MoPub Analytics for Publishers

MoPub Analytics is reserved for our premier publishers. Note that MoPub Analytics reporting is not used for billing.

MoPub Analytics UI

For a visual tour of MoAnalytics, watch our video walkthrough below.

Login

Access MoPub Analytics at https://analytics.mopub.com. A login is created for your team’s MoPub Analytics account. If you do not have the login or password, contact your MoPub account team and we will send you your login information.

Publisher Auctions Data Cube

Use the [Publisher] → My Auctions data cube to view data for all of the auctions on your inventory. You can review data as far back as July 17, 2018. Common use cases include:

  • Identify trends for specific applications
  • Track net revenue over a time period
  • Review performance by country

Platform Performance Data Cube

Use the [Publisher] → Platform Performance data cube to view data for all demand sources (Network, Marketplace, Direct Serve) in one view. You can review data as recent as two days ago and as far back as July 10, 2019. Common use cases include:

  • Track estimated revenue for all demand sources by referring to the Ad Server Revenue metric at ad unit, format, ad group, and app level.

    Ad Server Revenue is estimated revenue based on MoPub ad server calculation for all demand sources (Network, Marketplace, Direct Serve).

  • Review accurate revenue for supported networks by referring to the NR Revenue metric at the account level. MoPub scrapes and aggregates NR revenue at the account level. The label “NR” indicates that this revenue metric is network-reported, which means that this is the revenue reported by the individual network that MoPub is mediated through MoPub. To review the NR Revenue metric, enable Network Reporting in the MoPub Publisher UI.

    From MoPub Analytics, the best way to examine metrics that come from “Network Reporting” is to select only NR Revenue, NR Impression, or NR eCPM from the MEASURE dropdown menu, and then select NR Network Type, Country from SHOW or FILTER without any other dimensions and measures selected. Refer to the example below:

    Example

  • Review performance for all demand sources using the Requests, Attempts, Demand source show rate, and Inventory show rate metrics. You can further analyze these metrics by various dimensions such as ad unit, format, ad group, app, etc.

    • The Requests metric is the number of times an ad unit has requested an ad.
    • The Attempts metric is the number of times MoPub made an ad attempt to an ad demand source (network, marketplace, Direct Serve) for a given ad request.
    • The Inventory show rate metric measures the number of times an ad was shown to users (impressions) compared to the number of times an ad was requested from a specific ad unit (requests). Formula: Impressions / Requests.
    • The Demand source show rate measures the number of times an ad was shown to users (impressions) compared to the number of times MoPub made an ad attempt to an ad demand source (attempts). Formula: Impression / Attempts. A given ad request can make multiple ad attempts to the same ad source depending on the waterfall setup.

Bookmarks and Reports

While you cannot yet bookmark a view in the MoPub Analytics UI, each view does generate a unique URL. You can bookmark that URL for later use with your browser’s bookmark feature. We recommend opening the links once a week to ensure the bookmarks are saved.

MoPub Analytics also enables you to download a report containing a maximum of 5,000 rows. If you need additional rows, contact your account manager.

Alerts Functionality

Alerts are currently under construction. Please check back soon!

Metric and Dimension Definitions

Auctions Metrics

Dimension Description
Net revenue Estimated publisher revenue. This is not used for billing, but aligns within 1-3% of your final invoice.
Auctions Number of MoPub auctions. This should be roughly equivalent to the number of Marketplace attempts, which may be more or less than the number of ad requests, depending on how your waterfall is setup.
Clicks Number of clicks.
Auctions Won Number of auctions where at least one DSP bid above your price floor.
Win Rate Formula: number of auctions won / total number of auctions
Cleared Number of impressions from the MoPub Marketplace.
Clear Rate Formula: number of cleared / total number of auctions
CTR Click-through rate.
Survival Rate Formula: number of cleared / number of auctions won
eCPM Estimated cost per thousand impressions within the MoPub marketplace.
Winning Bid Avg The average highest bid for the selected auctions.
Bid Avg The average bid for the selected auctions.
Bid Max The highest bid for the selected auctions.
Bid Depth Average number of non-zero bids per auction.
Floor Avg The average price floor.
Competitive Factor Average number of non-zero bids per auction with a winner.
Uniques Number of unique devices.
Video Start This event is used to indicate that an individual creative within the ad was loaded and playback began.
Video Midpoint The creative played for at least 50% of the total duration.
Video Complete The creative was played to the end at normal speed.
Video Completion Rate This is the rate of the number of videos that started playing compared to the number that were watched to the Video Complete status.

Formula: Videos Complete / Impressions

Auctions Dimensions

Dimensions Describing Inventory

Dimension Description
Publisher The name of the publisher (you).
Application The name of your application.
Site The name of your mobile websites.
Category The application’s app store category, as you specified in the MoPub UI, along with its IAB category. In response to requests by DSPs, MoPub created a mapping from each app store category to an IAB category.
Global AID For Android apps, this is the package name. For iOS apps, this is the iTunes ID.
Ad Size The ad size of the impression opportunity in the bid request expressed as a non-format array, width x height.
Format Array Sizes The ad size of the impression opportunity in the bid request expressed as a format array, width x height.
Winning Creative Ad Size The ad size of the auction winner.
Adunit Format The placement format as defined by the publisher, shown as follows, with legacy format names in parentheses:
  • banner (320x50)
  • native
  • medium_rectangle (300x250)
  • full
  • rewarded_video
  • banner (728x90)
  • full (full_tablet)
  • custom
  • deprecated (160x600)
Adunit Internal MoPub ID for your ad unit.
Adgroup ID Internal MoPub ID for the ad group (line item).
Blocked Categories The IAB categories you have chosen to block from your inventory.
Request Video Duration Whether you have enabled 15-second video, 30-second video, or both kinds, to be shown on your inventory.
Price Floor The price floor in buckets for the selected auctions.
Video Inventory Enabled Inventory enabled for video only, both video and static, or not enabled for video.
Interstitial Playables Enabled Interstitial playable inventory.
Adgroup ID Internal MoPub ID for the ad group (line item).
Blocked Categories The IAB categories you have chosen to block from your inventory.
Request Video Duration Whether you have enabled 15-second video, 30-second video, or both to be shown on your inventory.
Price Floor The price floor in buckets for the selected auctions.
Video Inventory Enabled Inventory enabled for video only, both video and static, or not enabled for video.
Interstitial Playables Enabled Interstitial playable inventory.
Ad Group Name The name of the ad group (line item).
Ad Group Priority The priority level of the ad group (line item).
App Version The release version of your applications.
Inventory Package The MoPub-created Inventory Packages to which your inventory belongs.
Publisher ID Internal MoPub ID for the publisher (you).
Request Deal ID The deal ID for the Inventory Package to which your inventory belongs.

Dimensions Describing Users

Dimension Description
Device Device type.
Device Model Model of the device.
OS Operating system.
OS Version Operating system version.
SDK Version MoPub SDK version.
Country User’s country, based on the user’s IP address.
Has Location Whether the SDK passes latitude and longitude data.
Has Device ID Whether or not the SDK passes the device ID.
Age Range The age ranges of your users (if passed in the bid request).
Carrier Name User's mobile carrier.
Gender The gender of your users (if passed in the bid request).

Dimensions Describing Creatives

Dimension Description
Bidder Name Name of the Demand-Side Platform buying on the MoPub marketplace.
Adomain Advertiser domain of the ads shown.
Buyer Seat The DSP’s ID for their buyers. Only a few DSPs pass this information correctly.
Creative MRAID Whether or not the shown ad is MRAID.
Creative Video Served Type of video that was served.
Response Video Duration Winning bid response’s video duration.
Creative ID The ID of the creative.
Impression Latency Time between the auction and impression.
Response Deal ID The deal ID for the Inventory Package to which your inventory belongs, sent in a bid response, indicating that the buyer has bought your inventory as part of the Inventory Package.
Viewability Vendors Response The viewability partner whose tag was used to measure viewability on your inventory.
GDPR Applies Whether or not users are subject to GDPR. GDPR applies when MoPub detects that a user opened a given application for the first time in the European Economic Area, United Kingdom, or Switzerland; or when the publisher determines that GDPR applies for apps using MoPub SDK v5.1 or higher.
Consent Status Whether or not the user has provided consent, if they are subject to GDPR.
  • Yes: The user has provided consent to receive personalized ads.
  • No: The user has not provided consent to receive personalized ads. We will only serve contextual ads.
  • N/A: GDPR Does Not Apply; users are not subject to GDPR. This field should only be available when 'GDPR Applies' is equal to 'No'. Filter on 'GDPR Applies' = 'Yes' before reviewing consent status data.
Consent Status Details More granular breakdown of consent status when the user is subject to GDPR.
  • Explicit Yes: The user either consented to the MoPub dialog or the publisher notified our SDK that the user provided consent via the publisher’s custom consent mechanism. MoPub and our partners have consent to serve personalized ads.
  • Explicit No: No consent; only contextual ads served. The user explicitly said no to the MoPub dialog or the publisher notified our SDK that the user did not provide consent via the publisher's custom consent mechanism.
  • Do Not Track Enabled: No consent; only contextual ads served. The user has their device's Limit Ad Tracking preference enabled.
  • Pending MoPub Approval: No consent; only contextual ads served. Publishers must be approved to use their own consent mechhanism. You will see this value if MoPub has approved your use of your own consent mechanism.
  • Unknown: No consent; only contextual ads served. This is reported when the following occurs:
    • Users have not been asked to provide consent in apps that contain MoPub SDK v5.0 or higher. This may occur if publishers do not ask the user to provide consent.
    • Users have been asked to provide consent in apps that contain MoPub SDK v5.0 or higher, but they either closed the app before responding or clicked on the exit button of the MoPub consent dialog.
    • Apps that are on a MoPub SDK older than v5.0.
  • N/A: GDPR does not apply; the user is not subject to GDPR.
Unified Auction Denotes whether or not an auction was unified (Advanced Bidding) or was a traditional line item auction.

Platform Performance Metrics

Metric Name Definition
AdServer Revenue Estimated revenue reported by MoPub ad server for all demand sources (network, Direct Serve, Marketplace). It can be broken down into a variety of dimensions, such as ad unit, format, ad group, app, etc.

This metric includes the revenue of network, and it is estimated and calculated based on:
  • MoPub impressions
  • Auto CPM or manually-entered eCPM
Direct Serve and Marketplace revenue are accurate revenues reported by the MoPub ad server.

This is not used for billing.
AdServer Impressions Total number of times the ad was shown to users. This is reported by MoPub.
AdServer CTR Click Through Rate is the number of times an ad was clicked compared to the number of times an ad was shown to users. This is reported by MoPub.
AdServer Clicks Number of times an ad was clicked. This is reported by MoPub.
AdServer eCPM Effective cost per thousand impressions. This is reported by MoPub.
AdServer Requests The number of times an ad unit has requested an ad. This is reported for all demand source types.
AdServer Attempts The number of times MoPub made an ad attempt to an ad demand source (network, Marketplace, Direct Serve) for a given ad request. Some of these attempts may be counted more than once as the attempt moves through the mediation waterfall. The attempt includes both client and server sides of the attempt, based on demand source type.
Inventory Show Rate Represents the number of times an ad was shown to users (impressions) compared to the number of times an ad was requested from a specific ad unit (requests).

Formula: Impressions / Requests
Demand Source Show Rate Demand source show rate represents the number of times an ad was shown to users (impressions) compared to the number of times MoPub made an ad attempt to an ad demand source (attempts).

Formula: Impressions / Attempts

Note: If the same line item occurs multiple times in the waterfall, the number of attempts for this line item will be counted multiple times, which may cause the demand source show rate to drop.
Uniques The number of unique device IDs reported by MoPub. This excludes “DNT” and “GDPR non-consenting” devices.
Impressions per Unique Number of impressions generated per unique user.
Trackable DAUs Trackable daily active users is based on unique device IDs. This is calculated by averaging the daily unique ID count over a period of time. Much like “Uniques”, this metric excludes “DNT” and “GDPR non-consenting” users.
NR Revenue Revenue reported by supported network when network reporting is enabled.

Note that the time zone of supported network may be different as shown below:
  • PDT: AdMob, Facebook
  • ET: Verizon
  • UTC: AppLovin, AdColony, Vungle, Tapjoy, Yahoo! Flurry, Unity, ironSource, Lifestreet, Chartboost
  • Customized Timezone: Conversant (NOTE: Can be adjusted in network's portal)
This is not used for billing.
NR Impressions Total number of times the ad was shown to users. NR Impressions is reported by the supported network in question when network reporting is enabled.
NR eCPM Effective cost per thousand impressions. This is reported by supported networks when network reporting is enabled.

Formula: (NR Revenue / NR Impression) * 1000
AdServer Revenue per 1k Uniques The estimated revenue generated per 1000 unique users.
Fills (Beta) Fills (for example, Successful Loads) is the total number of creatives that successfully loaded on the device. This is before the Impression renders on the device.
Inventory Fill Rate (Beta) The proportion of Fills to Ad Requests. Inventory Fill Rate represents how often Ad Requests end up in Fills. Use only with inventory dimensions, such as ad unit, app, format, etc.

Formula: Fills / Requests
Demand Source Fill Rate (Beta) Represents the number of times an ad was successfully loaded (fills) compared to the number of times MoPub made an ad attempt to an ad demand source (attempts). Refer to our note about the Render Rate metric for extra context on this metric.

Formula: Fills / Attempts
Render Rate (Beta) The proportion of Fills that resulted in Impressions. This metric can be used with both Inventory and Demand Efficiency metrics. Refer to our note about the Render Rate metric for more details about this metric.

Formula: Impressions / Fills
Avg. Request to Attempt Response Time (Beta) The average time duration from the moment an ad request is initiated to the last attempt. The response is returned regardless of failure or success.

Note: If the last ad attempt result response is 'success', this indicates that the ad request received a fill. If the last ad load result response is 'failure', this indicates that the ad request went through all ad groups in the waterfall and didn’t receive a fill.
Avg. Attempt to Attempt Response Time (Beta) The average time duration from the moment an ad attempt response is sent by an ad group until the ad load result response is returned for the ad group, regardless of the result (failure or success).
Attempt Response (Beta) The total number of attempted calls that generated a response, regardless of the response content (Success, Timeout, Error, or any other result).

Platform Performance Dimensions

Dimension Name Definition
AdGroup Type A dimension defined by MoPub. Use this dimension to break the metrics down by individual Ad Group Type as shown below:
  • Guarantee
  • Non-guaranteed
  • Promotion
  • Marketplace/MPX Line Item
  • PMP/PMP Line Item
  • Network
AdGroup Network Type A dimension defined by MoPub. Use this dimension to break the metrics down by individual Ad Group Network Type. To review more accurate data for a “Supported Network” that has network reporting enabled, refer to the second bullet point in the Platform Performance Data Cube section.
AdGroup ID The internal MoPub ID for the ad group (line item). Use this dimension to break the metrics down at the individual ad group level. You can find your ad group ID by referring to the Line Item Key in the MoPub Publisher UI.
AdGroup Name The name of the ad group (line item). Use this dimension to break the metrics down at the individual ad group level. Your ad group name is the same as the Line Item Name in the MoPub Publisher UI.
Country Country breakdown that is merged with existing ad server cube and network reporting.
Adunit Format A dimension defined by MoPub. Use this dimension to break the metrics down by individual format, as shown below, with legacy sizes in parentheses:
  • banner (320x50)
  • native
  • medium_rectangle (300x250)
  • full
  • rewarded_video
  • banner (728x90)
  • full (full_tablet)
  • custom
  • deprecated (160x600)
Adunit ID The internal MoPub ID for your ad unit. Use this dimension to break the metrics down at the individual ad unit level. To find your ad unit ID, copy the Ad Unit Key in the MoPub Publisher UI.
Adunit Name The name for your ad unit. Use this dimension to break the metrics down at the individual Ad Unit level. Your name by is the same as the Ad Unit Name in the MoPub Publisher UI.
OS The operating system. Use this dimension to break the metrics down by individual OS or platform (Android, iOS).
App ID The internal MoPub ID for your application. Use this dimension to break the metrics down at the individual App level. You can find your App ID by copying the App Key from MoPub Publisher UI.
App Name The name for your application. Use this dimension to break the metrics down at the individual App level. You can find your application name by copying the App Name from MoPub Publisher UI.
Time Use this dimension to customize your data’s date range. The granularity of the Platform Performance data cube is at the daily level.
NR Network Type Supported network name with network reporting enabled. Refer to the second bullet point in the Platform Performance Data Cube section for more details on how to use this dimension.
Attempt Response Type (Beta) This is a dimension you can apply when you select the Attempt Response metric to reveal the individual result of the attempt response.

Inventory Efficiency Metrics (Beta)

Dimension Name Definition
Fill Rate (Inv.) The proportion of Fills to Ad Requests. This represents how often Ad Requests end up in Fills. Use only with inventory dimensions such as adunit, app, format, etc.

Formula: Fills / Requests
Show Rate (Inv.) The proportion of Impressions to Ad Requests. This represents how often Ad Requests end up in Impressions. Use only with inventory context dimensions such as ad unit, app, format, etc.

Formula: Impressions / Requests
Waterfall Latency Previously known as Avg. Request to Attempt Response Time. Waterfall Latency reveals the average time duration from the moment an ad request is initiated to the last attempt for a response. A response is returned whether the result is failure or success.

Note: If the response to the last ad attempt was 'success', this indicates that the ad request received a fill. If the response to the last ad load is 'failure', this indicates that the waterfall of this ad request went through all ad groups without receiving a fill.
Waterfall Depth Number of Attempt Responses Per Ad Request. Waterfall Depth is strongly correlated with Waterfall Latency. Use only with inventory context dimensions, such as ad unit, app, format, etc.

Demand Efficiency Metrics

Dimension Name Definition

Fill Rate (Demand) 

The proportion of Attempts that resulted in Fills. Most useful with demand source context dimensions, such as ad group type, ad group network type, and ad group name.

Formula: Fills / Attempt Responses

Show Rate (Demand)

The proportion of Attempt Responses that result in Impressions. Most useful with demand source context dimensions, such as ad group type, ad group network type, ad group name, etc.

Formula: Impressions / Attempt Responses

Latency (Demand)

Previously known as Avg. Attempt to Attempt Response Time. Latency (Demand) reveals the average time duration from the moment an ad attempt response is sent by an ad group until the ad load result response is returned for the ad group, regardless of the result (failure or success). Most useful with demand source context dimensions, such as ad group type, ad group network type, and ad group name.
Fill Time (Demand) The average time duration to receive an Attempt Response after attempting a demand source (ad group) when the Attempt Response results in a Fill (successful load). This enables you to directly compare Fill durations across multiple line items. Most useful with demand source context dimensions, such as ad group type, ad group network type, and ad group name.
Timeout Rate The proportion of Attempt Responses that resulted in a Timeout response. Most useful with demand source context dimensions, such as ad group type, ad group network type, and ad group name.
Missing Adapter Rate The proportion of Attempt Responses that resulted in a Missing Adapter response. Most useful with demand source context dimensions, such as ad group type, ad group network type, and ad group name.

Video Walkthrough

We created a three-part video series to walk you through accessing and using MoPub Analytics.

Introduction to MoPub Analytics

Functionality and Features

Visualizing your Data

For additional details that aren’t covered in the videos, keep reading.

Note About GDPR Data

MoPub reporting does not support a unique device identifier for users who have not provided consent. Therefore, the Uniques metric for non-consenting users is zero, and Uniques only represent consenting users. This only apply to users for whom GDPR Applies.

MoPub sends contextual-only auctions when GDPR Applies to users and they do not provide consent (Consent Status is ‘No’). For more information, refer to the GDPR section of the MoPub OpenRTB integration guide.

Unique users who have consented will be updated monthly. Reviewing consenting unique users across months will result in incorrect data. To report on unique users from month to month, pull the number of unique users for each month. and then average the count across multiple months.

If the Number of Unique Devices is lower than 1000 for the data granularity you applied, both the My Auctions and Platform Performance data cubes will aggregate all metrics into an Others bucket to comply with GDPR data protection laws.

Note About the Render Rate Metric

Within the inventory context, this metric represents the render rate for a unique ad request. You can find it when applying dimensions at the ad unit or app level. It’s available in both the Ad Server and Platform Performance cubes.

Within the demand source context, this metric represents the render rate for the ad group. Analyze render rate broken down by demand source in the Platform Performance cube.

Unfortunately, when applying ad group level dimensions in the Ad Server cube, the render rate is always 100% because the Ad Server cube’s impressions are aggregated based only on ad requests that generated impressions. Consider these examples of how render rate in the demand source context in the the Ad Server cube can end up always being 100% when analyzed by ad group level dimensions:

Ad Request 1 (got the impression from Ad Group B)

  • Ad Group A: fill = 0, imp = 0
  • Ad Group B: fill = 1, imp = 1

Ad Request 2 (didn’t have impression at all)<

  • Ad Group A: fill = 0, imp = 0
  • Ad Group B: fill = 0, imp = 0
  • Ad Group C: fill = 0, imp = 0

Ad Request 3 (didn’t have impression at all)

  • Ad Group A: fill = 0, imp = 0
  • Ad Group B: fill = 1, imp = 0

The render rate when applying ad group level dimensions will be:

  • Ad Group A: 0/0 = 0%
  • Ad Group B: 1/1 = 100% because the data from Ad Request 3 won’t be captured when applying ad group level dimensions. The correct Ad Group B render rate should be 1/2 = 50%.
  • Ad Group C: 0/0 = 0%

FAQ

The most recent data in the Platform Performance data cube is one day old for all measures except NR Revenue, NR Impression, and NR eCPM; the most recent data for these three metrics is two days old.

At this time, we do not have a one-to-one relationship between ad unit ID, line item ID, and network ID. We are still developing the ability to report on this precise network revenue data at a level more granular than account, without double-counting. Currently, you can drill down into network reporting metrics by the Country dimension.

However, we do provide estimated network revenue data at more granular levels based on Auto CPM or manually entered CPM, which serves well as a replacement in providing a comprehensive view of your monetization efforts.

The demand source show rate formula is Impressions / Attempts. The denominator is the number of times MoPub makes an attempt to an ad demand source in the waterfall.

For all demand sources except Marketplace, MoPub only counts the attempt to the demand source when MoPub has reached the priority assigned to the demand sources. Until MoPub reaches the priority assigned to the demand source, no attempts are counted for that demand source. For example, if there are two demand sources, each assigned its own priority, and the ad request is filled by the first demand source in the highest priority, the second demand source in the second priority does not trigger an “attempt.”

However, the “attempt” for Marketplace specifically is always triggered to prepare an auction when the waterfall starts, regardless of Marketplace’s priority. This means there are always attempts counted toward MoPub Marketplace, as long as it is one of the demand sources within the waterfall. As a result, the number of “attempts” for Marketplace may be inflated as compared to other demand sources, making its fill rate (Impressions / Attempts) artificially lower.

Yes. The Demand Source Show Rate formula is Impressions / Attempts. The “attempt” is not a unique attempt by network, so if the same line item is reached multiple times in the waterfall, the “number of attempts” is counted multiple times as well. For example, if line item AdMob 12345 appears four times in a waterfall and recieves the impression from the fourth attempt, the total attempts for AdMob 12345 is ‘4’ instead of ‘1’, and as a result, the Demand Source Show Rate is 25% instead of 100%.

If you have more than one account (meaning more than one account key) with MoPub and use the same network reporting credentials for all of your accounts, the network reporting data you see in the data cube will be aggregated for all of your accounts.

For Ad Group level dimensions, the AdServer Requests count and the AdServer Impressions count are always equal because of how these datasets were joined. Viewing AdServer Attempts instead of AdServer Requests is more meaningful for understanding your Ad Group level performance, because the attempt count relates to the activity between your ad group and demand sources.

Yes. Filter on SDK Version to be ‘5.3’ and higher for the Inventory Fill Rate, Demand Source Fill Rate, Fills, Render rate, Avg. Attempt to Attempt Response Time, and Avg. Request to Attempt Response Time metrics. These metrics are supported by the client-side waterfall which is available for the newer MoPub SDK versions (v5.3 and higher).

Leveraging Order Name and Order Advertiser dimensions can help you find individual network segment data.

Metrics with “Inventory” in their name denote measures within the inventory context, such as ad unit, or app. When analyzing these metrics, apply app or ad unit level dimensions, such as ad unit name, ad unit format, etc. Metrics with “Demand Source” in their name denote measures within the demand source context, such as ad group, or line item. When analyzing these metrics, apply ad group-level dimensions, such as ad group name, ad group type, ad group network type, etc. The diagram below shows the ad lifecycle funnel, with inventory and demand source context.

Diagram

There are several possibilities:

  • Small ad unit: Some metrics are still a product of sampling. Be mindful when dealing with smaller ad units (adunits that generate small amounts of requests), which may suffer from sampling errors.
  • Altered SDK: Modifying or customizing our SDK may cause strange numbers, because all the metrics are derived from the SDK events assuming proper integration with no technical modifications.

No, for MPX and AB this means that the ad has cached successfully and is ready to play. The meaning varies for each network because networks return “success” based on their own logic.

Latency metrics measure how different demand sources perform in the waterfall. For Marketplace to be in the waterfall, it must have run an auction that produced a winning bidder immediately prior to the waterfall. A Marketplace attempt in the waterfall is the act of loading the creative ID supplied by the successful auction. Therefore, when Marketplace is included in the waterfall, we already know that there is a creative ready to load. This works differently for networks; we send networks an attempt without having prior knowledge about whether or not they intend to fill.

Marketplace’s high fill rates mean that Marketplace fills creatives the vast majority of the time. Networks’ low fill rates mean that networks respond with a ‘No Fill’ response more often. ‘No Fill’ responses take significantly less time than filling. We recommend styding the Fill Time (Demand) metric, in which you will find that Marketplace is comparable, and actually faster platform-wide.

This again is attached to the high fill rate scenario. A demand source can only time out if it’s attempting to fill. For Marketplace line items, the high fill rate opens them up to more timeouts. Network timeouts can only occur when they decide to fill, which is a rare occurrence.

Changelog

September 16, 2019

  • Added new dimensions to the Auctions data cube
    • Added Adunit Format dimension
    • Added Format Array Sizes dimension
    • Added Winning Creative Ad Size dimension
  • Added a new dimension to the Platform Performance data cube
    • Added unified format versions to the Adunit Format dimension

Last updated May 18, 2020

TWITTER, MOPUB, and the Bird logo are trademarks of Twitter, Inc. or its affiliates. All third party logos and trademarks included are the property of their respective owners.

© 2020 MoPub (a division of Twitter, Inc.)