Avery Vale

How every figure below was reached, and what its confidence marks mean.

This dossier assesses Avery Vale (@averyplays), streaming primarily on Twitch with a secondary presence on YouTube, against the brief. The same checks were run as for every creator: attested reach and engagement metrics were read from platform-native sources and cross-checked; engagement was compared against the tier and vertical benchmark; audience composition was sought where the platform exposes it; expected delivery was modelled as a range with a central estimate; the probability of clearing the brief's delivery threshold was modelled with an error band; and the content history was scanned for the safety concerns the brief names. Findings are graded on a three-tier confidence framework: high-confidence values are stated plainly with their source and read date; medium-confidence values are given directionally with the basis named; low-confidence values are marked as estimates with the governing assumption stated. The observation window is the trailing 90 days, running to 2026-07-02, the date the content-history scan was run. Where a value could not be obtained to a usable confidence, that is recorded as such rather than inferred.

The reach and engagement figures read from the platforms, each with its source and read date.

The account's reach and engagement figures were read from platform-native surfaces during the trailing-90-day window. Followers and subscribers, true reach, median views, engagement rate, and posting cadence each carry a source stamp, a read date, and a confidence tier in the underlying record, and each is marked as observed or derived by name. For this sample record the individual metric rows are not reproduced; the derived figures they support — the modelled delivery range and its central estimate — are set out in the expected-delivery section, where each input's confidence is shown. True reach is a derived value rather than an observed one, and is treated as such wherever it feeds a downstream estimate.

How the audience actually shows up, measured against the brief.

Engagement was assessed against the benchmark for the creator's tier and vertical. The record carries the comparative view — the observed rate, the band median for the peer group, and the delta between them — together with coherence checks on the shape of that engagement, which look for the patterns that distinguish organic interaction from inflated or purchased activity. On the evidence read within the window, the engagement is consistent with the audience-fit reading that follows and shows no coherence signal that would qualify it. The specific rate and peer-band median sit in the metrics record and are not restated here.

Who the audience is, and how much overlaps other channels.

Audience composition — demographics, geography, and language — could not be obtained to a usable confidence for this account within the observation window; the platform did not expose the segment data the brief would require to score it, and no reliable derived substitute was available. This is recorded as an obtainability gap rather than a finding: the absence reflects what the platform surfaces, not a judgement about the audience itself. The new-audience overlap the brief calls for is therefore not scored here. The directional read on audience fit that informs the verdict rests on the engagement evidence above rather than on segment-level composition.

The delivery range this creator is modelled to land in, and how it was derived.

Modelled delivery for this engagement falls in the range 6,019 to 12,198, with a central estimate of 8,569. The range is derived from the account's observed reach and engagement over the trailing 90 days, discounted to the conversion pattern the brief's format implies; the low and high bounds mark the modelled spread around that central figure rather than a floor and ceiling that bound every outcome. The central estimate clears the brief's delivery floor with room in hand. Each input to the derivation carries its own confidence in the record: the reach and engagement observations are high-confidence platform reads, while the conversion assumption that turns reach into delivery is the medium-confidence input, stated directionally and based on the format's typical performance. The single active caution, noted in the risk register, is descriptive and left the delivery range unchanged.

Expected delivery 6,019 to 12,198, central estimate 8,569.

The modelled chance of clearing the brief's delivery threshold.

The modelled probability that delivery clears the brief's threshold is 71.65 per cent, carried with the error band the delivery distribution implies. This is a modelled estimate derived from the delivery range above and the position of the brief's floor within it; it is a modelled estimate, not a commitment as to the outcome. The active caution recorded in the risk register is descriptive and does not move this figure. The estimate rests on the observation window holding — a material change in the account's reach, cadence, or the sponsored-segment conduct noted below would move it.

72% (modelled estimate)

Anything in the content history the brief asked us to check.

One caution stands on this record; it does not clear the register to empty. Detection: a content-history scan of the trailing 90 days measured the account's sponsored segments against the brief's stated sensitivity to family-friendly tone during sponsored content. Observation: strong language appeared during sponsored segments in two clips within the window, dated 2026-05-28 and 2026-05-05, both on the primary Twitch channel (source clips are held in the record and redacted for this sample). This is graded a caution in the brand-safety family. Quantified effect: this finding is descriptive — it leaves the modelled delivery range unchanged and does not adjust the modelled hit probability. The remaining scope the brief named was checked and found clear: strong language beyond the two clips noted, gambling references, competitor endorsements, and platform standing — strikes or other enforcement history — surfaced nothing across the same window. The clear result on those scopes is the outcome of an active check, not an absence of checking.

• cautionbrand_safety

  Two clips in the trailing 90 days contain strong language during sponsored segments.

  Observed: 28 May 2026; 5 May 2026

  Source: https://twitch.tv/averyplays (clip, redacted for sample)

  Method: Content-history scan of the trailing 90 days against the brief's stated sensitivities (family-friendly tone during sponsored segments).
  Band: n/a — descriptive Hit probability: n/a — descriptive

The call, and the reasoning behind it.

Proceed with noted cautions. Avery Vale clears the brief's modelled delivery floor with room in hand — a central estimate of 8,569 against the threshold — and the engagement evidence supports a directionally strong read on audience fit, though segment-level composition could not be scored for want of obtainable data. One caution governs the recommendation: strong language appeared during sponsored segments in two clips within the trailing 90 days, dated 2026-05-28 and 2026-05-05, which cuts against the brief's family-friendly requirement for the sponsored window. The recommendation is to proceed with a briefed content note setting tone expectations for the sponsored segments; the modelled hit probability of 71.65 per cent, carried with its error band, is stated as a modelled estimate and not as a commitment.

proceed with cautions

Avery clears the brief's floor on modelled delivery with room to spare, and the audience fit is directionally strong. One caution stands: strong language has appeared during sponsored segments, which cuts against the brief's family-friendly requirement. Proceed with a briefed content note on tone during the sponsored window.