diff --git a/.gitignore b/.gitignore
index 76d0f357..95de70d4 100644
--- a/.gitignore
+++ b/.gitignore
@@ -24,3 +24,4 @@ un_sdg/metadata/
 *.Rproj
 .Rproj.user
 venv
+credentials/
diff --git a/requirements.txt b/requirements.txt
index e3d11fa1..57b17ae5 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -15,4 +15,6 @@ black==21.8b0
 flake8==3.9.2
 xlrd==2.0.1
 pytest==6.2.5
-mypy==0.910
\ No newline at end of file
+mypy==0.910
+google-auth-oauthlib==0.4.6
+google-api-python-client==2.19.1
diff --git a/site_analytics/README.md b/site_analytics/README.md
new file mode 100644
index 00000000..6fd8ab87
--- /dev/null
+++ b/site_analytics/README.md
@@ -0,0 +1,49 @@
+# Site analytics
+
+Exposes simple site analytics, such as page views by slug URL on the [ourworldindata.org](https://ourworldindata.org) website. For internal use by OWID staff only.
+
+This service uses the [Google Analytics Reporting API v4](https://developers.google.com/analytics/devguides/reporting/core/v4). 
+
+## Setup
+
+In order to use this service, you must:
+
+1. Have a client secrets file located in `site_analytics/config/credentials/owid-analytics-client-secrets.json`. 
+2. Add the following variables to `.env` at the root of this repository:
+
+```bash
+GA_VIEW_ID = "{GA_VIEW_ID}"
+GA_ACCOUNT_ID = "{GA_ACCOUNT_ID}"
+GA_PROPERTY_ID = "{GA_PROPERTY_ID}"
+```
+
+Ask an OWID developer for these files/variables (available to OWID staff only).
+
+
+## Examples
+
+Execute a single simple [Google Analytics report request](https://developers.google.com/analytics/devguides/reporting/core/v4/rest/v4/reports/batchGet) with one metric and one dimension over one time range:
+
+```python
+from site_analytics.request_report import execute_report_request
+
+df = execute_report_request(
+  metric="pageviews",
+  dimension="pagePath",
+  start_date="7daysAgo",
+  end_date="yesterday",
+  filters_expression="ga:dimension1==0"  # excludes views of embedded pages
+)
+print(df.tail(5))
+#       start_date   end_date             pagePath  pageviews
+# 41932   7daysAgo  yesterday         /covid-cases    90053.0
+# 41933   7daysAgo  yesterday  /covid-vaccinations   123601.0
+# 41934   7daysAgo  yesterday         /coronavirus   167887.0
+# 41935   7daysAgo  yesterday                    /   176392.0
+# 41936   7daysAgo  yesterday  /covid-vaccinations   711477.0
+
+```
+
+For acceptable values of the `metric`, `dimension`, etc parameters, check out the [Google Analytics UA query explorer](https://ga-dev-tools.web.app/query-explorer/).
+
+> Note: set `filters_expression="ga:dimension1==0"` to exclude visits to embedded OWID pages.
diff --git a/site_analytics/__init__.py b/site_analytics/__init__.py
new file mode 100644
index 00000000..32f3d8b4
--- /dev/null
+++ b/site_analytics/__init__.py
@@ -0,0 +1,6 @@
+import os
+
+CURRENT_DIR = os.path.dirname(__file__).split("/")[-1]
+CONFIGPATH = os.path.join(CURRENT_DIR, "config")
+CREDSPATH = os.path.join(CONFIGPATH, "credentials")
+CLIENT_SECRETS_PATH = os.path.join(CREDSPATH, "owid-analytics-client-secrets.json")
diff --git a/site_analytics/request_report.py b/site_analytics/request_report.py
new file mode 100644
index 00000000..42ccb80c
--- /dev/null
+++ b/site_analytics/request_report.py
@@ -0,0 +1,84 @@
+import os
+import pandas as pd
+from typing import Optional
+from dotenv import load_dotenv
+from apiclient.discovery import build
+from google.oauth2.credentials import Credentials
+
+from site_analytics.utils import google_analytics_authenticate
+
+load_dotenv()
+
+GA_ACCOUNT_ID = os.getenv("GA_ACCOUNT_ID")
+GA_PROPERTY_ID = os.getenv("GA_PROPERTY_ID")
+GA_VIEW_ID = os.getenv("GA_VIEW_ID")
+
+
+def execute_report_request(
+    metric: str,
+    dimension: str,
+    start_date: str,
+    end_date: str,
+    filters_expression: Optional[str] = None,
+    credentials: Optional[Credentials] = None,
+) -> pd.DataFrame:
+    """Request a single simple Google Analytics report with one metric and one
+    dimension in a single time range.
+
+    Uses the Google Analytics Reporting API v4. See
+    https://developers.google.com/analytics/devguides/reporting/core/v4/rest/v4/reports/batchGet
+    for valid values of the `metric`, `dimension`, `start_date`, etc parameters.
+    """
+    if not credentials:
+        credentials = google_analytics_authenticate()
+    analytics = build("analyticsreporting", "v4", credentials=credentials)
+    ga_rows = []
+    report_request = {
+        "viewId": f"ga:{GA_VIEW_ID}",
+        "dateRanges": [{"startDate": start_date, "endDate": end_date}],
+        "metrics": [{"expression": f"ga:{metric}"}],
+        "dimensions": [
+            {"name": f"ga:{dimension}"},
+        ],
+        "pageSize": "100000",
+        "orderBys": [
+            {
+                "fieldName": f"ga:{metric}",
+                "sortOrder": "DESCENDING",
+            }
+        ],
+    }
+    if filters_expression:
+        report_request["filtersExpression"] = filters_expression
+    response = (
+        analytics.reports()
+        .batchGet(body={"reportRequests": [report_request]})
+        .execute()
+    )
+    for row in response["reports"][0]["data"]["rows"]:
+        dims = row["dimensions"]
+        assert "(other)" not in dims
+        val = row["metrics"][0]["values"][0]
+        ga_rows.append([start_date, end_date] + dims + [val])
+    while response["reports"][0].get("nextPageToken"):
+        report_request["pageToken"] = response["reports"][0].get("nextPageToken")
+        response = (
+            analytics.reports()
+            .batchGet(body={"reportRequests": [report_request]})
+            .execute()
+        )
+        for row in response["reports"][0]["data"]["rows"]:
+            dims = row["dimensions"]
+            assert "(other)" not in dims
+            val = row["metrics"][0]["values"][0]
+            ga_rows.append([start_date, end_date] + dims + [val])
+    df = pd.DataFrame(
+        ga_rows,
+        columns=["start_date", "end_date", dimension, metric],
+    )
+    try:
+        df[metric] = df[metric].astype(float)
+    except:
+        pass
+    df = df.sort_values(by=metric).reset_index(drop=True)
+    return df
diff --git a/site_analytics/test_request_report.py b/site_analytics/test_request_report.py
new file mode 100644
index 00000000..1c06d521
--- /dev/null
+++ b/site_analytics/test_request_report.py
@@ -0,0 +1,56 @@
+import pytest
+
+import os
+import pandas as pd
+from apiclient.discovery import build
+
+from site_analytics.utils import google_analytics_authenticate
+from site_analytics.request_report import execute_report_request
+
+from dotenv import load_dotenv
+
+load_dotenv()
+
+GA_ACCOUNT_ID = os.getenv("GA_ACCOUNT_ID")
+GA_PROPERTY_ID = os.getenv("GA_PROPERTY_ID")
+
+
+@pytest.fixture(scope="module")
+def credentials():
+    yield google_analytics_authenticate()
+
+
+def test_dimension1_name(credentials):
+    nm = (
+        build("analytics", "v3", credentials=credentials)
+        .management()
+        .customDimensions()
+        .get(
+            accountId=GA_ACCOUNT_ID,
+            webPropertyId=GA_PROPERTY_ID,
+            customDimensionId="ga:dimension1",
+        )
+        .execute()["name"]
+    )
+    assert nm == "Page is embedded"
+
+
+@pytest.fixture(scope="module")
+def df_simple_request(credentials):
+    df = execute_report_request(
+        metric="pageviews",
+        dimension="pagePath",
+        start_date="yesterday",
+        end_date="yesterday",
+        filters_expression=None,
+        credentials=credentials,
+    )
+    yield df
+
+
+def test_is_frame(df_simple_request):
+    assert type(df_simple_request) == pd.DataFrame
+
+
+def test_gt_zero_rows(df_simple_request):
+    assert df_simple_request.shape[0] > 0
diff --git a/site_analytics/utils.py b/site_analytics/utils.py
new file mode 100644
index 00000000..f9a6bb31
--- /dev/null
+++ b/site_analytics/utils.py
@@ -0,0 +1,16 @@
+from google_auth_oauthlib.flow import InstalledAppFlow
+from site_analytics import CLIENT_SECRETS_PATH
+
+
+def google_analytics_authenticate():
+    flow = InstalledAppFlow.from_client_secrets_file(
+        CLIENT_SECRETS_PATH,
+        scopes=[
+            "openid",
+            "https://www.googleapis.com/auth/userinfo.email",
+            "https://www.googleapis.com/auth/analytics.readonly",
+        ],
+    )
+    flow.run_local_server()
+    credentials = flow.credentials
+    return credentials
diff --git a/standard_importer/posts_to_update.py b/standard_importer/posts_to_update.py
new file mode 100644
index 00000000..99ca159d
--- /dev/null
+++ b/standard_importer/posts_to_update.py
@@ -0,0 +1,184 @@
+"""Constructs a dataframe that counts the number of charts per OWID page that
+have been updated.
+
+Each row is in the dataframe is an OWID page.
+
+The exported csv file is intended to make it easy for OWID staff to see which
+public OWID pages need to be checked for textual inconsistencies after one or
+more suggested chart revisions have been approved as part of a bulk dataset
+update.
+
+Example:
+
+    >>> from standard_importer import posts_to_update
+    >>> df = posts_to_update.main(dataset_id=5357, since="2021-08-01", include_google_analytics=True)
+    >>> print(df.head())
+                             post_title                                          post_slug  num_charts_updated  pageviews_from_20210608_to_20210608
+    1                Women's employment     https://ourworldindata.org/female-labor-supply                   2                               9755.0
+    2  Working women: Key facts and ...  https://ourworldindata.org/female-labor-force-...                   2                               6250.0
+    0           Teachers and Professors  https://ourworldindata.org/teachers-and-profes...                   2                               1894.0
+
+"""
+
+import os
+import re
+import time
+import datetime as dt
+import grequests
+import json
+import pandas as pd
+import numpy as np
+import logging
+from tqdm import tqdm
+from dotenv import load_dotenv
+
+from db import get_connection
+from utils import assert_admin_api_connection, batchify
+from site_analytics.request_report import execute_report_request
+
+load_dotenv()
+
+logging.basicConfig()
+logger = logging.getLogger(__name__)
+logger.setLevel(logging.INFO)
+
+
+# config for google analytics report request
+GA_CONFIG = {
+    "metric": "pageviews",
+    "dimension": "pagePath",
+    "filters_expression": "ga:dimension1==0",  # exclude embedded page views
+    "start_date": (dt.datetime.utcnow() - dt.timedelta(days=90)).strftime("%Y-%m-%d"),
+    "end_date": (dt.datetime.utcnow() - dt.timedelta(days=1)).strftime("%Y-%m-%d"),
+}
+
+SITE_SESSION_ID = os.getenv("SITE_SESSION_ID")
+SITE_HOST = os.getenv("SITE_HOST")
+DEBUG = os.getenv("DEBUG") == "True"
+
+
+def main(
+    dataset_id: int, since: str, include_google_analytics: bool = False
+) -> pd.DataFrame:
+    """Constructs a dataframe that counts the number of charts from a dataset
+    per OWID page that have been updated.
+    """
+    if include_google_analytics:
+        df_ga = get_ga_data()
+
+    df = get_charts_updated_data(dataset_id, since)
+
+    if include_google_analytics:
+        metric = GA_CONFIG["metric"]
+        dimension = GA_CONFIG["dimension"]
+        start_date = GA_CONFIG["start_date"]
+        end_date = GA_CONFIG["start_date"]
+        df = df.merge(
+            df_ga[[dimension, metric]],
+            left_on="post_slug",
+            right_on=dimension,
+            how="left",
+            validate="1:1",
+        ).drop(columns=[dimension])
+        df.sort_values(by=[metric, "num_charts_updated"], ascending=False, inplace=True)
+        df.rename(
+            columns={
+                metric: f"{metric}_from_{re.sub('-', '', start_date)}_to_{re.sub('-', '', end_date)}"
+            },
+            inplace=True,
+        )
+    else:
+        df.sort_values(by=["num_charts_updated"], ascending=False, inplace=True)
+
+    df["post_slug"] = "https://ourworldindata.org/" + df["post_slug"]
+    return df
+
+
+def get_ga_data() -> pd.DataFrame:
+    df_ga = execute_report_request(**GA_CONFIG)
+    assert (df_ga[GA_CONFIG["dimension"]] == "(other)").sum() == 0
+    if "pagePath" in df_ga.columns:
+        df_ga["pagePath"] = (
+            df_ga["pagePath"]
+            .str.replace(r"\?.*", "", regex=True)
+            .str.replace(r"/?\.?$", "", regex=True)
+            .str.replace(r"^/", "", regex=True)
+        )
+        df_ga = (
+            df_ga.groupby(["start_date", "end_date", GA_CONFIG["dimension"]])[
+                GA_CONFIG["metric"]
+            ]
+            .sum()
+            .reset_index()
+        )
+
+    df_ga.sort_values(GA_CONFIG["metric"], ascending=False, inplace=True)
+    return df_ga
+
+
+def get_charts_updated_data(dataset_id: int, since: str) -> pd.DataFrame:
+    """returns a dataframe of the number of charts updated since YYYY-MM-DD, by page."""
+    # retrieves all charts that use a variable from {dataset} and have been updated.
+    assert_admin_api_connection()
+    logger.info("Retrieving updated charts...")
+    df = pd.read_sql(
+        f"""
+        SELECT
+            charts.id, charts.updatedAt, charts.createdAt, charts.lastEditedAt, charts.publishedAt
+        FROM charts
+        INNER JOIN chart_dimensions
+        ON charts.id = chart_dimensions.chartId
+        WHERE variableId IN (
+            SELECT id
+            FROM variables
+            WHERE datasetId = {dataset_id}
+        )
+            AND charts.updatedAt >= "{since}"
+        ORDER BY updatedAt DESC
+    """,
+        get_connection(),
+    )
+    if DEBUG:
+        logger.warning(
+            "DEBUG mode is on. Only retrieving post references for the first "
+            "10 updated charts."
+        )
+        df = df.iloc[:10]
+
+    logger.info("Retrieving chart references to OWID pages...")
+    wait = 2
+    batch_size = 50
+    n_batches = int(np.ceil(df.shape[0] / batch_size))
+
+    responses = []
+    for batch in tqdm(batchify(df, batch_size=batch_size), total=n_batches):
+        requests = []
+        for _, row in batch.iterrows():
+            # refs = get_references_by_chart_id(row.id)["references"]
+            url = f"{SITE_HOST}/admin/api/charts/{row.id}.references.json"
+            res = grequests.get(url, cookies={"sessionid": SITE_SESSION_ID})
+            requests.append(res)
+        responses += grequests.map(requests)
+        time.sleep(wait)
+
+    references = []
+    for resp in responses:
+        refs = json.loads(resp.content)["references"]
+        for ref in refs:
+            ref["chartId"] = row.id
+        references += refs
+
+    # df_refs = pd.read_csv('refs.csv')
+    df_refs = pd.DataFrame(references).rename(columns={"id": "refId"})
+    df = (
+        df_refs.value_counts(["title", "slug"])
+        .reset_index()
+        .rename(
+            columns={
+                0: "num_charts_updated",
+                "title": "post_title",
+                "slug": "post_slug",
+            }
+        )
+    )
+    return df
diff --git a/utils.py b/utils.py
index 0a836756..33ce3780 100644
--- a/utils.py
+++ b/utils.py
@@ -1,7 +1,16 @@
+import os
 import re
+import json
+import requests
 from typing import Any, Generator, List, Collection
 from dataclasses import dataclass, field
 
+from dotenv import load_dotenv
+
+load_dotenv()
+SITE_HOST = os.getenv("SITE_HOST")
+SITE_SESSION_ID = os.getenv("SITE_SESSION_ID")
+
 
 def write_file(file_path, content):
     with open(file_path, "w") as f:
@@ -94,3 +103,22 @@ def from_values(xs: List[int]):
 
     def to_values(self):
         return [self.min, self.max]
+
+
+def assert_admin_api_connection() -> None:
+    """raises an AssertionError if unable to successfully connect to the admin API."""
+    res = False
+    try:
+        charts = json.loads(
+            requests.get(
+                f"{SITE_HOST}/admin/api/charts.json?limit=1",
+                cookies={"sessionid": SITE_SESSION_ID},
+            ).content
+        )
+        res = len(charts["charts"]) > 0
+    except Exception:
+        res = False
+    assert res, (
+        "Failed to connect to admin API, have you set SITE_HOST and "
+        "SITE_SESSION_ID correctly in .env?"
+    )