Examine Trends over Time

{
  "base_audience": {
    "id": "1",
    "name": "All Internet Users",
    "expression": null
  },
  "locations": [
    "s2_1"
  ],
  "waves": [
    "q1_2024",
    "q2_2024",
    "q3_2024",
    "q4_2024"
  ],
  "rows": [
    {
      "id": "1",
      "name": "fb_boomers",
      "expression": {
        "and": [
          {
            "question": "q42011a",
            "datapoints": [
              "q42011a_3"
            ],
            "suffixes": [
              1,
              2
            ]
          },
          {
            "question": "q3new",
            "datapoints": [
              "q3new_1"
            ]
          }
        ]
      }
    },
    {
      "id": "2",
      "name": "fb_millennials",
      "expression": {
        "and": [
          {
            "question": "q42011a",
            "datapoints": [
              "q42011a_3"
            ],
            "suffixes": [
              1,
              2
            ]
          },
          {
            "question": "q3new",
            "datapoints": [
              "q3new_3"
            ]
          }
        ]
      }
    }
  ],
  "columns": [
    {
      "id": "1",
      "name": "Q1 2024",
      "expression": {
        "question": "waves",
        "datapoints": [
          "q1_2024"
        ]
      }
    },
    {
      "id": "2",
      "name": "Q2 2024",
      "expression": {
        "question": "waves",
        "datapoints": [
          "q2_2024"
        ]
      }
    },
    {
      "id": "3",
      "name": "Q3 2024",
      "expression": {
        "question": "waves",
        "datapoints": [
          "q3_2024"
        ]
      }
    },
    {
      "id": "4",
      "name": "Q4 2024",
      "expression": {
        "question": "waves",
        "datapoints": [
          "q4_2024"
        ]
      }
    }
  ]
}

import requests
import json
import pandas as pd

with open('req.json', 'r') as file:
    request_json = json.load(file)

url = 'https://api.globalwebindex.com/v2/query/crosstab'
headers = {"Authorization": "Bearer YOUR_API_KEY"}

response = requests.post(url, json=request_json, headers=headers)
response.raise_for_status()

# Process the response (JSON Lines format)
results = []
for line in response.iter_lines():
    if line:
        obj = json.loads(line)
        if obj.get('row_index') and obj.get('column_index'):  # Check for non-zero index values
            row_name = obj['row']['name']
            column_name = obj['column']['name']
            percentage = obj['intersect']['percentage']
            results.append((row_name, column_name, percentage))

# Convert the results to a DataFrame and pivot the table for readability
df = pd.DataFrame(results, columns=['Row Name', 'Column Name', 'Percentage'])
pivot_table = df.pivot(index='Row Name', columns='Column Name', values='Percentage').reset_index()
print(pivot_table.to_string(index=False))

{
  "row": {
    "id": "1",
    "name": "fb_boomers",
    "expression": {
      "and": [
        {
          "question": "q42011a",
          "options": [
            "q42011a_3"
          ],
          "suffixes": [
            1,
            2
          ]
        },
        {
          "question": "q3new",
          "options": [
            "q3new_1"
          ]
        }
      ]
    }
  },
  "row_index": 1,
  "column": {
    "id": "1",
    "name": "Q1 2024",
    "expression": {
      "question": "waves",
      "options": [
        "q1_2024"
      ]
    }
  },
  "column_index": 1,
  "audiences": {
    "column": {
      "audience": "1",
      "size": 257758496,
      "sample": 24922,
      "percentage": 100,
      "intersect_percentage": 100
    },
    "row": {
      "audience": "1",
      "size": 36643034,
      "sample": 3728,
      "percentage": 14.2,
      "intersect_percentage": 14.2
    }
  },
  "base": {
    "size": 257758496,
    "sample": 24922
  },
  "intersect": {
    "size": 36643034,
    "sample": 3728,
    "percentage": 14.2,
    "index": 100
  }
}
{}
{}
...

     Row Name  Q1 2024  Q2 2024  Q3 2024  Q4 2024
   fb_boomers     14.2     14.8     15.0     15.5
fb_millenials     16.5     15.9     15.5     15.1

To explore how behaviours or preferences shift over time, a powerful method is to use a Crosstab analysis. A Crosstab structures data into a matrix where rows or columns represent audiences, specific data points, or variables like time periods or locations. That way, Crosstabs allow you to examine large amounts of data and spot trends, compare audience behaviours, and generate actionable insights. For example, using GWI’s API, you can track how Facebook usage among Boomers and Millennials has changed over the last four quarters. Such analysis helps identify behavioural shifts and refine marketing strategies based on evolving trends. You can use the waves question and its datapoints in Crosstab columns to see the shift in trends over time. For that, you can set appropriate audiences (we’ll call them fb_boomers and fb_millennials) as rows and the waves of the last 4 quarters as columns:

curl -X POST "https://api.globalwebindex.com/v2/saved/crosstabs" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d @req.json

The response is potentially a large JSON Lines file but you can extract the relevant data from response with a script like the one provided in examples.

{
  "base_audience": {
    "id": "1",
    "name": "All Internet Users",
    "expression": null
  },
  "locations": [
    "s2_1"
  ],
  "waves": [
    "q1_2024",
    "q2_2024",
    "q3_2024",
    "q4_2024"
  ],
  "rows": [
    {
      "id": "1",
      "name": "fb_boomers",
      "expression": {
        "and": [
          {
            "question": "q42011a",
            "datapoints": [
              "q42011a_3"
            ],
            "suffixes": [
              1,
              2
            ]
          },
          {
            "question": "q3new",
            "datapoints": [
              "q3new_1"
            ]
          }
        ]
      }
    },
    {
      "id": "2",
      "name": "fb_millennials",
      "expression": {
        "and": [
          {
            "question": "q42011a",
            "datapoints": [
              "q42011a_3"
            ],
            "suffixes": [
              1,
              2
            ]
          },
          {
            "question": "q3new",
            "datapoints": [
              "q3new_3"
            ]
          }
        ]
      }
    }
  ],
  "columns": [
    {
      "id": "1",
      "name": "Q1 2024",
      "expression": {
        "question": "waves",
        "datapoints": [
          "q1_2024"
        ]
      }
    },
    {
      "id": "2",
      "name": "Q2 2024",
      "expression": {
        "question": "waves",
        "datapoints": [
          "q2_2024"
        ]
      }
    },
    {
      "id": "3",
      "name": "Q3 2024",
      "expression": {
        "question": "waves",
        "datapoints": [
          "q3_2024"
        ]
      }
    },
    {
      "id": "4",
      "name": "Q4 2024",
      "expression": {
        "question": "waves",
        "datapoints": [
          "q4_2024"
        ]
      }
    }
  ]
}

import requests
import json
import pandas as pd

with open('req.json', 'r') as file:
    request_json = json.load(file)

url = 'https://api.globalwebindex.com/v2/query/crosstab'
headers = {"Authorization": "Bearer YOUR_API_KEY"}

response = requests.post(url, json=request_json, headers=headers)
response.raise_for_status()

# Process the response (JSON Lines format)
results = []
for line in response.iter_lines():
    if line:
        obj = json.loads(line)
        if obj.get('row_index') and obj.get('column_index'):  # Check for non-zero index values
            row_name = obj['row']['name']
            column_name = obj['column']['name']
            percentage = obj['intersect']['percentage']
            results.append((row_name, column_name, percentage))

# Convert the results to a DataFrame and pivot the table for readability
df = pd.DataFrame(results, columns=['Row Name', 'Column Name', 'Percentage'])
pivot_table = df.pivot(index='Row Name', columns='Column Name', values='Percentage').reset_index()
print(pivot_table.to_string(index=False))

{
  "row": {
    "id": "1",
    "name": "fb_boomers",
    "expression": {
      "and": [
        {
          "question": "q42011a",
          "options": [
            "q42011a_3"
          ],
          "suffixes": [
            1,
            2
          ]
        },
        {
          "question": "q3new",
          "options": [
            "q3new_1"
          ]
        }
      ]
    }
  },
  "row_index": 1,
  "column": {
    "id": "1",
    "name": "Q1 2024",
    "expression": {
      "question": "waves",
      "options": [
        "q1_2024"
      ]
    }
  },
  "column_index": 1,
  "audiences": {
    "column": {
      "audience": "1",
      "size": 257758496,
      "sample": 24922,
      "percentage": 100,
      "intersect_percentage": 100
    },
    "row": {
      "audience": "1",
      "size": 36643034,
      "sample": 3728,
      "percentage": 14.2,
      "intersect_percentage": 14.2
    }
  },
  "base": {
    "size": 257758496,
    "sample": 24922
  },
  "intersect": {
    "size": 36643034,
    "sample": 3728,
    "percentage": 14.2,
    "index": 100
  }
}
{}
{}
...

     Row Name  Q1 2024  Q2 2024  Q3 2024  Q4 2024
   fb_boomers     14.2     14.8     15.0     15.5
fb_millenials     16.5     15.9     15.5     15.1

Compare Results by Location Guide

⌘I

Getting started

Use Cases

Transition from V1 to V2 API

Platform API

Examine Trends over Time