against your source data.

The Right Mental Model

Think of AI as a very fast, very capable junior analyst who just joined your team. They can execute tasks incredibly quickly, but they need clear direction and their work needs to be reviewed. You bring the context, the strategy, and the judgment. AI brings the speed and the ability to never get tired of repetitive work.

Pro Tip

The analysts who get the most value from AI aren't the ones who ask it to do everything. They're the ones who know exactly which parts of their workflow to delegate and which parts require human judgment.

See It In Action

Here's a real example of using AI to write a SQL query you might need for a marketing attribution report:

Prompt Example

any

Ask any AI tool to write this query for you:

Write a SQL query that shows marketing channel attribution for the last 30 days. I need: channel name, number of conversions, total revenue, and cost per acquisition. The data is in tables called 'events' (with columns: event_date, channel, event_type, revenue) and 'ad_spend' (with columns: date, channel, spend). Group by channel and sort by revenue descending.

AI Generatedclaude

SELECT e.channel, COUNT(CASE WHEN e.event_type = 'conversion' THEN 1 END) AS conversions, SUM(CASE WHEN e.event_type = 'conversion' THEN e.revenue ELSE 0 END) AS total_revenue, COALESCE(SUM(a.spend), 0) / NULLIF(COUNT(CASE WHEN e.event_type = 'conversion' THEN 1 END), 0) AS cost_per_acquisition FROM events e LEFT JOIN ad_spend a ON e.channel = a.channel AND e.event_date = a.date WHERE e.event_date >= CURRENT_DATE - INTERVAL '30 days' GROUP BY e.channel ORDER BY total_revenue DESC;

That query would take most analysts 10-15 minutes to write and debug. With AI, you describe what you need in plain English and get working SQL in seconds. But notice — you still need to know what question to ask and whether the output makes sense for your data.

Try It Yourself

Open ChatGPT or Claude and paste this prompt. Replace the example with your actual data structure if you have one handy.

I'm a marketing analyst. I have a Google Sheet with columns: Date, Campaign Name, Channel (paid search, social, email, organic), Sessions, Conversions, and Revenue. Can you help me:
1. Write a formula to calculate conversion rate per channel
2. Suggest which channels I should investigate for optimization
3. Create a summary I could paste into a Slack message for my team

Key Takeaways

AI excels at repetitive, pattern-based tasks: data cleaning, SQL, formulas, summarization
AI struggles with business context, strategic judgment, and accuracy guarantees
Think of AI as a fast junior analyst — give clear direction, always review the output
The biggest wins come from knowing which parts of your workflow to delegate

In the next lesson, we'll look at the specific AI tools available and help you choose the right one for your workflow.

What AI Can (and Can't) Do for Analysts