Question

我一直在努力关闭使用Python和Facebook营销API在Adset中表现不佳的Facebook广告。我有点担心我的代码对API的访问次数。另一个问题是我使用'get_insights'方法来访问我想用来构建逻辑决策的参数，但我需要使用'get_ads'来打开/关闭它们，所以我觉得我做的事情增加了一倍。这是我到目前为止使用API v2.9做的一个例子：

from facebookads.api import FacebookAdsApi
from facebookads import adobjects
from facebookads.adobjects.adaccount import *
from facebookads.adobjects.campaign import *
from facebookads.adobjects.adset import *
from facebookads.adobjects.ad import *
from fctn import * # this is just a file where I centralized some functions
import credentials
import copy

# Auth
my_app_id = credentials.my_app_id
my_app_secret = credentials.my_app_secret
my_access_token = credentials.my_access_token
api = FacebookAdsApi.init(my_app_id, my_app_secret, my_access_token)

ad_account = AdAccount(credentials.ad_account)

# Batch creation
my_batch = api.new_batch()

# Desired fields
fields = ['campaign_name', 'adset_name', 'ad_name', 'ctr', 'impressions']

# Getting all Adsets
ad_sets = ad_account.get_ad_sets(fields=[AdSet.Field.name, Ad.Field.created_time, Ad.Field.status],
    params={
        'effective_status': ['ACTIVE'],
        'date_preset': 'last_30d',
        'limit': 5000})

# We'll iterate over each adset because we want to campare just the ads inside the same adset       
for ad_set in ad_sets:

    ads = ad_set.get_ads(fields=[Ad.Field.name, Ad.Field.created_time, Ad.Field.status],
                              params={'effective_status': ['ACTIVE'],
                                      'date_preset': 'last_30d',
                                 'limit': 5000})

    ads_insights = ad_set.get_insights(fields=fields,
                              params={'level': 'ad',
                                      'date_preset': 'last_30d',
                                 'effective_status': ['ACTIVE'],
                                 'limit': 5000})

    # this is an external funtion to get the median in relation with some metric
    median_ctr = median_metric(ads_insights, 'ctr')

    print(median_ctr)
    print(ads_insights[0]['campaign_name'])
    print(ad_set['name'])
    print('BEFORE')
    print(ads)

    for i in range(0, len(ads)):

        if dias_ate_hoje(ads[i]['created_time'][:10]) < 10:
            # If Ad running less than 10 days, keep going
            continue
        else:
            if float(ads_insights[i]['impressions']) < 300:
                # If impressions less then 300, keep going (just an arbitrary decision here)
                continue
            else:
                if float(ads_insights[i]['ctr']) < median_ctr:
                    # If Ad is in the worst half in relation with CTR: turn off
                    ads[i].api_update(params={'status': 'PAUSED'}, batch=my_batch)
                else:
                    continue

    my_batch.execute()

    print('AFTER')
    print(ads)

希望任何已经做过类似事情的人可以帮助我更好地使用这些代码，同时减少对API的访问和双倍的代码。感谢。

Answer 1

这是你可以做的：

从帐户级ad_account.get_ads()
从帐户ad_account.get_insights(fields=fields,params={'level': 'ad', ...})

在您的见解API调用中，您需要将adset_id和ad_id添加到fields，以便您可以计算广告中位数和暂停广告。

通过这种方式，您不需要遍历广告并为每个广告设置API调用。如果洞察数据太大，您可以尝试使用async insights api https://developers.facebook.com/docs/marketing-api/insights/best-practices#asynchronous

使用Python中的Facebook Marketing API v2.9关闭广告

1 个答案: