Microsoft Invest - Adaptive pacing
Manually managing the correct pace for a line item - determining how much to spend per hour and day, and implementing that with budgets - is error-prone, and can take an enormous amount of a trader's time. The first problem, determining how much to spend per day, is addressed by Lifetime Pacing. The second problem, determining how much to spend per hour and per minute within a day, is addressed by Adaptive Pacing.
Adaptive pacing controls a line item's pacing and prevents overpaying. Using the line item's daily budget (either generated by Microsoft Advertising automatically or manually entered by the user), Microsoft Advertising will calculate an ideal spend curve so that advertiser dollars are distributed evenly throughout the day. A dynamically calculated bid multiplier is applied to bids to control the amount of spend. If a line item is hitting its target delivery easily, the bid modifier will shade bids down, but only to the extent that delivery does not suffer. If a line item is not meeting its target delivery, adaptive pacing will shade bids less or not at all in order to meet target spend.
The spend curves are tailored to a country's internet traffic patterns. That is, instead of forcing line items to pace evenly throughout the day, Microsoft Advertising adjusts the spend so that more is spent during the hours when there are more users online. If a line item is targeting multiple countries, we will take a weighted average of the countries based on availability.
Adaptive pacing benefits line items in a few ways:
- It helps performance by spending less per impression.
- If you're getting paid on a fixed event basis, it helps line items retain more margin by spending less per impression.
- If you're a direct advertiser clients or you book on a Cost Plus basis, it helps advertiser dollars go farther. If the line item spends less per impression, those dollars can be used to buy more impressions.
Additionally, although adaptive pacing was developed to distribute spend, it is also used as a signal in many of our different optimization applications, such as Discovery.