How to read analytics TikTok, Instagram and YouTube: what metrics are important for an affiliate marketer

Views are the affiliate’s drug. 50,000 views on a video feels like a victory, but 800 feels like a failure. But here’s the paradox: a video with 50,000 views and zero CTR brings exactly the same number of conversions as a video with 0 views - zero. Analytics TikTok, Instagram and YouTube store dozens of metrics, but 90% of arbitrage traders look at one - views. And then they wonder why the connection doesn’t pay off. In this article, we analyze all the key metrics of the three platforms through the eyes of an affiliate marketer, not a blogger. We explain what numbers really influence ROI, how to calculate the engagement rate and cost per result, how to conduct A/B testing of creatives using analytics data, and how to determine the winner from five creative options within 48 hours.

TikTok Analytics: metrics that determine the fate of a video

TikTok Analytics is the most powerful native tool among the three platforms. It is available on Pro accounts (free switch in settings) and shows data that Instagram and YouTube hide behind third-party services.

Views - TikTok counts views from the first second of playback. If a user even accidentally stops at your video, this is a +1 view. For an affiliate marketer, this means that the absolute number of views is overestimated: a significant part of the “viewers” ​​swiped after 0.5 seconds. Views is a contextual metric, not a target one. Use it only as a denominator for calculating other indicators.

Watch time (total viewing time) - the total time that all viewers spent on your video. Critical to the algorithm: TikTok promotes videos that keep people on the platform longer. For a 60-second video, watch time 200 hours is a good signal. But watch time itself is not very informative - it needs to be normalized by the number of views.

Average watch time — watch time / views. The main metric of content quality. For a 30-second video, the average time of 12-15 seconds is above average. If the average watch time is below 5 seconds for any video, the problem is in the hook. If it’s 5–10 seconds, the problem is in the body of the video. If it’s above 15 seconds, the content is holding, scale it up.

Completion rate (percentage of inspections to completion) is the most valuable TikTok metric for an affiliate marketer. Shows what percentage of viewers watched the entire video. In TikTok Analytics is shown in the retention graph - look for the value at the right end of the graph. The average completion rate for 30-second videos is 18–22%. Videos with a completion rate above 35% receive powerful algorithmic push. For affiliate marketing, the completion rate is critical, because the CTA usually stands at the end of the video - if the viewer did not reach the end, he did not see the call to action.

Shares (reposts) - the number of times the video was sent to private messages or shared on other platforms. For TikTok this is one of the strongest algorithmic signals. One share “weighs” approximately 5–7 likes according to the algorithm’s calculations. For an affiliate marketer, shares mean free secondary coverage - your video reaches people whom the algorithm might not have reached.

Profile visits (profile visits) - how many people went to your profile after watching the video. A key metric for UBT schemes where the link is in the profile bio. Conversion formula: profile visits / views × 100 = profile visit rate. A good indicator is 3–5%. If it’s below 1%, the viewer lacks motivation to switch; strengthen the CTA in the video.

Link clicks - available for business accounts with a link in bio. TikTok does not show which specific video the click came from - only the total number for the period. Use UTM tags and trackers to link clicks to specific videos.

Affiliate marketing formula for TikTok: creative effectiveness = (profile visits + link clicks) / views × 100. This is your real CTR. If it is below 0.5%, the creative does not convert traffic into conversions, even if it gains views. If above 2% is a winner, scale immediately.

Instagram Insights: what lies behind reach and impressions

Instagram Insights is available on business and creator accounts. The main trap of Instagram is the duplication of metrics with different names, which confuse even experienced affiliate marketers.

Reach (reach) - the number of unique accounts that saw your content. One person looking at Reels three times is 1 reach. For an affiliate marketer, reach shows the potential size of the audience touched by the creative.

Impressions - The total number of times the content was shown, including repeat views by the same user. Impressions are always greater than or equal to reach. The impressions / reach ratio shows the “virality” of the content: if the ratio is above 1.5, viewers are re-watching your video, which is a strong signal for the algorithm.

Saves - the number of times users have bookmarked your content. In 2026, saves is the main algorithmic signal Instagram, overtaking likes and comments. Instagram interprets save as “this content is so valuable that people want to come back to it.” One save “weighs” as much as 3-4 likes in the calculation of the Explore algorithm. For an affiliate marketer, high-save content means long-lasting impact: the user will return to the video and, possibly, click on the link later.

Shares (reposts) - sending content via Direct or to Stories. The second strongest algorithmic signal after saves. Shares open access to a “warm” audience—friends trust their friends’ recommendations more than algorithmic selections.

Profile actions - transitions to the site, clicks on email, calls and referrals. For an affiliate using a link in bio, the only significant metric here is website clicks. Instagram shows it in the Insights → Activity section. Consider website click rate: website clicks / reach × 100. The guideline is 0.3–1.2% for organic traffic in Reels.

Story exits (exits from Stories) - the percentage of viewers who closed your Story before it ended. If you're using Stories to warm up before a CTA (link in last slide), story exit rate is critical. The average is 5-8% exits on each slide. If the exit rate is higher than 15% on a specific slide, the content on this slide repels the audience; remake or delete it.

Reels-specific metrics: Instagram shows plays (analogous to views TikTok), likes, comments, saves and shares separately for each Reels. Completion rate is not directly shown, but it can be estimated through “average percentage watched” in Reels Insights (available from 2025). A value above 50% for 30 seconds Reels is an excellent result.

Affiliate marketing formula for Instagram: creative effectiveness = (saves + shares + website clicks) / reach × 100. Saves and shares are included because they directly affect programmatic promotion, which means future free reach. CTR below 1% - weak creativity. Above 3% - scale through uniqueness to the entire network of accounts.

YouTube Studio: CTR, AVD and hidden signals of the

YouTube Studio provides the most detailed analytics of the three platforms. For Shorts and long videos, the metrics are different - we analyze both categories.

Impressions (impressions) - the number of times your video preview was shown to users on YouTube (in the feed, in recommendations, on the search page). Impressions are not equal to views: the user can see the preview and not click. For Shorts impressions are calculated differently - this is the number of times Shorts appeared in the user's Shorts feed.

CTR (Click-Through Rate) - the percentage of users who clicked on your video after seeing the preview. Formula: views / impressions × 100. Average CTR to YouTube - 4–6%. For long videos, CTR depends on the preview and title. For Shorts CTR is less relevant because the videos automatically play when scrolling. However, “implied CTR” for Shorts is the percentage of viewers who stopped on your video instead of swiping. YouTube shows this as a “swipe-away rate” (the lower the better).

AVD (Average View Duration) - average viewing duration. The main metric YouTube for assessing the quality of content. For long videos, AVD 50%+ of the total length is an excellent result. For Shorts, AVD is measured in seconds: if the average viewing time of a 15-second Shorts is 11 seconds, that's 73% retention, which is strong.

Retention graph is the most informative tool in YouTube Studio. Shows the percentage of viewers for each second of the video. Look for “dips”—moments where viewers leave en masse. For an affiliate marketer, failures before the CTA are a disaster: the viewer did not reach the link. If the CTA is at 25 seconds and the flop is at 20, move the CTA earlier or strengthen the content at 18-22 seconds.

External traffic - YouTube shows where viewers come from: search YouTube, recommendations, external sites, direct links. For an affiliate marketer, external traffic is important in the context of cross-platform promotion: if you duplicate content from TikTok to Shorts, check whether the platforms are cannibalizing each other.

Shorts-specific metrics: in 2026 YouTube Studio allocates Shorts to a separate analytics section. Key indicators: views, likes, comments, shares and “percentage viewed” (analogous to completion rate). YouTube also shows “remix stats” - how many times your Shorts has been used as the basis for someone else's content - this is a powerful viral signal.

Affiliate marketing formula for YouTube: for Shorts - effectiveness = (likes + comments + shares) / views × 100 - your engagement rate. For long videos, add CTR × AVD as an “algorithmic value” metric. If the ER is above 5% for Shorts it is a potential winner.

Metrics that decide ROI: engagement rate, CTR, cost per result

Native platform metrics - raw data. To make decisions, an affiliate marketer needs derived indicators normalized to the task - earnings.

Engagement Rate (ER) - involvement

Universal formula: ER = (likes + comments + saves + shares) / reach × 100. Variation for TikTok (reach is not shown): ER = (likes + comments + shares + saves) / views × 100. Average ER by platform in 2026:

• TikTok: average ER - 4-6%. Above 8% – high engagement. Above 12% – viral content
• Instagram Reels: average ER - 2-4%. Above 6% is high. Above 10% – viral
• YouTube Shorts: average ER - 3-5%. Above 7% is high. Above 11% – viral

For an ER affiliate, a proxy metric of “resonance”: how much the content touched the audience. A high ER means the theme, hook, and delivery are on point. But ER does not equal conversion - a video can get likes and comments, but not generate link clicks. Therefore, ER is used for primary screening: out of 10 videos, you filter out 7 with low ER, and analyze the remaining 3 using CTR.

Click-Through Rate (CTR) - conversion to transition

CTR in the context of affiliate marketing is not CTR preview of YouTube, but CTR of your CTA: how many viewers completed the target action (click on the link in bio, go to profile, swipe on the link in Stories). Formula: CTR = target actions / views × 100.

Organic traffic guidelines:

• TikTok (profile visits): 1–3% is the norm, above 5% is the winner
• Instagram (website clicks): 0.3–1% is the norm, above 1.5% is the winner
• YouTube (end screen clicks + description links): 0.5–2% is the norm, above 3% is the winner

CTR is the only metric that is directly converted into money. A video with 2,000 views and CTR 5% gives 100 clicks. A video with 20,000 views and CTR 0.2% - 40 clicks. The first video is 2.5 times more effective with ten times less coverage. This is counterintuitive - and this is why affiliate marketers who only look at views lose.

Cost Per Result (CPR) - cost of the result

For paid traffic, CPR is automatically considered an advertising account. For organic organic free traffic (UBT), CPR must be calculated manually: CPR = (infrastructure costs / period) / number of conversions per period.

Infrastructure costs include: proxy, anti-detect browser, SIM cards, 360° Uniquizer for uniqueness, numbers for verification. Let's say the total cost is $300/month, and a network of 30 accounts generates 900 conversions. CPR = $300 / 900 = $0.33 per conversion. If an offer pays $5 per conversion, your ROI = ($5 − $0.33) / $0.33 × 100 = 1,415%. It is CPR that is the final arbiter of the effectiveness of the connection, not views and not ER.

How to determine the winner: data analysis algorithm

Metric theory is useless without a practical decision-making algorithm. Here is a step-by-step methodology that top affiliate marketers use to identify winning creatives from an array of analytics data.

Step 1: initial screening (after 24 hours)

24 hours after uploading, check the basic metrics of each creative:

• Completion rate below 15% - creativity is dead, don’t waste time on analysis. The problem is in the hook or the first seconds
• Completion rate 15–25% - average. May work if ER is high. Leave for observation
• Completion rate above 25% is a potential winner. Go to in-depth analysis

At this stage, 50–70% of creatives are eliminated. This is normal - in affiliate marketing, one winner out of five to seven options is considered a good result.

Step 2: in-depth analysis (after 48 hours)

For the remaining creatives, consider derived metrics:

1. ER: calculate using the formula for each platform. Compare between options - not with the “market average”, but with each other
2. CTR: profile visits / views for TikTok, website clicks / reach for Instagram. This is your first level conversion
3. Retention graph: look at where the audience leaves. If the gap is before the CTA, move the CTA earlier. If there is a smooth decline, the content is normal, but not catchy
4. Save rate: saves / views × 100. A high save rate (above 2%) with a low ER means that the content is useful, but not emotional - viewers save, but do not like

Step 3: decision matrix

Make a table for each creative with four columns: completion rate, ER, CTR, save rate. Assign a score from 1 to 5 to each indicator (within the group of tested options). Summarize. The creative with the highest score is your winner.

Weightings for arbitrage (not for blogging!):

• CTR - weight 3x: is money. High CTR = more clicks = more conversions
• Completion rate - weight 2x: affects the algorithm (more coverage) and the CTA (viewer reached the end)
• ER - weight 1.5x: algorithmic signal, but not directly converted into money
• Save rate - weight 1x: long-term asset, but slow in conversion

Example: creative A - CTR 3% (score 5 × 3 = 15), completion 30% (5 × 2 = 10), ER 6% (4 × 1.5 = 6), save rate 1.5% (3 × 1 = 3). Total: 34 points. Creative B - CTR 1.5% (3 × 3 = 9), completion 35% (5 × 2 = 10), ER 8% (5 × 1.5 = 7.5), save rate 3% (5 × 1 = 5). Total: 31.5 points. The winner is creative A, despite the lower ER and save rate, because CTR is the main metric for arbitrage.

Step 4: Winner Scaling

We've determined the winner based on analytics data—now scale it up. This is where analytics intersects with uniqueness: you know which creative works, and you need to upload it to the maximum number of accounts without detecting duplicates.

Upload the winning video to 360° Uniquizer and create a unique version for each account in the grid. The software changes the video at all levels of verification - visual hash, audio fingerprint, metadata, neural network analysis - while maintaining the content and quality that showed the best metrics. Each account receives a technically unique file, but viewers see the same content that won the A/B test.

After scaling, the analytics goes into monitoring mode: monitor the same metrics (completion rate, ER, CTR) for each account. When the indicators begin to decline, this is a signal of creative burnout, and it’s time to launch a new testing cycle.

Tools and A/B testing: putting the system together

Native analytics of three platforms is a basis, but it is not enough for large-scale work with account grids. Manually checking the metrics of 30 accounts on three platforms - 90 dashboards daily - is physically impossible. Need automation tools.

Native instruments

TikTok Analytics (Pro Account): free, shows data for 7, 28 and 60 days. Sufficient to work with 1–5 accounts. Limitation: no data export and no API for automation.

Instagram Insights (Business/Creator): free, 7, 14, 30 and 90 day data. Shows separate analytics for Reels, Stories, posts. Restriction: you cannot compare accounts with each other.

YouTube Studio: the most detailed native tool. Real-time data, export to CSV, comparison of periods, retention graphs accurate to the second. Limitation: Only works for one channel.

Third-party dashboards for large-scale work

Analisa.io - analytics TikTok and Instagram without access to the account (by username). Convenient for monitoring competitive accounts and your network without logging in to each one. Shows ER, posting frequency, audience demographics.

Not Just Analytics (ninjalitics) - deep analytics Instagram: subscriber growth, ER by post, best publication time. The free version covers basic needs.

vidIQ / TubeBuddy - extensions for YouTube Studio with additional metrics: keyword score, competition score, SEO recommendations. For Shorts, the “Shorts performance comparison” function is useful - it compares your Shorts to CTR and retention in one table.

Custom tables - the optimal tool for advanced affiliate marketers: Google Sheets or Notion table with formulas for calculating ER, CTR, CPR for each account and creative. Manual data entry once a day takes 15–20 minutes for 20 accounts, but gives a complete picture and history of changes.

A/B testing methodology through analytics

A/B testing of creatives in affiliate marketing differs from classical marketing: you don’t have an advertising account with automatic split, but you have a network of accounts - and this is your advantage.

Test structure:

1. Define a variable: test one thing at a time - hook, CTA, duration, music, text on screen. If you change everything at once, you won’t understand what exactly influenced the result
2. Prepare options: 3-5 versions with one variable. For example, the same video with five different hooks. Or one hook with five different CTAs at the end
3. Unique each option via 360° Uniquizer for each test account. Critical: upload different options to different accounts, but not different options to one account - the algorithm can “punish” for frequent content changes
4. Upload simultaneously (within 1-2 hours) to all test accounts. Post time affects initial reach, so timing is important for test integrity
5. Collect data after 48 hours: completion rate, ER, CTR, save rate from each account
6. Average the metrics for each option (for 3-5 accounts) - this removes the error of individual accounts
7. Apply the decision matrix from the previous section

Common A/B testing mistakes:

• Insufficient sample: testing on one account is pointless - the algorithm can “kill” a video for reasons independent of the content (shadowban, publication time, competition in the feed). Minimum 3 accounts per option
• Impatience: making decisions after 6 hours instead of 48. Algorithm needs time to distribute coverage - early data is skewed
• Multiple variables: changing the hook, music and CTA at the same time invalidates the test
• Ignoring platform specifics: testing the same option on TikTok and Reels and comparing the results is incorrect because the audiences, algorithms and calculation methodologies are different

A systems approach to analytics transforms arbitrage from a lottery into an engineering process. You don’t “guess” creative that works—you find it through data. And when you find it, you scale it through uniqueness to the entire grid, preserving the very metrics that made the creative a winner.