A marketing analyst interview checks two things at once: can you get the numbers right, and can you explain them to a campaign manager who doesn't want to hear it. Expect a technical screen, a case built on messy campaign data, and at least one conversation with a marketer who'll be your main customer. Prepare stories where your analysis changed where money went.
| The question | What they want to know | Your angle |
|---|---|---|
| Google Ads says a campaign drove a lot of conversions, but the CRM shows far fewer new customers. How do you figure out what's going on? | Platform numbers never match the CRM, and this is the conversation you'll have most often in the job. They want to see whether you know why, not just that it happens. | Name the usual causes: different attribution windows, view-through conversions, and each platform counting a sale it only touched |
| Walk me through how you'd set up an A/B test for a new email subject line. | Testing is where analysts either earn trust or burn it. They're checking whether you plan the read before the send, not after. | Pick one primary metric up front, usually clicks or conversions rather than opens, since open tracking is unreliable |
| What attribution model would you use for a business with a long sales cycle, and why? | This separates people who've read about attribution from people who've had to defend a model to finance. There's no right answer, so they're watching your reasoning. | Say why last-click undercounts early channels like paid social and content when deals take a while to close |
| Write a query that shows cost per acquisition by channel for last month. | It's a basic SQL check wrapped in marketing vocabulary. They want clean joins and a clear definition of acquisition. | Ask what counts as an acquisition before writing anything: first purchase, signup or qualified lead |
| Tell me about a time your analysis changed how a marketing team spent money. | The job exists to move budget. They want proof you've done it, or come close. | Set up the situation briefly: the channel, the question, who owned the budget |
| Conversions dropped sharply overnight. What do you check first? | Tracking breaks happen all the time. They want someone who suspects the plumbing before blaming the campaign. | Check whether the drop shows up in every source or only one, such as Google Analytics but not the CRM |
| How would you decide which customer segments to target for a new product launch? | Segmentation is a core marketing analyst task, and it shows whether you think in terms of value or just demographics. | Start from existing customers who bought something similar and look at what they share |
| A campaign manager asks you to report only the metrics that make their campaign look good. What do you do? | This happens. They want to know you'll protect the numbers without making an enemy. | Acknowledge what the campaign did well, since there's usually something |
Picking one source as the truth right away, or saying the platform is just wrong without explaining how you'd check.
Name the usual causes: different attribution windows, view-through conversions, and each platform counting a sale it only touched. Check the conversion definition in the ad account against what the CRM counts as a new customer. Look for tracking breaks, such as a tag that fires twice or a form that doesn't pass UTM parameters. Say which number you'd report upward and how you'd label the gap so nobody is surprised.
Planning to check results every hour and stop as soon as one version looks ahead.
Pick one primary metric up front, usually clicks or conversions rather than opens, since open tracking is unreliable. Split the audience randomly and keep everything else the same. Decide the sample size and how long the test runs before it starts. Explain what you'd do if the result is flat: report it as flat, not as a small win.
HeroApply applies to Marketing Analyst jobs that match you, every day. 209 jobs are open today.