A Homepage Ad Slot Experiment: One Placement at a Time

I run a small site and I’ve been treating the homepage like a lab bench. One change at a time, written down, so I can actually tell what moved.

Start with a slot map. I sketch the homepage in a notebook — header, first screen (above fold), mid-scroll, footer — and label each box A, B, C. Then I list which ad unit sits where and what size it is. It sounds tedious, but you can’t compare placements you can’t name.

Above fold is the tempting spot. It gets the most impressions, and it also competes with the headline people came for. In my first test I put a single 300x250 just under the intro paragraph. Traffic on the page went up in the reports, but time on page dropped noticeably. So I moved the same unit below the first section, and the drop mostly went away.

Report segments matter more than totals. I break the homepage numbers by device, new vs returning, and traffic source. In my case, mobile made up most of the visits, and the above-fold unit pushed the first paragraph off the screen entirely on a phone. That’s a layout problem, not a revenue problem.

Practical checks I do on mobile:

• Load the homepage on an actual phone, not a resized browser window
• Scroll the first screen and note what you see before any scrolling
• Check that tapping the menu isn’t crowded by the ad
• Confirm one ad doesn’t push the first heading below the fold

I also keep a short homepage ad placement log — dates, slot, what changed, and the two or three numbers I care about. Writing it down once a month is enough for me; more often and I start chasing noise.

Monthly check: pick one slot, change one thing, wait, compare the same segment. If reports mix mobile and desktop, a bad mobile layout can look like a bad ad unit. Check the provider’s own current terms before assuming anything about limits or payouts — mine changed once without much notice.

— Daniel Hayes