Automated bid strategies optimize your line item's bid based on the likelihood that:
- An impression will lead to a chosen outcome (a click, conversion, view, etc.).
- An impression is available at an optimal price (compared to the market).
Some variables that automated bid strategies use for prediction include inventory signals, audience lists, creative effectiveness, and line item strategies.
Use automated bidding
- New line items are automatically defaulted to the appropriate automated bid strategy.
- Go to Line item > Details to enable or disable automated bidding.
- To optimize toward conversions (without historical data in Display & Video 360), opt into the enhanced automation pool.
Insights about bid strategy performance
View bidding insights
- Navigate to your campaign's Combined view.
- Select Filter > Bidding insights to view insights based on availability.
- Hover over the intelligence icon
to view bidding insights.
- Choose View bidding insights to see details and additional insights about the top:
- Websites
- Exchanges
- Devices
- Browsers
- Day of the week
- Time of day
- Environment
to view bidding insights.Automated bid strategies optimize toward your goal even when insights aren't available.
Interpret bidding insights
- Bidding insights give an in-depth look at how your automated bidding strategy performed. Because of this, they can be difficult to interpret.
- Generally, impressions should be bid on by likeliness to convert to your goal (conversion rate) at an optimized cost (below market price), and then successfully won (win rate).
| Insights chart name | What to expect* | Example image |
|---|---|---|
| Calibrated conversion rate | The automated strategy performed well when a staircase-like graph is shown. "Most likely" to convert impressions is greater than "Likely," which is greater than "Unlikely," which is greater than "Least likely." |  |
| Average cost and market price | The automated strategy performed well when the best impressions (most likely to convert) were bought at a fair price relative to the market. Naturally, impressions with a higher likelihood of converting tend to cost more than those with a lower likelihood. "Market price" is the average highest bid for a line item's impressions and isn't available for all exchanges. |  |
| Win Rate | The automated strategy performed well when impressions were won successfully. Win rate should reflect the first two charts, but with more detail on the dynamics between cost and conversion likeliness. For example, "Most likely" to convert impressions might be lost due to high market price, or "Less" likely may be won because they were a bargain. |  |
* Insights might not match the examples depending on campaign settings, unique inventory, and market competition.
Take action
Settings changes aren't needed based on bidding insights. However advanced users may find them helpful for advanced troubleshooting.
Types of automated bid strategies
Maximize performance
-
Maximize {KPI} while prioritizing spending my full budget optimizes for the best performance for your chosen KPI. Flight dates must last 2 weeks for the entire budget to be spent. Your line item's bid will be raised until your budget can be spent or the line item's bid is capped by the Do not exceed average CPM of setting.
-
Maximize clicks/Maximize conversions, while prioritizing a target {KPI} optimizes for a specific performance goal. Meeting the line item's performance goal will take priority over spending your line item's budget. If your line item's goal is overly aggressive, the line item budget might not be spent.
-
(App install line items only) Maximize installs optimizes for app install performance for your budget, or for a target cost per install. You can also limit the average CPM (see below).
Considerations
- Don't use this setting with impression-based budgets. Because they aren't based on currency, they may lead to overspending.
- Don't use this setting if you must stay under an absolute maximum CPM at all times. While it may limit the line item's maximum average CPM over the line item's lifetime, average CPM may be exceeded at isolated points during its flight.
- Maximum average CPM is calculated over a line item's entire history, whether or not automated bidding was enabled on the flight start date.
- Check the box for "Apply advertiser floor price for targeted deals" to prioritize pacing of targeted deals, with potential impact on maximizing performance.
- When checked, deal floor prices are taken into account, and bids are adjusted accordingly.
- If there is still under-delivery towards the deal due to bids below the floor, you can increase the floor price in My Inventory > Settings.
Considerations for deals
- Don't use this setting if you want to maximize performance across Deals and Open Auction. If an insertion order has separate line items for Deals and Open Auction inventory, the system may prioritize deals if the setting is applied. To control budget allocation while maximizing performance across Deals and Open Auction:
- Set up an insertion order that exclusively consists of line items targeting deals, and apply this setting to ensure pacing of targeted deals is not impacted by bids below the floor, or
- Apply automated bidding settings at the line item level, with specific budgets set for line items targeting Deals or Open Auction respectively.
Maximize brand
With the Maximize Brand, while prioritizing spending my full budget strategy, your line item's bid will be raised until your budget can be spent.
Maximize for:
-
Completed in-view and audible impressions (video only) to prioritize impressions where the video was visible at completion of the video and the audio was audible.
-
Impressions that were viewable for at least 10 seconds (video only) to prioritize impressions where the video was visible for 10 seconds or more.
-
Viewable impressions (display and video) to prioritize impressions where the display or video ad was viewable according to Active View.
-
Use While prioritizing a target viewable CPM to optimize your bid based on the probability that an impression will be viewable (powered by Active View).
-
Advanced troubleshooting
- For an accurate overview of automated bidding performance, ensure your insertion order optimization goal aligns with the chosen bidding strategy on your automated bidding line items.
- If Apply advertiser floor price for targeted deals is applied, the algorithm raises bid prices to respect the deal floor, which may lead to overvaluing the deals inventory.
- If While prioritizing spending my full budget strategy (recommended) is in use, the algorithm may raise bid prices to improve win rate (especially if market price is high). Consider the While prioritizing a target KPI strategy and set a target performance value; however, this may limit pacing.
- When average CPM is less than Market price (Chart 2):
- If Do not exceed average CPM is applied, average CPM may need to be adjusted to improve win rate.
- If using the While prioritizing a target KPI strategy, the target value may be too low compared to market price.
- Consider using the While prioritizing spending my full budget strategy. The bidder automatically adjusts targets to balance pacing and performance.
Bidding Insights only apply to bid/auction-related impression loss reasons. When unrelated to bids/auction, you may want to review suggestions, troubleshoot deals and line items, obtain creative Approval, check the Inventory Availability report, adjust targeting, or work with your publishers.
Limitations
- The bidding algorithm can take up to 4 weeks to learn and calibrate before judging performance.
- 7 days or more can pass before enough data is available to show automated bidding insights. If the bid strategy is changed 7 days or more of fresh data is needed.
- Using ActiveView based autobidding and viewability targeting at the same time may severely limit your ability to spend your budgets or reach your desired performance goal.
- Specific KPI target values can lead to underspending (if too aggressive). In most cases, Maximize {KPI} while prioritizing spending the full budget will work best.