Above Fold Lab
Above Fold Lab
@AboveFoldLab

Deep dive: a protocol for choosing the one hero image that earns its weight

The hero image is the heaviest single asset on most landers and the most thoughtlessly chosen — a stock handshake, a generic dashboard, a smiling model. Two research threads should govern the choice. First, gaze studies show people look where faces in an image look: an eye-tracking finding (the often-cited baby-direction-of-gaze study) is that a face looking at your headline pulls the viewer's gaze there too. Second, authenticity research suggests genuine, specific imagery outperforms polished stock for trust. The image is a persuasion tool, not decoration.

The selection protocol:

— Step 1: Define the image's job before browsing options. Is it proving the product is real (show the actual product/UI), humanizing the brand (show real people), or directing gaze (use a face looking toward the call-to-action)? One job, chosen first.
— Step 2: Reject generic stock by default. A photo a hundred competitors also use signals "template," and template signals "low effort," which leaks into perceived product quality.
— Step 3: If using people, use real and specific ones. Authentic faces — real customers, real team — read as more trustworthy than airbrushed models, and they avoid the uncanny "too perfect" tell that triggers skepticism.
— Step 4: Exploit gaze direction. If a person is in the frame, orient their look toward your headline or button, not out at the viewer. You're borrowing their gaze to steer the visitor's.
— Step 5: Pay the weight cost only if the image does a job. A heavy hero that merely decorates is a performance tax with no conversion return — replace it with a lighter asset or none.
— Step 6: Test image versus no image. On some landers, removing the hero and giving the copy room lifts conversion; assumption, then measurement.

Mechanism: images are processed faster than text and set an emotional and credibility frame before a word is read. The right image accelerates comprehension and trust; the wrong one contradicts the copy or wastes the eye's first, most valuable fixation.

The caveat: gaze-direction effects are real but modest and easily swamped by stronger hierarchy cues like contrast and motion. Don't over-engineer the photo at the expense of basic clarity — the image supports the message, it doesn't replace it.

TL;DR
— Assign the hero image one job (prove / humanize / direct gaze) before choosing it.
— Reject generic stock; use real specific faces and point their gaze toward your call-to-action.
— Only pay the image's weight cost if it does a job — test image vs none.
This post was published in the Telegram channel Above Fold Lab. Subscribe here: @AboveFoldLab.
tech

Latest posts in “Tools and infrastructure”

All channels in this category →

start

Готовы запустить рекламу через сеть public.tg?

Новый оффер, продукт, GEO, кейс, событие или партнёрский запуск — соберём маршрут под задачу и отдадим медиаплан.

Telegram для медиаплана: @AFFtop_connect. Быстрый тест: $20 за канал, $1000 за пакет по сети.