Existing Forecast

One of the simplest forecasts we can execute is one attempting to predict impressions and unique users for the ads running in your network. The payload for an “existing” forecast request looks like the following:

{
  "Type": "existing",
  "EndDate": "2023-03-20"
}

It has type “existing” and you must specify an end date (into the future) until which to forecast impressions and unique users. Once the forecast is finished, you should get back a response like the following:

{
  "resultStatus": "success",
  "result": {
    "total": {
      "123456789": {
        "impressions": 0,
        "uniqueUsers": 0
      },
      "123456799": {
        "impressions": 76974,
        "uniqueUsers": 31738
      },
      "123456999": {
        "impressions": 290126,
        "uniqueUsers": 94924
      }
    }
  },
  "id": "29d418c3-4ac2-40cc-8c20-8c9c900408ce",
  "status": "finished",
  "progress": 100.0,
  "desc": "Forecast is finished"
}

This is telling us that there are 3 ads running, the ad with ID 123456789 will have no impressions, while ads with IDs 123456799 and 123456999 will get a few. To customize the response, you can use the params object in the request. One possible customization option is the Sampling, which can be a number between 0 (for higher accuracy) and 3 (for faster responses). The following payload triggers a forecast request for existing ads using sampling level 3:

{
  "Type": "existing",
  "EndDate": "2023-03-20",
  "Params": {
    "Sampling": 3
  }
}

Forecasts for existing ads can take several minutes to complete. We’re actively working on improving the response times.