Intent Spike Triage
The intent spike list, calibrated — above-threshold accounts with a rising topic trend and the default angle that actually converts.
Sample output for this workflow will appear here once it is captured.
Run the workflow in Claude, ChatGPT, or HG Insights Playground using the buttons below to see real output.
Overview
Triage the accounts currently at or above an intent threshold on your category, with signals in the last 30 days and a rising topic trend. The single-trigger entry point for an SDR onboarding into a territory: returns those accounts with their current intent score, buyer-journey stage and a default opening angle per account, calibrated by intent stage and topic relevance per `intent-data-calibration`.
Use cases
Onboarding's first prospecting run that doesn't waste a week
Your new SDR runs this on day-three with a territory filter and gets a ranked list of accounts above the intent threshold with a rising topic trend and signals in the last 30 days, with the buying function and default angle per account. They build intuition by sequencing real, calibrated signals — not by hunting through dashboards.
Calibrated, not aggregated — `Researching` doesn't lead
The workflow respects `intent-data-calibration`: only `Evaluating` rows make the action list. `Researching` rows go to the watch list as corroboration. SDRs stop opening messages with 'I noticed you're researching X' — the move that destroys credibility on the first email.
View workflow prompt
# Intent Spike Triage
## Parameters
- `{{your_product}}` *(required)* — Your product or category — used to resolve canonical intent topic_ids. Example: `Datadog`
- `{{score_threshold}}` *(required)* — Intent score threshold (0-100). Use 80 — the canonical "strong intent" threshold — unless you have territory-specific tuning.. Example: `80`
- `{{territory_filter}}` *(optional)* — Optional firmographic filter (industry, country, employee band) to scope the territory. Example: `industry: SaaS, country: US`
## Purpose
You are an SDR onboarding into a territory or grading the week's intent surface for {{your_product}}. Return accounts whose current topic score is at or above {{score_threshold}} (default 80), with signals in the last 30 days and a positive `trend`, with default opening angles calibrated by intent stage. HG returns the current score and trend, not threshold-crossing history, so never claim an account "newly crossed" the line. Per [`intent-data-calibration`](https://phoenix.hginsights.com/gtm/skills/intent-data-calibration), `buyers_journey_name: "Evaluating"` is commercial-trigger-grade; `"Researching"` is education-stage and corroborates but does not lead (HG reports only these two stages).
## Process
1. **Resolve topics** — `list_intent_topics` once for {{your_product}}'s category. Capture each topic's `id` and `name`.
2. **Cohort scan** — `search_companies` with `intent.topics.ids` (hex ids from `list_intent_topics`, `inclusion_method: "ANY_PRESENT"`), optionally `intent.signal_score: "MEDIUM"` as a floor. Returns the company list scoring on those topics. Apply {{territory_filter}} as a post-fetch firmographic gate via `company_firmographic` instead.
3. **Per-account intent detail** — for each surfaced domain, call `company_intent` with `topic_ids` (the resolved ids), `fields: ["topics"]`, `granularity: "global"` and `signal_date: { from: <today-30> }`. Keep accounts where a resolved topic's `score` ≥{{score_threshold}} AND its `trend` is positive (the "spike" cut). If no row comes back for the topic, treat as inconclusive (drop or move to Watch).
4. **Stage calibration** — note `buyers_journey_name` per row: `"Evaluating"` rows lead and rank first; `"Researching"` rows go to the watch list, not the action list.
5. **Function hint** — `company_fai` once per surviving account to suggest the buying-function contact via `contact_search` titles only (no enrichment in this workflow — it's a triage step, not a contact-list build).
## Output Format
- `# 📈 Intent Spike Triage — accounts above the line and rising`
- `## Action list` — table: rank | domain | topic | buyers_journey_name | score | trend | suggested function | default opening angle
- `## Watch list` — `"Researching"`-stage rows (corroborate but don't lead)
- Per-row citation: `[company_intent]` for score, `[search_companies]` for the topic cohort, `[company_fai]` for function
## Quality Checklist
- Every Action-list row is `buyers_journey_name: "Evaluating"` — `"Researching"` rows ALWAYS go to Watch
- Spike check confirms the topic `score` is ≥{{score_threshold}} AND `trend` is positive, with signals dated in the last 30 days — "newly crossed" is never claimed
- Topic-absence is inconclusive (not "no intent") — fall back to the `search_companies` intent filter per the skill
- No contact enrichment in this workflow (its role is triage, not contact-list building)