Skip to main content
Every frame carries explicit confidence flags. Before you act on a direction or verdict, you should know how much evidence sits behind it. CoDomain makes that information available in every frame so you can calibrate your decisions — not just read a label.

Three Confidence States for Field Sentiment

CoDomain distinguishes between three states for the discourse corpus backing each frame’s field sentiment read.

Validated sentiment

The discourse corpus cleared the minimum publication threshold: n≥20 sourced statements within the 90-day rolling window. The sentiment distribution is statistically meaningful within CoDomain’s methodology.

Directional only

A corpus is present but below threshold. Figures are indicative, not validated. The direction signal may be correct, but it carries more uncertainty than a validated read.

Insufficient signal

The corpus is absent or too sparse for any responsible inference. The frame is held or marked accordingly rather than published with a manufactured direction.
When you see Insufficient signal, the map is being honest rather than manufacturing a direction. A held frame is more trustworthy than a fabricated one.

Spend Confidence: Ranges, Not Point Estimates

CoDomain never expresses spend as a single number. Every spend figure is a range with a confidence band. The width of the band reflects the quality and tier of the underlying sources.
Source tierBand widthInterpretation
Tier 1 — Primary financial disclosures (SEC filings, disclosed ARR, earnings calls)NarrowerHigher-confidence estimate derived from direct financial evidence
Tier 2 — Analyst reports (Grand View Research, Mordor Intelligence, MarketsandMarkets, etc.)WiderEstimate derived from third-party research methodologies
Tier 3 — Procurement signal (CIO surveys, spending surveys)WidestIndicative procurement signal; useful for directional reads
When a spend estimate is triangulated across multiple tiers, the confidence band reflects the mix. A frame backed by two Tier 1 sources will carry a tighter band than one backed by a single Tier 2 report.
Low confidence doesn’t mean wrong — it means the estimate carries more uncertainty and should be read with proportionally more caution.

The 90-Day Rolling Window

The 90-day rolling window is the corpus measurement period for field sentiment. CoDomain counts sourced practitioner statements that fall within the 90 days prior to the frame’s publication date. Statements outside that window do not count toward the publication threshold. Why a rolling window matters:
  • It ensures field sentiment reflects current practitioner experience, not historical discourse that may no longer be relevant.
  • It prevents a single high-volume period — a conference, a product launch, a viral post — from permanently inflating the corpus for a pattern that has since gone quiet.
  • It means that if a pattern stops generating practitioner discourse, its confidence state will decline on the next refresh, which is itself a meaningful signal.
When the rolling window clears n≥20 statements, the frame’s sentiment read is marked Validated. When the corpus exists but sits below threshold, it is marked Directional only. When the window is empty or too sparse, the frame is held or marked Insufficient signal.

The Spend Data Lag

Spend data lags actual deployment by approximately two quarters. Financial disclosures, analyst reports, and procurement signal all reflect capital that has already been committed — not what is being deployed or experienced in production today.
A frame marked “accelerating trajectory” reflects the direction of capital commitment, not confirmed deployment velocity. The field sentiment signal is more current — practitioner discourse responds to real-world experience faster than financial disclosures do — which is why divergence between the two signals is itself meaningful information. When you see a Capital ahead direction, part of that gap may be explained by the lag: capital committed two quarters ago hasn’t yet produced the practitioner experience that would generate enthusiastic field sentiment. Keep the lag in mind when interpreting direction labels, especially in fast-moving patterns.

Applying Confidence to Your Decisions

Use confidence flags to calibrate — not to discount. A Directional only frame is not worthless. It means the evidence points in a direction but the corpus hasn’t yet reached the validation threshold. If the direction aligns with other signals you’re tracking, it may still be decision-relevant. A practical approach:
1

Check the confidence flag first

Before reading the direction or verdict, note whether the frame is Validated, Directional only, or Insufficient signal. This sets your uncertainty budget for what follows.
2

Read the spend band width

Narrow bands (Tier 1-backed) warrant more weight than wide bands (Tier 2 or Tier 3 only). If the spend figure spans a wide range, factor that into how much you rely on it for a specific investment or deployment decision.
3

Account for the data lag

Add approximately two quarters to the spend signal timeline. If the frame shows accelerating spend, the practitioner experience that follows that capital is likely still forming. Field sentiment, when it clears threshold, will confirm or contradict the trajectory.
4

Combine with your own context

CoDomain frames are market-level reads. Your specific customer segment, geography, and use case may sit at a different point in the curve than the aggregate signal suggests. Use confidence flags to know where the aggregate signal is strong enough to rely on and where you need to supplement with primary research.