Digital Marketing Manager Interview Questions

First I separate channels by intent — branded vs non-brand search, prospecting vs retargeting. I'd reallocate from the worst-ROAS prospecting cohorts into mid-funnel (retargeting, lookalikes off high-LTV seeds) and creative testing, then fix the leakiest step in the funnel (usually landing-page conversion) before buying more traffic. I cut CAC 31% YoY this way while growing volume 2.4× — most of the win was creative velocity + a landing-page rebuild, not bid tweaks.

By marginal ROI and payback, not vanity reach. Paid funds predictable near-term volume; organic + lifecycle compound and lower blended CAC over time. I anchor on a target blended CAC:LTV and a payback window, then shift spend monthly toward whatever's beating the threshold. Lifecycle almost always wins on ROI once you have a base.

I stop arguing about last-click. I pick one source of truth for spend decisions (usually a media-mix or data-driven model), use platform numbers only for in-platform optimization, and validate with geo-holdout or incrementality tests on the big channels. I consolidated three conflicting tools into one pipeline and ended nine months of 'whose number is right.'

S: A product-launch push had strong CTR but poor conversion. T: Diagnose and fix mid-flight. A: I traced it to a message-match gap — the ad promised a use case the landing page buried. I rewrote the LP hero to mirror the top ad, added a sharper CTA, and paused the two worst creatives. R: Conversion rate recovered ~40% within the week and the campaign hit plan.

I don't force last-click on it. I track leading indicators (branded-search volume, direct traffic, assisted conversions, organic pipeline) and run periodic holdouts or geo-lift to estimate incremental impact. The honest answer is brand is measured in trend + incrementality, not a single ROAS number.

Authenticate (SPF/DKIM/DMARC), warm new domains, segment by engagement, and sunset cold contacts before they tank your sender reputation. I watch inbox-placement and complaint rates, not just open rate (which is noisy post-MPP). On a migration I lifted subscription-to-active conversion 18% mostly by suppressing dead segments.

One variable, a pre-registered hypothesis and primary metric, enough sample for power, and I don't peek-and-stop. I prioritize tests by expected impact × ease, and I kill the loser fast. Most of my wins came from creative and offer tests, not button colors.

Weeks 1–3: audit the funnel end to end and find the biggest leak. Weeks 4–8: fix it (usually conversion or lifecycle, not top-of-funnel) and stand up clean attribution. Weeks 9–12: scale the channels that survive a payback test and kill the rest. I lead with the leak, not the spend.