Top Ways to Analyze TikTok Ads Data (Manually and Automatically)

There are four common ways to analyze TikTok Ads data: the reports in TikTok Ads Manager, a CSV export into a spreadsheet, a scheduled pipeline into a BI tool or warehouse, and an AI assistant that queries the data for you. Each one fits a different question, and most teams use two of them.

This guide covers the metrics worth tracking, what each method can and cannot do, and how to set up the automated routes with Windsor.ai.

What is TikTok Ads analytics?

TikTok Ads analytics is the measurement of what your paid TikTok campaigns deliver: reach, clicks, video plays, conversions and revenue, split by campaign, ad group, ad, audience and placement. The goal is to see which creatives, audiences and bids produce results at an acceptable cost.

Key TikTok Ads metrics to track

These are the metrics most analysis starts from. The field names in brackets are the ones Windsor.ai uses for TikTok Ads; the connector has 954 fields in total (see the TikTok Ads field reference).

• Spend (spend): what the campaign cost in the period, the base for every cost metric.
• Impressions, reach and frequency (impressions, reach, frequency): how often ads were shown, to how many people, and how many times each person saw them on average.
• CTR (ctr): clicks divided by impressions. Windsor.ai returns it as a fraction, so 0.0026 means 0.26%.
• CPC and CPM (cpc, cpm): spend divided by clicks, and spend per 1,000 impressions.
• Conversions and cost per conversion (conversion, cost_per_conversion): the optimization events TikTok attributes to your ads, and what each one cost.
• Purchases and ROAS (complete_payment, complete_payment_roas): purchases TikTok attributes to your ads, and purchase value divided by spend.
• Video plays (play_duration_2s, play_duration_6s, focused_view_6s, focused_view_15s, average_video_play): plays of at least 2 or 6 seconds, focused views and average play time. Divide 2-second plays by impressions to get a hook rate.

Plays at 25%, 50%, 75% and 100% of each video are in play_first_quartile, play_midpoint, play_third_quartile and play_over. Divide each by play_duration_2s to see where viewers stop watching, and play_over by impressions for a completion rate.

Way 1. Reports in TikTok Ads Manager

TikTok Ads Manager shows campaign, ad group and ad results, with breakdowns such as age, gender, placement and country, and custom reports you can save. It is the fastest way to check yesterday’s delivery, review a new creative or follow a campaign in its learning phase.

Its limits are about scope. Ads Manager covers TikTok only, so comparing TikTok with Meta Ads or checking its conversions against GA4 or your CRM means exporting from each platform. The numbers also follow TikTok’s attribution settings, which differ from those of other platforms.

Way 2. A CSV export into a spreadsheet

Exporting a report from Ads Manager into Google Sheets or Excel works for a one-off analysis, such as a quarterly creative review. You can add your own formulas, for example hook rate, or a cost per conversion that uses your CRM’s numbers.

For a weekly report, the export has to be repeated each time, with the same columns and date range, and every edit to the sheet risks breaking a formula. If you send the same report every week, a scheduled connection saves that repetition.

Way 3. A scheduled pipeline into a BI tool or warehouse

A connector such as Windsor.ai pulls TikTok Ads data on a schedule into the tool where you build reports: Data Studio (formerly Looker Studio), Google Sheets, Excel, Power BI, Tableau, BigQuery, Snowflake and other destinations. TikTok Ads then sits next to Meta Ads, Google Ads, GA4 or your CRM in the same report, with field names that follow one schema across platforms.

How often the data refreshes depends on your Windsor.ai plan: daily on the free forever plan and the Basic plan, hourly from Standard, and every 15 minutes on the larger plans. The free forever plan (no credit card) covers 1 source and the last 30 days; TikTok Ads next to Meta Ads, or a longer history, needs a paid plan from $19 a month for 3 sources. See pricing for the current limits.

How to set it up

1. Connect TikTok Ads. Create a Windsor.ai account, select TikTok Ads as a data source and sign in with your TikTok for Business account. Select the advertiser accounts you want to report on. For cross-channel reports, connect Meta Ads, Google Ads, GA4 or your CRM as well.

2. Choose fields. Pick the dimensions and metrics for your report, such as campaign, ad group, ad name, placement, spend, conversions and the video play fields, and preview the table before you send it anywhere.

3. Choose a destination. Select the BI tool, spreadsheet or warehouse and follow its setup steps; the destination guides cover each one.

4. Start from a template if you want one. The TikTok Ads dashboard templates include a Data Studio TikTok Ads overview report and a Tableau TikTok Ads dashboard, plus paid media and paid social templates for Power BI that combine TikTok with other ad platforms.

For analysis beyond 30 days in a warehouse, TikTok Ads to BigQuery keeps every day you load, so you can compare creatives and seasons over longer periods.

Way 4. An AI assistant

An AI assistant answers questions about your TikTok Ads data in plain language and can calculate rates, compare periods and explain changes in the same chat. Two connectors are listed in Claude’s directory:

• TikTok for Business is TikTok’s own verified connector, added in August 2026. It is free, covers TikTok Ads only and has tools that create and edit Smart+ campaigns, ad groups and ads.
• Windsor.ai puts TikTok Ads next to Meta Ads, GA4, BigQuery, HubSpot and other sources in one chat. Claude can pause or enable campaigns, ad groups and ads and set budgets after you approve each change, but it cannot create campaigns or ads.

The setup and prompts are in how to connect TikTok Ads to Claude; for ChatGPT, see how to connect TikTok Ads to ChatGPT.

When we tested the Windsor.ai connector on Windsor.ai’s own TikTok test accounts, two details stood out: the ctr field came back as a fraction (0.0026 for 0.26%), and campaigns with no spend were listed with zeros in every metric. Ask for CTR as a percentage, and for campaigns with spend above zero, to keep answers clean. An assistant reads the numbers and names of each ad; it does not watch the videos, so descriptive ad names help it group creatives by format or hook.

A prompt that works with the fields above:

For each ad in the last 7 days, show spend, impressions, play_duration_2s divided by impressions as the hook rate, conversions and cost per conversion. Which ads stop the scroll but do not convert?

Which method to use

• Daily checks on one account: TikTok Ads Manager.
• A one-off deep dive with your own formulas: a CSV export into a spreadsheet.
• A weekly or client report, or TikTok next to other channels: a scheduled pipeline into Data Studio, Power BI, Sheets or BigQuery.
• Ad hoc questions across TikTok, Meta Ads and GA4: an AI assistant with the Windsor.ai connector.
• Building TikTok campaigns from a chat: TikTok’s own connector in Claude.

Checks before you act on a number

• Compare one total with Ads Manager for the same dates and the same time zone before you trust a breakdown.
• Name the formula for every rate you calculate, such as hook rate as 2-second plays divided by impressions, so reports stay comparable.
• Expect TikTok-reported conversions to differ from GA4 or your CRM, because each counts with its own attribution rules. Compare them side by side; do not add them up.

Conclusion

Ads Manager and a spreadsheet cover quick checks and one-off reviews. A scheduled pipeline is the better fit for recurring reports and for TikTok next to other channels, and an AI assistant answers questions across those sources without building a report first.

🚀 Connect TikTok Ads to Windsor.ai and build your first report: start with the free forever plan.