How to create a best-in-class monetization analytics setup?

Digital publishers manage multiple monetization avenues (subscriptions, programmatic ads, direct ads, commerce) in an effort to maximize their revenues. This leads to ominous challenges around managing all the data and staying on top of things. In today’s business scenario, monetization teams are increasingly expected to do more with less. So, setting up a top-notch analytics system is a crucial part of the strategy. Setting up analytics & reporting is always an iterative activity, but starting with a framework can really help speed the process up and get the basic things right from the get go.

Since several marquee publishers across the globe use Tercept, we’ve had a chance to work closely with the best in the industry in setting up their monetization analytics. Here, we’ve captured the best practices to keep in mind when doing so.

  1. Completely automate all reporting requirements

    Why is it important?

    • Saves time & money.
    • Makes reporting more accurate.
    • Gives access to deep insights at the click of a button.
    • Frees from frustrating repetitive activities.
    • Detects bugs and setup issues faster.
    • Enables faster decision-making.

    What all should a publisher automate?

    • Automate fetching data from the ad-server/s & programmatic partners (Ad Exchanges, SSPs, DSPs & Ad Networks).
    • Automate fetching data from the customers’ ad-server/s so that the team can compare metrics with the ad-server numbers for direct campaigns.
    • Automate fetching data from all the analytics platforms (web click stream data, app click stream data, app attribution data, Comscore data, page/app performance data, etc.) that are in use.
    • Automate fetching all the transaction data that the properties generate: on-site commerce, in-app commerce, commerce partnerships, etc.
    • Automate fetching highly granular data including hourly/real-time data where possible.
    • Automate dashboards, saved queries & scheduled reports for the senior management (high level dashboards), middle management (in-depth dashboards) and analysts (in-depth dashboards + multi-dimensional pivots).

    The average ad-ops person spends 81% of her time aggregating data & preparing reports and only 19% of her time analyzing the data & generating insights. Ideally, the plan should be to bring the 81% down to zero.

  2. Normalize all the data

    Why is it important?

    • Maintains consistency in the data across partners.
    • Makes reporting more accurate.
    • Saves time & rework down the line.
    • Reduces confusions around what every metric or dimension signifies.
    • Enhances data accuracy and enables more accurate decision-making.

    What all should be normalized?

    • Ensure consistent naming convention of every metric & dimension on every partner.
    • Ensure consistency in reporting currency & reporting time zones of every partner.
    • Group dimension members meaningfully & create additional custom dimensions to ensure consistency in dimension members across partners. For example, most of our customers get traffic from several geos, but end up grouping countries into meaningful buckets using custom groupings & setup a custom dimension to access these groupings. Similarly, publishers create a custom dimension called Platform where the data is split across Mobile Web, Desktop Web, App, AMP, Others. Another example is grouping their partners into Programmatic/HB/Networks/House/etc.
    • Break up the data meaningfully to achieve an optimum level of granularity. For example: many publishers break up Google Ad Exchange data into AdX Mediation, AdX First-look, EBDA, etc.)
    • If there are multiple properties, group ad-units accordingly so that the team is able to track aggregated data for every property.
    • Setup reporting rules carefully: bake in revenue share percentages, any off-line pricing arrangements, ad-server adjustments, discrepancy adjustments, reconciliation adjustments, bank charges, etc. to make sure the reports are super accurate.

  3. Combine the data into meaningful tables

    Why is it important?

    • Enables deep insights from metrics that have never been tracked consistently before.
    • Eliminates data silos (worst enemies). Data silos result in partial information that are often times dangerous for decision making.
    • Enables end-to-end understanding of user behaviors starting from marketing behavior to on-site/in-app behavior to monetization behavior.
    • Helps create surprising efficiencies across functions including strategy, marketing, product & monetization teams.

    What are some must-have combined data tables?

    • ROI by UTM Source: Track monetization at a user-level to get an end-to-end picture of marketing/acquisition, in-app monetization as well as ad monetization. This helps tie up campaign performance to monetization data, eliminates guesswork & builds profitability.
    • PageRPM/ScreenRPM, SessionRPM, Revenue by DAU: Combine analytics data with monetization data to get a deep understanding of monetization by web pages, app screens, user sessions, user activity, etc.
    • Header Bidding Discrepancy: Combine data from the ad-server, header-bidding provider & individual header-bidding partners to setup discrepancy tables so that the team is on top of any issues with zero effort.
    • Direct campaign performance: Combine data from the ad-server with the buyers’ (brand or agency) ad-server to automatically track the metrics (conversions, post-click page-views, etc.) that matter to the buyer as well as to track discrepancy between clicks & landing-page visits.
    • Order management, billing & ad-server reconciliations: Automatically map data from the order management software, billing software & the ad-server to completely eliminate the need for tedious reconciliations.

  4. Setup automated alerts

    Why is it important?

    • Enables peace of mind & saves time that’s wasted in constant monitoring of metrics.
    • Helps catch bugs, setup issues, inconsistencies & unusual behaviors quickly so that the team can take timely action and minimize damage.
    • Helps the team capitalize on trends by comparing daily & weekly numbers and taking action early.

    What are some best practices for alerts?

    • When setting up alerts, the primary focus has to be on ratio metrics. Absolute metrics like Impressions, Ad-Requests, Revenue, Clicks, etc. can fluctuate quite significantly. But, ratio metrics should remain steady.
    • Some key metrics on Google Ad Manager that need alerts are CPM, Fill Rate, Render Rate, Delivery Rate, Viewability and Ad-Request CPM.
    • Metrics related to combined tables with data from analytics & monetization that require alerts are: PageRPM/ScreenRPM, Session RPM, Revenue per User, Revenue per Visit & Revenue per DAU.
    • Focus on key dimensions and setup alerts for the above metrics across each of these dimensions. Key dimensions for most publishers are partner, platform, geo, ad-unit, pricing rules & buyers.
    • Setup hourly alerts for extremely critical metrics.
    • Ensure that alerts have been setup to be delivered when the metric drops/jumps by a more than certain percentage as well as when certain upper & lower limits are breached.
    • Ensure that alerts are delivered to the most used communication channels in the team: Slack, SMS, and Email.
    • Setup alerts to be delivered to the right people based on the severity. Define an escalation matrix: lower priority alerts to analysts and severe alerts to middle & senior management. Define severity based on the metric & extent of drop/deviation.
    • Although we have covered a whole gamut of alerts above, spend time figuring out what’s important for the business and avoid over burdening the system with too many alerts as it can lead to fatigue. Also, ensure to maintain focus on alerts, which result in some immediate action items, rather than just informative alerts.

  5. Create a simple dashboard & alerts setup for the direct sales team

    Why is it important?

    • Direct sales teams often work with insufficient information.
    • Better understanding of buyer (brands & agencies) behavior will dramatically increase their chances of closing a deal and get better deal terms. It also empowers them to negotiate better.
    • Timing is everything when striking a deal – so, being able to act on an opportunity promptly when favorable conditions exist can make all the difference in closure rate.

    What are some must-have setups for direct sales teams?

    • Track Revenue & CPM by date by Brand, Advertiser, Buyer Network & Bidder/DSP for top 5 programmatic partners individually as well as aggregate. Slice the data by geo, platform and ad-units (if necessary).
    • Some key metrics on Google Ad Manager that need alerts are CPM, Fill Rate, Render Rate, Delivery Rate, Viewability and Ad-Request CPM.
    • Setup a dashboard with this data that is automatically populated daily (or hourly when available).
    • Setup alerts for any change in Spend or CPM by each of the above.

Gourav Chindlur
CEO, Tercept Inc.

  • Share:

Leave a Comment