Getting Started: FreeWheel Setup Guide

This guide walks you through setting up Watching That with FreeWheel from start to finish — what you'll end up with, how to get there, and what you can do on day one.

Total hands-on time: under 1 hour. Typical time to data in platform: under a week.

Related pages:

• Metrics, Dimensions & Filters — platform-wide concepts

• FreeWheel: Metrics & Dimensions — event types and dimensions overview

• FreeWheel Metrics Reference — complete standard metrics catalogue

• FreeWheel Log File Fields — raw field reference

On This Page

What You'll End Up With

By the end of this setup you'll have:

• FreeWheel v4 log files flowing into Watching That continuously — every ad request, impression, error, and completion event

• Optionally, bid-level logs for full programmatic auction visibility

• FreeWheel API access enriching your log data with friendly names, campaign metadata, and configuration details — powering Validation Monitors

• Up to 10 event types and up to 50 data fields configured for your organisation — determining your metrics and dimensions

• Reports, Analysis, Inspect, and Monitoring all working across your FreeWheel data from the moment data starts flowing

Step 1: Grant Log File Access

Once done, send your AWS Account ID and FreeWheel Network ID to your Watching That contact. We configure the ingestion on our side.

Optional — Bid-Level Logs: If your FreeWheel account includes bid-level logs and you want programmatic auction visibility, the same CloudFormation template covers both. No additional setup. See FreeWheel Bid Level Log Fields for the field reference.

Step 2: Set Up API Access

We provide the exact form answers to give FreeWheel — no guesswork.

What changes immediately: Friendly names throughout the platform instead of numeric IDs. Validation Monitors become available — automated business rule checks running every ~10 minutes against your FreeWheel configuration.

Step 3: Configure Your Event Types and Data Fields

Your account team works with you to select the right configuration. This is a configuration conversation, not a technical project.

Choosing Your Event Types (up to 10 of 16 available)

A typical selection of 10 covers the full ad delivery lifecycle plus bid-level analytics:

Left for later: Three mid-quartile events (FIRST QUARTILE, MIDPOINT, SECOND QUARTILE) for granular viewability, SERVER RESPONSE for demand partner response analysis, SKIP/MUTE for engagement tracking.

Trade-off: If you don't need bid-level data, those four slots free up for quartile events + SERVER RESPONSE — full viewability and demand partner analysis instead of auction visibility.

Choosing Your Data Fields (up to 50 of 60+ available)

Left for later: Audience fields, content detail (genreId, programmerId, tvRating), privacy (privacyChoices), custom key-values (AllRequestKV), additional identity fields. Swap in as priorities shift.

For the complete field reference, see FreeWheel Log File Fields. Fields marked ⚡ are available on request.

Step 4: Data Starts Flowing

Once Steps 1–3 are complete, Watching That begins ingesting your log files. No deployment, no code, no agents to install.

What happens automatically:

• Log files picked up from your S3 bucket

• Events parsed, enriched with API metadata, indexed

• Metrics and dimensions available in all modules

• Historical data backfilled (your account team confirms the window)

Timeline:

What You Can Do on Day One

With data flowing, everything below works immediately — no additional configuration.

Reports

• Delivery health — Gross Impressions, Error Rate (%), Fill Rate (%) by distributionPartnerId on timeseries. Partner-level trends at a glance.

• Revenue by partner and position — Gross Ad Server Revenue by sellingPartnerId and adType.

• Device and geo splits — Fill Rate (%) by deviceType and country. Platform-specific or regional issues.

• Bid performance — Received Bid Rate and Bid Win Rate by marketBuyerSeatId.

Analysis

• Revenue dropped Tuesday? Gross Impressions by distributionPartnerId → pivot by errorCode → drill into placementId. Minutes, not days.

• Demand partner competitiveness: Bid Participation Rate and Clearing Price by marketBuyerSeatId and marketDealId.

• Live vs VOD: Fill Rate (%) by siteSectionId and isLive.

Monitoring (Anomaly)

• Impression flatline — distributionPartnerId drops to zero → partner outage caught early

• Error spike — Error Rate (%) deviates from 7-day pattern → config change caught

• Revenue anomaly — Gross Ad Server Revenue relative change → CPM drops caught

• Bid timeout spike — Bid Response Error: Timeouts by marketBuyerSeatId → demand partner issues caught

Monitoring (Validation)

With the API connected (Step 2), Validation Monitors check your FreeWheel configuration every ~10 minutes. Your first 3 are free. See the Validation Monitor of the Month series for inspiration.

Inspect

Session-level event viewer. Pull a request ID → see every event in the session: requested, filled, errored, bid on. JSON export. Minutes to resolve.

Custom Metrics

Build your own derived metrics (formulas across base metrics) and filtered metrics (any metric filtered by any dimension) from day one. The FreeWheel Metrics Reference documents the standard starting point — your custom metrics extend from there.

Evolving Your Setup

For pixel deployment, see FreeWheel Pixel Deployment. Event types and fields can be added to your licence at any time.

Summary