The following are a few Frequently Asked Questions (FAQs) on Synamedia Iris Forecasting:
The Synamedia Iris forecaster needs a 3-month historical basis of data before it can generate a forecast. If you have requested that the forecasting service be enabled, then it will be collecting and processing data in the background.
Synamedia Iris uses historical impression data and trend analysis as the basis of the forecast. We recommend enabling the data collection period once production levels of traffic can be observed so that any test or integration data (which may have different observable patterns or volumes to real production data) is excluded from the prediction algorithms.
The ‘all inventory’ forecast does not take existing campaigns and booked impressions into account when generating the forecast. The number of ad opportunities predicted will be the estimated total of all upcoming ad impressions that meet your criteria. You will need to subtract any existing booking from this number to understand how many impressions are available to sell. Alternatively, run a Line Item campaign forecast to understand if delivery for a specific booking may be achievable.
For Line Items, the forecaster will only consider the next 120 days when generating predicted ad opportunities and analyzing competing campaigns. If you run a forecast against a Line Item with run dates beyond the 120 day limit, a forecast will still be returned but it will be based on the next 120 days only and you should take into consideration that some future bookings or traffic trends outside that time frame will not have been available to the forecaster. The prediction period is today plus 120 days and not from the start date of the Line Item.
Once the base three months historical data have been provided, any new segments or KVPs that have been added to your system will require four weeks' worth of data. If sufficient data has been provided, it may be that the forecasting service does not predict that there will be any impressions that meet your criteria.
Synamedia Iris does not currently support manual overrides on the forecast data.
Synamedia Iris uses sophisticated models to forecast available inventory that will take observable traffic trends into account when making a prediction. As the system sees long-term patterns, then the forecasts will become more accurate. However, the forecaster is a prediction and not a guarantee. Traffic patterns may have changed during the delivery period from the observable historical data. Also, other bookings may have been made or adjusted since the forecast was initially generated and those changes have an impact on the ability of the system to deliver all campaigns. It is possible to rerun the forecast for any Line Item in Synamedia Iris to view the latest forecast.
Using the targeting criteria specified when asking the system for a prediction, the forecaster will predict the number of ad opportunities that might be seen by Synamedia Iris for an upcoming date range. An ad opportunity is the predicted possibility to serve an advertisement in the future and it is not a guarantee that an ad will be selected, shown or viewed. At the point of delivery, Synamedia Iris will select an appropriate ad from those in flight at that time. Once an ad is selected and played by the viewer, then an impression will be generated. Opportunities are, therefore, forward looking and impressions record an activity in the past.