Skip to main content

Understanding TikTok Benchmark

Learn how Orphex benchmarks your TikTok ads against similar ads, how scores are calculated, and what each metric means.

Written by Furkan Savrun

How to Access TikTok Benchmark

  1. Log in to your Orphex account

  2. From the left sidebar, click Creative Performance

  3. Select Benchmark from the dropdown menu

⚠️ TikTok Benchmark is only available for accounts with an active TikTok integration. If you don't see Benchmark in the menu, go to Settings → Connections and make sure your TikTok account is connected.


What is TikTok Benchmark?

  • TikTok Benchmark shows how your TikTok ads perform compared to similar ads running on TikTok.

  • Each metric is scored on a percentile scale from 0% to 100%.

  • A score of 60% means your ad outperforms 60% of comparable ads in terms of targeting, format, and placement.


Score Ranges

Score

Label

What it means

0–19%

Poor

Significantly below average

20–39%

Below Avg

Room for improvement

40–59%

Average

In line with similar ads

60–79%

Good

Above average performance

80–100%

Excellent

Top performing ads


Which ads are included?

Only ads with at least 1,000 impressions within the selected time window are included in the benchmark. Ads below this threshold are automatically excluded. This ensures scores are statistically meaningful.


Available Metrics

The benchmark table includes the following TikTok metrics:

Cost, Impressions, Clicks, CPC, CPM, CPV, CTR, CVR, CPA, Video Views, Video Watched 2s, Video Watched 6s, Video Watched 25% / 50% / 75% / 100%, Average Video Play Rate, Profile Visit Rate, Page View Rate.


How is the data updated?

Benchmark data is refreshed daily. Due to TikTok's data pipeline, there is a 30–48 hour latency on scores. The last update date is shown at the top of the page.


Compare by: choosing your comparison dimensions

The Compare by picker at the top-right of the benchmark table controls which ads TikTok uses to build the comparison pool. The more dimensions you enable, the more relevant the benchmark — but the smaller the sample becomes.

Four dimensions are available:

  • Location : Only compares against ads delivered in the same country or region. Recommended for most cases, since CPM and CTR vary heavily by market. Turkish-market ads compared against US-market ads would produce misleading benchmarks.

  • Placement : Only compares against ads running on the same placement (TikTok feed, Pangle network, etc.). Useful if you're running Pangle, since Pangle benchmarks differ substantially from in-feed TikTok placements.

  • Interest Category : Only compares against ads targeting the same TikTok interest category (for example, Beauty & Personal Care, Apparel & Accessories, Travel). This is TikTok's audience interest taxonomy, not your industry classification — so a jewelry brand targeting "Apparel & Accessories" will be compared against other ads targeting that interest, not just other jewelry brands.

  • Conversion Event : Only compares against ads optimizing for the same conversion event, such as Purchase, Complete Payment, or Add to Cart. Important for conversion campaigns, since CPA expectations differ substantially between event types.

By default, all four dimensions are enabled. If you see a low confidence warning on a metric, the comparison pool may be too small — try disabling one dimension at a time, usually Conversion Event or Interest Category first.


Frequently Asked Questions

  • Why are some of my ads not showing up?

    • Ads with fewer than 1,000 impressions in the selected time window are excluded from the benchmark table. Try selecting a longer time window to include more ads.

  • Can I compare my performance against a specific competitor?

    • No — TikTok Benchmark compares your ads against all similar ads on TikTok based on targeting and format. Competitor-specific comparison is not available here.

  • Does changing the time window affect the benchmark scores?

    • Yes. TikTok recalculates percentile scores based on the selected window, so scores may differ between time ranges.

Did this answer your question?