TikTok Reach vs Engagement: How to Tell Which Is Stronger

TikTok is a platform surface of the same InfScore, not a separate formula. On model 2026-08-18.1 the published mix is still audience quality 32%, engagement 27%, reach & momentum 23%, and platform strength 18%, scored from public TikTok-facing creator signals. Private TikTok analytics and For You myths do not move the free score. "Stronger" means the higher observed factor score on that 0–100 row, not the bigger weight and not a private Insights screenshot. Engagement's 27% is importance in the mix; reach & momentum's 23% is the same kind of weight. This guide shows how the two rows differ, how weight vs score interact, and how dashes stay honest when a provider field is missing. For the four-factor tour, start with How to Measure Influencer Performance.

Same InfScore, TikTok-facing inputs

Search a public TikTok profile on infscore.com and you get the same free 0–100 composite as an Instagram or YouTube lookup. No account is required. There is no separate TikTok-only formula. The scorecard still shows four independently calculated rows, scoring coverage, and the model id. The platform you typed is the surface. It is not a fork of the math.

Brands still ask which TikTok number "matters more," as if the app kept its own weights. It does not. The displayed mix is the mix the current model uses: audience quality 32%, engagement 27%, reach & momentum 23%, platform strength 18%. Independently calculated still applies. A loud TikTok engagement row cannot lend points to a missing reach row, and a huge follower total cannot invent an engagement rate.

What does change is which public creator signals sit behind those rows when TikTok is the profile you searched:

Factor	Weight	TikTok-facing public input
Audience quality	32%	Provider-measured audience credibility when that signal exists
Engagement	27%	Public TikTok engagement rate, mapped onto a fixed 0–8% scale
Reach & momentum	23%	Logarithmic public follower total, blended with growth when available
Platform strength	18%	How many public audiences were observed, and whether they are badly lopsided

Those inputs are public provider data: not a dump of TikTok Analytics, not a For You placement report, and not a creator-uploaded CSV. If a field is missing, the factor is a dash. The missing weight is held out. Remaining observed weights are rebalanced in their published proportions. See Missing Data on Creator Metrics for the coverage arithmetic.

Reach still sums observed public audiences. If the provider also returned Instagram or YouTube for the same creator, those follower totals join the reach input. Platform strength counts how many of those graphs were observed. TikTok-only lookups still run the same two rows this article compares; they just sit on a thinner platform graph. That is a coverage fact, not a verdict that posting on one network is a mistake.

The worked public card for this guide is Khaby Lame on TikTok. Example cards on the homepage are illustrative; a score URL is the product. For the same composite on a different surface, see Emma Chamberlain on Instagram or the MrBeast YouTube score, not a different TikTok formula. The Instagram-facing map is Instagram InfScore Factors.

What "stronger" means on a scorecard

Two different numbers sit on each of those rows, and people mix them up.

Weight is the published share of the model: engagement 27%, reach & momentum 23%. When both rows are present, engagement counts slightly more in the mix. That is importance in the composite. It is not a trophy for the engagement line.

Factor score is the 0–100 grade on that row for this creator, on this fetch, on this model version. "Stronger" in this article means the higher of those two observed scores. A reach row at 80 is stronger than an engagement row at 50, even though 23% is smaller than 27%.

A third reading is weighted contribution: factor score × weight. On the same example, reach contributes 80 × 23 = 1840 and engagement contributes 50 × 27 = 1350, so reach also moved the composite more. Ask the contribution question when you care how the 0–100 InfScore was assembled. Ask the factor-score question when you care which public row actually looks stronger. Do not treat the 27% vs 23% split as "always prefer engagement."

None of this replaces audience quality at 32% or platform strength at 18%. Those rows still run when their inputs exist. Comparing reach and engagement while cropping the rest of the card is the same trap as screenshotting one green line. How to Read an Influencer Scorecard is the full reading order: coverage, dashes, independent rows, then the composite.

If a compared row is dashed, it has no factor score to rank. You cannot call it stronger or weaker. You can only say it did not run.

Engagement (27%) on TikTok

Engagement strength converts a public TikTok engagement rate onto a fixed 0–8% scale. That factor is 27% of InfScore when the rate exists. Eight percent is the top of the scale: it maps to 100 on the factor. Values above 8% are capped so a single outlier posting burst cannot take over the composite.

On model 2026-08-18.1, the public percentage is divided by 8, multiplied by 100, then clamped to 0–100 and rounded. A 4% public rate maps to 50 on the factor. A 2% rate maps to 25. The math is linear until the cap, then it stops. The conversion is identical for every public Instagram, TikTok, and YouTube profile on this version. TikTok does not get a friendlier scale, and it does not get a harsher one.

What InfScore skips: inventing a "good TikTok engagement rate for your niche," comparing you to an unpublished peer set, or treating comment volume from accounts that never watch the work as a quality bonus. The public engagement percentage is the input. The 0–8% scale is the conversion. Context is the rest of the card. The dedicated guide is What a Good Engagement Rate Looks Like in 2026.

If the public rate is missing, the factor is a dash. InfScore will not substitute likes-per-follower from a different platform, a private analytics average, or a screenshot of "video views last week" to hide the gap. Coverage then falls by 27, to 73 (32+23+18), and the remaining mix stays 32 : 23 : 18.

Comment pods and bought activity are not a scoring win. They can inflate a raw ratio for a moment and still leave a weaker audience-quality row. InfScore is built from observed public signals, not from manufactured comments.

Reach & momentum (23%) on TikTok

Reach is logarithmic, not a richest-follower contest. On the current model, logarithmic reach runs from about 1,000 public followers (0 on that piece) through 100 million (100). TikTok follower vanity does not move this row one-for-one. The log scale compresses the gap between a large account and a gigantic one, and it keeps mid-size creators on the same chart.

The input is the public follower total InfScore can actually see. When TikTok is the only observed audience, that total is the TikTok count. When the provider also returned other public graphs for the same creator, those counts are summed. A platform the provider did not return does not add to the total.

When a growth percentage exists, the factor becomes a blend: 70% logarithmic reach and 30% momentum. Momentum is a bounded transform of that growth signal. If growth is missing, reach still scores on its own. Growth is the optional 30% blend, not a veto on the row. The factor dashes only when the public follower total is missing or not usable. Coverage then falls by 23, to 77 (32+27+18).

InfScore will not estimate "true TikTok reach" from impressions it cannot see. Views, watch time, and fill-rate live in denser provider modules when they exist. They are not smuggled into the free composite. A screenshot of accounts reached last week does not rewrite this row, and a claim about For You distribution does not either.

Read that twice if you have been taught that more TikTok followers always win. InfScore still moves as public reach grows. It will not treat 80 million as eight times 10 million on this factor, and it will not treat a bought spike as a credibility upgrade. Size and quality stay on separate rows. Follower count is the size label. Audience quality, when present, is the credibility grade at 32%. The longer split is Audience Quality vs Follower Count.

Reading both rows together

Put the two 0–100 scores next to each other. Ignore which weight is larger until you have named both scores, both dashes, and scoring coverage.

High engagement, weaker reach. The public rate mapped onto the 0–8% scale landed higher than the log-reach row. The comments look busy relative to the audience the provider can see. The audience is smaller, or the log curve already compressed a large total, or growth is missing so the 30% momentum blend never ran. Engagement did not "fix" reach. Reach still has its own 23% when the follower total exists.

High reach, weaker engagement. The log-reach row is the stronger of the two. A large public TikTok following can do that even when the rate is ordinary. This is the usual vanity trap in reverse: the follower screenshot looks like a win, and the engagement factor is the quieter number. Weight does not rescue it. 27% of a lower score is still a lower score on that row.

One row dashed. If reach is missing and engagement is present, coverage is 77. If engagement is missing and reach is present, coverage is 73. The present row stays itself. It does not stand in for the gap. Independently calculated means you do not average the two into a fake pair.

Both dashed. Coverage then sits on whatever else ran. If only audience quality and platform strength returned numbers, coverage is 50 (32+18). You have no reach-vs-engagement comparison at all. Guessing either row would be inventing the card.

A checkable case from the model tests, using round numbers: audience quality missing, engagement at 4% public rate (50 on the factor), reach & momentum 40, platform strength 45. Coverage is 68. The remaining mix is 27/68, 23/68, and 18/68. The composite is (50 × 27 + 40 × 23 + 45 × 18) ÷ 68, which rounds to 45. On that card, engagement (50) is the stronger of the two compared rows. Reach is 40. Neither row is the InfScore. Paying does not fill the audience-quality dash, and it does not flip 50 vs 40.

What does not settle the comparison

A lot of TikTok metrics feel like they should pick a winner. Most of them never enter the two rows.

Private analytics and creator-app Insights. The free InfScore uses public creator signals. Insights screenshots, unpaid access to non-public fields, and a logged-in view of For You performance are not free-score inputs. Paying $0.99 does not purchase a pipe into the creator's app, and it does not fill a dash with a screenshot you pasted.

Follower vanity as a linear grade. Headline TikTok count is size. Reach scores that size on a log scale as 23% of the model. Audience quality scores credibility as 32%. A bigger number on the profile does not, by itself, mean reach won the comparison.

Bought follows, pods, and traffic that never comes back. Those usually work against you. Hollow followings tend to weaken credibility. Pods can inflate a raw ratio for a moment and still leave a weaker AQ row. InfScore will not publish a recovery score for people who did those things, and it will not treat a purchased spike as momentum you can keep.

One green row. A strong TikTok engagement factor is 27% of the model when the rate exists. A strong reach factor is 23%. Cropping the card to the prettier line is not a reading.

A homepage tile. Example cards on the homepage are illustrative. A live /score/ URL is the product.

Paying. The one-time $0.99 report unlocks denser provider modules when they exist. It leaves the 32/27/23/18 weights alone, leaves the 0–8% engagement scale alone, and does not convert a dash into a 0, a 50, or a 100.

InfScore is an assessment of public signals, not a growth-hack tool. There is no published "raise InfScore by posting Y" table, and this article will not invent one.

How to check a live TikTok InfScore

Read the TikTok card the same way you read any other public InfScore: coverage first, then dashes, then the four rows, then the 0–100 number and the model id. When the job is reach vs engagement, name those two factor scores only after the dashes.

1. Scoring coverage. If every factor is present, coverage is 100 and the 32/27/23/18 mix applies in full. If AQ is dashed, coverage is 68. Engagement dashed, 73. Reach & momentum dashed, 77. Platform strength dashed, 82.
2. Dashes. A dash is a gap in public provider data. Treating it as zero punishes a blind spot. Treating it as 100 invents a perfect row. Missing Data on Creator Metrics is the rebalance rule.
3. Independence. Each factor uses its own TikTok-facing inputs. Engagement cannot stand in for logarithmic reach. Logarithmic reach cannot stand in for a missing growth series or for AQ.
4. The two factor scores, only if both exist. The higher 0–100 row is the stronger of the two. The 27% vs 23% split is still just weight.
5. The 0–100 composite, after the rows. Cite model 2026-08-18.1. A 72 on this version is this mix. A 72 on a later version might mean something else.
6. A live URL. Open Khaby Lame on TikTok if you need a worked public card. Do not paste a homepage tile.

Platform strength still sits at 18% when at least one public audience was observed. On the current model the coverage floors are 45 for one observed platform, 70 for two, and 90 for three or more, plus a small balance bonus when follower counts are not concentrated on a single network. A TikTok-only lookup can still return a number near that 45 floor. It does not pick TikTok as a "best" app, and it is not a request to spam two empty handles so the floor can move.

The public InfScore stays free. No signup. Pricing is one report, one dollar, no recurring fee. Use the paid report when you need to explain the score to someone else. Use the free score when you need the comparable number quickly. Both views are honest about gaps. The paid view is simply denser.

This versioned composite uses only available public creator signals. Missing factors stay unavailable and their weight is rebalanced across observed factors. It is not a percentile, price estimate, or guarantee of campaign results.

The TikTok InfScore worth trusting is the one you can check: four published factors, public inputs, visible dashes, and two rows you can actually compare. Search a public profile on infscore.com. If you want the modules behind a specific scorecard, the full report is there for $0.99, and only when you actually need it.