Analytics

Pozy collects metrics from every connected account on a schedule and serves them through one set of endpoints. You get per-post numbers, account-level insights, follower history, timing recommendations and inbox activity, all filtered by the same profileId, accountId and platform parameters.

The dashboard's Analytics pages read these same endpoints, so a chart you see in the app is a request you can make yourself.

#What Pozy measures

#One call for an overview screen

GET /v1/analytics/dashboard returns totals, the previous period for comparison, follower growth, a daily series, top posts and recent posts in a single response. It is the fastest way to fill an overview page.

curl "https://api.pozy.ai/v1/analytics/dashboard?profileId=6650c1e2a3b4c5d6e7f80911&fromDate=2026-10-01&toDate=2026-10-29&compare=previous_period" \
  -H "Authorization: Bearer $POZY_API_KEY"

const res = await fetch("https://api.pozy.ai/v1/analytics/dashboard?profileId=6650c1e2a3b4c5d6e7f80911&fromDate=2026-10-01&toDate=2026-10-29&compare=previous_period", {
  headers: { Authorization: `Bearer ${process.env.POZY_API_KEY}` },
});
const data = await res.json();
if (!res.ok) throw new Error(data.error);

console.log(data.totals.impressions, data.previousTotals?.impressions);

import os
import requests

resp = requests.get(
    "https://api.pozy.ai/v1/analytics/dashboard",
    headers={"Authorization": f"Bearer {os.environ['POZY_API_KEY']}"},
    params={
        "profileId": "6650c1e2a3b4c5d6e7f80911",
        "fromDate": "2026-10-01",
        "toDate": "2026-10-29",
        "compare": "previous_period",
    },
    timeout=30,
)
data = resp.json()
print(data["totals"]["impressions"])

{
  "dateRange": { "fromDate": "2026-10-01", "toDate": "2026-10-29" },
  "totals": { "impressions": 48210, "reach": 31987, "likes": 2114, "comments": 302, "shares": 140, "saves": 211, "clicks": 590, "views": 22340, "engagementRate": 5.86 },
  "previousTotals": { "impressions": 40112, "reach": 27650, "likes": 1802, "comments": 251, "shares": 98, "saves": 180, "clicks": 470, "views": 19011, "engagementRate": 5.8 },
  "followers": { "current": 12840, "gained": 312, "byAccount": [{ "accountId": "6650c1e2a3b4c5d6e7f80a22", "platform": "instagram", "current": 12840, "gained": 312 }] },
  "daily": [{ "date": "2026-10-01", "impressions": 1630, "reach": 1210, "engagement": 96, "views": 702, "followersGained": 9 }],
  "topPosts": [],
  "recentPosts": [],
  "dataAsOf": "2026-10-29T22:40:00.000Z"
}

#Two kinds of numbers

The dashboard mixes two views of the same data, and they answer different questions:

• totals, previousTotals and daily count engagement that arrived during the window, on any post. A post from last month that gets 10 likes this week adds 10 to this week. totals always equals the sum of daily.
• topPosts and recentPosts list posts published during the window, each with its lifetime metrics.

So the top posts do not add up to the totals, and they are not meant to. engagementRate is a percentage: likes, comments, shares and saves divided by impressions. Where a platform reports no impressions, Pozy falls back to reach, then views.

#Posts made outside Pozy

Every analytics endpoint takes source:

Pozy discovers external posts on each account about every 90 minutes. When someone pastes the URL of a post they published a minute ago, fetch it right away with POST /v1/posts/sync-external and the accountId plus url. That call is debounced per account, so use it for that moment, not as a bulk refresh. See External posts.

#How fresh the numbers are

Platforms release metrics on their own schedule. Show a "data as of" label instead of promising real time.

Use the freshness fields the API returns: dataAsOf on the dashboard, overview.lastSync and overview.dataStaleness on post lists, syncStatus per platform entry (synced, pending, unavailable) and analytics.lastUpdated per post. Show a spinner on pending rather than a zero.

#Keep your own copy in sync

If you store metrics in your own database, do not poll GET /v1/analytics once per account. Use the change feed:

Taking the cursor before the baseline matters. Changes that land while the baseline loads would otherwise be in neither. The overlap is safe because metrics are absolute values, not increments: a later row for a post replaces an earlier one.

Treat the cursor as opaque. Do not parse it, build it or sort on syncedAt, which can go backwards between entries.

#Get told when an account changes

Subscribe a webhook to analytics.synced. It fires once per account each time its analytics sync finishes, which is roughly hourly per account. The payload names the account and how many posts changed, and it carries no metrics: read the delta feed when it arrives.

On a busy workspace this event is high volume (1,000 connected accounts means about 1,000 events an hour). Give it its own webhook endpoint so a slow consumer cannot get your other events switched off, and return 2xx before you read the feed. See Retries.

The event is skipped when a sync fails, when the account was synced moments ago, when the account is disconnected, or when the platform has no analytics.

#Errors you may see

#Endpoints

#Webhook events