What Is Click-Through Rate (CTR) in Short-Form Video?

What it actually measures

In its strictest definition, CTR equals the count of unique clicks on a destination surface divided by the count of unique impressions where the destination was visible, expressed as a percentage. The denominator definition is platform-specific and is the source of most reporting confusion. Instagram exposes link-sticker CTR and product-tag CTR on the Professional Dashboard, with impressions as the denominator. TikTok exposes link CTR on Business Center accounts under the Creative Center analytics tab. YouTube Shorts exposes end-screen card CTR on Studio Analytics with impressions as the denominator. The four destination types on short-form video are not interchangeable as CTR surfaces, and reporting them under one number is the most common reporting error.

The four destination types matter because they isolate different funnel moments. Profile-tap CTR diagnoses top-of-funnel curiosity. Bio-link-tap CTR, computed from profile visits divided by reach upstream and then clicks divided by profile visits downstream, diagnoses account-page conversion. Link-sticker CTR diagnoses mid-funnel intent. Destination-card CTR on Shorts, which the viewer sees only after finishing the clip, diagnoses bottom-of-funnel intent. The four numbers move differently across content types.

Where the term gets misused is when teams treat CTR as a single ranker signal. The rankers on Reels and TikTok do read profile-tap and link-tap CTRs as positive interactions, but no platform has published tap-signal weights; the defensible claim is that watch-time, share, and save inputs are the ones the platform documentation actually names. Mosseri's January 21, 2025 framework named the three Reels signals in priority order as watch time, likes per reach, and sends per reach. Profile taps and link taps appear in the Reels ranker's broader input stack but sit below the top three.

How to calculate it

The formula is one line per destination type. CTR equals clicks divided by impressions, multiplied by 100, computed per destination surface.

Walk it through with a fictional brand for grounding. Vespera Skin is a 22K-follower Instagram skincare DTC brand running roughly $4M ARR. On a Tuesday in April 2026 Vespera ships a Reel demonstrating the niacinamide-versus-vitamin-C layering protocol. The Reel attaches a link sticker pointing to the brand's product-bundle page. Over the first 72 hours the Reel reaches 38,400 unique accounts, generates 41,800 impressions, earns 1,241 profile taps, 360 of the profile visits that follow convert to a bio-link tap, and 287 of the link-sticker impressions earn a direct sticker tap.

The CTR breakdown by destination is the operator-relevant view. Profile-tap CTR equals 1,241 divided by 41,800 impressions, multiplied by 100, which is 2.97 percent. Link-sticker CTR equals 287 divided by 41,800, multiplied by 100, which is 0.69 percent. The link-in-bio conversion computed downstream of the profile visit equals 360 bio-link taps divided by the 1,241 profile visits — one denominator, the profile-visit count — which is 29.0 percent. Three numbers, one denominator each, three diagnoses. The profile-tap CTR tells Vespera whether the clip earned curiosity. The link-sticker CTR tells Vespera whether the offer survived the in-flow tap moment. The link-in-bio CTR tells Vespera whether the account-page chrome converted the visit. Reporting one combined CTR hides which surface to fix.

What good looks like by platform

Set the operating range for CTR from the account's own destination-type median rather than from a published band; no industry report publishes verified profile-tap or link-sticker CTR bands by audience size. The structural relationship holds regardless of the account: the profile tap is a lower-friction action than the link tap, so sticker CTR runs below profile-tap CTR on the same clip, and TikTok's profile-tap CTR tends to run above the Reels number on the same content because the surface's click grammar differs.

The destination-surface consequence is a frictions argument, not a documented algorithm fact: clips that direct the click to the profile (lower friction) tend to earn higher CTRs than clips that direct the click to a link sticker (higher friction), and operators commonly read a roughly 2x to 3x gap between the two surfaces on the same clip as a friction effect. Buffer's TikTok algorithm guide does not rank profile-tap CTR or link-tap CTR anywhere in its engagement list.

CTR tends to peak on mid-length clips paired with a short caption, where the clip body has enough room to plant the curiosity and the caption is short enough to read in-flow. Clips with very long captions tend to generate lower CTRs because the in-flow read overflows the visible area and the click moment is interrupted.

Uncited working heuristics some operators use (not published benchmarks; validate against your own panel): 0.50 percent profile-tap CTR on Reels, 0.20 percent link-sticker CTR on Reels, 0.80 percent profile-tap CTR on TikTok, 0.30 percent end-screen-card CTR on Shorts.

What I look for when I audit this metric

I run a four-pass diagnosis on CTR before recommending any content or surface change.

The first pass is the destination-type breakdown. I pull profile-tap CTR, link-sticker CTR, link-in-bio CTR, and (on Shorts) end-screen-card CTR side by side on the last 30 clips. If the profile-tap CTR passes the floor but the link-sticker CTR does not, the curiosity is firing but the offer is the bottleneck. If both pass but the link-in-bio CTR fails, the account-page chrome is the bottleneck and the fix is upstream of the clip.

The second pass is the in-flow caption check. I read the caption against the clip in real time. If the caption requires a "see more" expansion to deliver the click prompt, the in-flow conversion is broken and the CTR will run roughly 30 to 50 percent below the same clip with the prompt above the fold. The fix is to move the actionable line to the first 80 characters of the caption.

The third pass is the friction audit on the destination page. I tap through from the same surface the viewer would land on (mobile in-app browser, not desktop Chrome). If the page requires a second tap to reach the offer, or if the page is not optimized for the in-app browser's viewport, the CTR-to-conversion fall-off accumulates and the operator misreads a destination-page problem as a clip problem.

The fourth pass is the cross-platform CTR comparison. I pull the same content distributed to two surfaces and compare the profile-tap CTR side by side. If the cross-platform delta is above 2.5x, the lower-CTR surface is mispositioned for the content category or the destination surface (the in-bio link versus the link sticker) is not optimized for that platform's grammar.

Common mistakes

The first mistake I see is reporting CTR as a single combined number across destination surfaces. The four destination types diagnose different funnel moments and combining them hides which surface to fix. Reporting profile-tap, link-sticker, link-in-bio, and end-screen-card CTRs separately is the only way to read the signal.

The second mistake I see is over-relying on caption CTAs to lift CTR. The agency-side default of "link in bio" CTAs is a consistent pattern that produces flat CTR numbers because the call is decoupled from the clip's payload. The lift comes from the clip's payload prompting the click, not from the caption telling the viewer to click.

The third mistake I see is treating CTR as a top-line ranker signal on Reels. Mosseri's January 21, 2025 framework named the three Reels signals in priority order as watch time, likes per reach, and sends per reach. Profile-tap CTR is read by the ranker as a positive interaction but sits below the top three. The right sequence is to draft for the retention curve first, then layer in the click-prompting elements. Treat in-feed clicks as positive but secondary to the watch-time and send signals the platform rankers actually name.

Where a planning-first tool fits

The brand-profile analysis I built in a planning-first tool tracks CTR by destination type across an account's last 30 clips and adjacent competitor accounts as one input among several (not a substitute for the destination-page friction audit above). The operator's call on which destination surface to attach to a given clip sits upstream of any dashboard.

Related Terms

Frequently asked questions