Purpose:

The Python script identifies and tags campaigns where the publisher cost for the prior day is at least 90% of the daily budget, indicating a potential budget cap situation.

To Elaborate

The script is designed to monitor advertising campaigns by analyzing their daily spending relative to their allocated daily budgets. It specifically identifies campaigns where the publisher cost reaches or exceeds 90% of the daily budget, which may suggest that the campaign is at risk of being budget-capped. Additionally, it considers other factors such as the ‘iDNE’ value and ‘Lost Impression Share’ to refine the alert criteria. If these conditions are met, the script tags the campaign with a “Budget Capped” alert, helping advertisers manage their budget allocations more effectively and avoid potential overspending or missed opportunities due to budget constraints.

Walking Through the Code

1. Data Preparation
   • The script begins by converting the ‘iDNE’ column in the input DataFrame to numeric values, handling any conversion errors by coercing them to NaN. This ensures that subsequent calculations involving ‘iDNE’ are performed correctly.
2. Calculation of Ratios
   • It calculates the ratio of publisher cost to daily budget for each campaign and stores this in a new column ‘Pub Cost / Budget’. This ratio helps determine how close the spending is to the budget limit.
3. Budget Cap Alert Logic
   • A threshold of 90% is set for the ‘Pub Cost / Budget’ ratio. This threshold can be adjusted by the user if needed.
   • The script checks if the ‘Pub Cost / Budget’ is greater than or equal to the threshold, ‘iDNE’ is less than 1.5, and ‘Lost Impression Share’ is greater than 10%. If all conditions are met, it tags the campaign with “Budget Capped” in the ‘Budget Alert’ column.
4. Output Preparation
   • Finally, the script selects specific columns (‘Account’, ‘Campaign’, ‘Budget Alert’) to create an output DataFrame, which is then printed. This output provides a concise view of campaigns that may require attention due to budget constraints.

Vitals

• Script ID : 883
• Client ID / Customer ID: 248500018 / 13095968
• Action Type: Bulk Upload
• Item Changed: Campaign
• Output Columns: Account, Campaign, Budget Alert
• Linked Datasource: M1 Report
• Reference Datasource: None
• Owner: smalina@marinsoftware.com (smalina@marinsoftware.com)
• Created by Stephen Malina on 2024-03-29 17:37
• Last Updated by Stephen Malina on 2024-04-29 14:11

> See it in Action

Python Code

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
#
# Tag campaign if Pub Cost for prior day is 90% of Daily Budget
#
#
# Author: Stephen Malina
# Date: 2024-03-27

RPT_COL_CAMPAIGN = 'Campaign'
RPT_COL_PUBLISHER = 'Publisher'
RPT_COL_PUB_COST = 'Pub. Cost A$'
RPT_COL_ACCOUNT = 'Account'
RPT_COL_DAILY_BUDGET = 'Daily Budget'
RPT_COL_CAMPAIGN_STATUS = 'Campaign Status'
RPT_COL_BUDGET_ALERT = 'Budget Alert'
RPT_COL_iDNE = 'iDNE'
RPT_COL_LOST_IMP_SHARE = 'Lost Impr. Share (Budget) %'
BULK_COL_ACCOUNT = 'Account'
BULK_COL_CAMPAIGN = 'Campaign'
BULK_COL_BUDGET_ALERT = 'Budget Alert'

# Convert the 'iDNE' column to numeric values
inputDf[RPT_COL_iDNE] = pd.to_numeric(inputDf[RPT_COL_iDNE], errors='coerce')

# Calculate the difference between daily budget and pub cost, include iDNE for consideration
inputDf['Pub Cost / Budget'] = inputDf[RPT_COL_PUB_COST] / inputDf[RPT_COL_DAILY_BUDGET]
inputDf['iDNE'] = inputDf[RPT_COL_iDNE]
inputDf['Lost Imp Share'] = inputDf[RPT_COL_LOST_IMP_SHARE]

# Output a "Budget Cap Alert" where daily spend is greater than 90% of daily budget and iDNE is less than 1.5
threshold = 0.90  # Adjust this threshold as needed
inputDf.loc[inputDf['Pub Cost / Budget'] < threshold, BULK_COL_BUDGET_ALERT] = ""
inputDf.loc[(inputDf['Pub Cost / Budget'] >= threshold) & (inputDf['iDNE'] < 1.5) & (inputDf['Lost Imp Share'] > 10), BULK_COL_BUDGET_ALERT] = "Budget Capped"

# Print the modified inputDf
print(inputDf)
cols = [RPT_COL_ACCOUNT, RPT_COL_CAMPAIGN, BULK_COL_BUDGET_ALERT]
outputDf = inputDf[cols]

Post generated on 2025-03-11 01:25:51 GMT