Bidstream Lab
Bidstream Lab
@BidstreamLab

Retargeting vs prospecting: the same auction, two opposite bidding postures

Retargeting vs prospecting: the same auction, two opposite bidding postures

Mixing one bid strategy across retargeting and prospecting is the most common way to overpay. They have different value certainty, so they need different tools.

1. Retargeting: you know the user (they visited, added to cart). Value-per-impression is high and known, the audience pool is small, and you are often bidding against everyone else who also retargets that user. Posture: bid high, shade conservatively, accept a higher win-cost because losing the impression loses a near-converter.

2. Prospecting: value-per-impression is uncertain and spread thin across a huge pool. Posture: bid moderate, shade aggressively, let volume and learning do the work because any single impression is cheap to lose.

3. The shared error: one global bid-shading setting and one bid cap across both. Aggressive shading that is correct for prospecting starves retargeting of must-win impressions; the high cap correct for retargeting overpays on prospecting.

4. The structural tell: retargeting win-rate is usually lower (more competition for known users) at a higher cost; prospecting is higher win-rate at lower cost. If they look similar, your settings are blurred across both.

Why it matters: a blended campaign report averages two opposite economies into one meaningless number. Split win-rate, shade margin, and cost per win by audience type in log-level data. The line items should look different — if they do not, you are running prospecting math on your most valuable users.
Этот пост опубликован в Telegram-канале Bidstream Lab. Подписаться можно по ссылке: @BidstreamLab.
traffic

Свежие посты в категории «Traffic Sources»

Все каналы категории →

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.