Skip to content

Find the creators worth betting your brand on.

SENSO gives marketing teams an intelligence layer for discovering, evaluating and measuring creators — with evidence behind every important number.

4.1K
Creators indexed
202.2K
Content read
74
Markets
Creator intelligenceNorthlight Studio

Northlight Studio

@northlightstudioYouTubeCanada

SENSO VerifiedLow risk

SENSO Health

Excellent

Audience authenticityweight 20%
93
Engagement qualityweight 15%
96
Growth patternweight 15%
84
Data confidence94%

High confidence — a separate axis from the score above

Product surfaces on this page are the real interface, rendered with illustrative creators. Coverage figures are live.

Accounts resolved
4.1K
Content items read
202.2K
Historical snapshots
14K
Markets covered
74

Live from the SENSO database as this page rendered. Collected from official platform APIs and creator-authorised accounts — no prohibited scraping, and no purchased list.

Don't choose creators by follower count.

Creator search6,847 creators match · 3 filters

Audience size

  • Micro · 10K–100K1,284
  • Mid · 100K–500K3,918
  • Macro · 500K–1M1,102
  • Mega · 1M+846

Audience quality

  • Low risk4,617
  • Medium risk912
  • High risk268

Verification

  • SENSO Verified2,046
  • Connection pending184
CreatorAudienceEngagementHealth ↓ConfidenceSignals
Northlight Studio@northlightstudioCanadaTechnology
1.2M268.9K views5.4%9194%Low riskActive
Saffron & Salt@saffronandsaltIndiaFood & Beverage
486K108.9K views7.1%8892%Low riskActive
Meridian Field@meridianfieldUnited KingdomTravel
742K191.4K views4.8%8488%Low riskActive
Atlas Grain@atlasgrainPortugalTravel
892K198.5K views1.1%7274%Medium riskSlowing
Kestrel & Co@kestrelandcoAustraliaFashion
2.4M412.7K views0.9%5861%High riskRecent
Halcyon Test Lab@halcyontestlabGermanyTechnology
318K84.2K views6.2%8143%Risk not assessedActive
Verdant Table@verdanttableUnited StatesFood & Beverage
1.1M204.6K views2.7%6979%Low riskSlowing
Filter on audience quality, not size
Engagement, authenticity risk, publishing cadence and verification state are filters, not footnotes.
See what a filter costs before you apply it
Every facet carries a live count against the index, so you know what narrowing removes.
Sort on the score, not the follower count
SENSO Health is a deterministic composite of nine weighted components, not a popularity rank.

Understand the creator behind the audience.

Nine components, published weights
Authenticity, engagement quality and rate, growth, view consistency, audience activity, comment quality, upload consistency, brand safety.
Confidence is its own axis
A creator can score 91 on health with 40% confidence. Folding the two together is the most misleading thing a tool in this category can do.
Absent is never zero
A component nobody could measure is withheld and the remaining weights renormalise. It is never scored as a nought.
Reproducible years later
Every score stores its components, its inputs and its formula version, so a number recorded last quarter can be recomputed today.
Creator dossierhealth-1.1.0

Northlight Studio

@northlightstudioYouTubeCanadaTechnology

SENSO VerifiedLow riskActive

SENSO Health

Excellent

94th percentile of 412 technology creators in the 1M+ follower band.

Audience authenticityweight 20%
93
Engagement qualityweight 15%
96
Engagement rateweight 15%
88
Growth patternweight 15%
84
View consistencyweight 10%
91
Audience activityweight 10%
95
Comment qualityweight 5%
87
Upload consistencyweight 5%
79
Brand safetyweight 5%
94
Data confidence94%

high confidence — separate from the score

Formula coverage100%

all nine components measurable

Followers
1.2M
Median views
268.9K
Engagement
5.4%
Brand safety
94

Content performance

  • Thermals after 200 hours: the honest numbers412.8K1.9×
  • What the spec sheet does not tell you298.1K1.3×
  • Six months with the workstation nobody reviewed241.6K1.1×

Each post measured against this creator's own median, not against other creators.

Every important number carries its evidence.

Most tools hand you a confident figure and no way to check it. Ours shows its working — on any number you might have to defend in a meeting.

Engagement rate

5.4%

Every figure on a profile carries this. Click it and the platform shows its working.

Engagement rate

5.4%Derived

Method
Calculated by SENSO from observed values using a published formula.
Source
Official platform API · tier 1 of 5
View source
Collected
2 hours ago
Derivation
Observed interactions ÷ observed views. Posts that hide likes or disable comments contribute no interactions and are not counted as zero.
Field confidence94%
  1. Verified

    Confirmed through the creator's own authorised platform connection.

  2. Observed

    Measured directly through an official platform API.

  3. Derived

    Computed from observed values by a published, versioned formula.

  4. Estimated

    A model estimate. Labelled as a range, never as a measurement.

  5. AI inferred

    Classified by a model from source material, with its evidence stored.

Sources are ranked, and the ranking is visible. An official API measurement outranks permitted public research, which outranks model inference — and where two sources disagree, the conflict is raised for a human rather than silently resolved in the platform's favour.

AI explains the score. It never sets it.

Profile intelligenceAI classified

Creator type

Independent hardware reviewer

Audience intent

Viewers arrive to decide whether to buy a specific device, and stay for teardown detail they cannot get from a spec sheet.

Strengths

  • Sustained technical depth that survives a long-form format
  • Discloses gifted hardware on camera, consistently
  • Comment threads carry substantive follow-up questions, not emoji

Risks

  • Reviews occasionally name competitor products directly
  • Publishing cadence drops around major launch cycles

Content themes

  • laptop and workstation reviews
  • thermal and acoustic testing
  • long-term durability follow-ups
openai gpt-5.1 · prompt 1.0.0 · stored with its evidence

Models classify, extract and explain. They are asked for one object matching a schema, in strict mode, and the reply is re-validated before anything is stored — so the prohibition is enforced by there being no field to fill, not by wording in a prompt.

A model may never produce

  • Follower counts
  • Engagement rates
  • View counts
  • Audience demographics
  • Bot percentages
  • Any historical metric

Every AI output stores its provider, model, prompt version, schema version and the evidence behind each claim.

From creator selection to measurable delivery.

Select talent from a shortlist, record the agreed rate, set a tracking hashtag, and SENSO attributes what each creator actually delivered.

Campaign deliverycampaign-1.0.0

Campaign score

14

Attributed posts

Matched to #OrbitSeries2026 across 4 confirmed creators.

Cost per engagement
£0.21
Reach
4.1M
Spend against budget£38.2K of £45K
Creator performanceRanked by campaign score
CreatorAgreedPostsReachEngagementCampaign score
Northlight Studio@northlightstudio
£12.0K41.4M5.4%95
Saffron & Salt@saffronandsalt
£9.5K51.2M7.1%88
Meridian Field@meridianfield
£10.2K3986K4.8%74
Atlas Grain@atlasgrain
£6.5K2512K1.1%55

A creator's campaign performance is scored separately from their SENSO Health — the first answers “how did they do for us?”, the second “who are they?”. The two are never merged, and both carry their own formula version.

Quality on one axis. Evidence on the other.

  • Excellent
  • Strong
  • Fair
  • Needs review
Data confidence against SENSO HealthEvery scored profile in the cohort
The top-right quadrant is where a shortlist should be drawn from: creators scoring above 70 on evidence that is itself above 70% confident. The cluster low and to the right is the case this axis exists to expose — a strong score standing on history too thin to rely on, which every other tool in this category would have shown you as a clean number.

Where the numbers come from.

  1. 01

    Platform APIs

    Authoritative structured metrics

  2. 02

    OAuth connection

    First-party creator analytics

  3. 03

    Normalise & snapshot

    Appended, never overwritten

  4. 04

    Deterministic scoring

    Versioned, components stored

  5. 05

    AI enrichment

    Explained and evidenced

A cohort of fewer than eight creators publishes no percentile. A rank computed against two accounts is noise wearing the costume of a statistic, and it would be the most quotable number on the page.

Start with five searches.

The free plan includes five influencer searches per month. Saved profiles, shortlists and comparisons stay available whether or not you have searches left.