Context determines response.
BHASM is watching the world your customers live in — every day, in every market, in every cultural moment. The same message in different contexts produces different outcomes. Here is what world intelligence does, and how it affects every send.
What world intelligence is
A continuous monitoring layer that reads external signals affecting customer behaviour. A customer who is buying normally suddenly approaches a festival — BHASM adjusts urgency and message tone before any internal data has shifted.
The intelligence is not a feature added to the brief. It is consulted on every score, every message generation, and every send decision. It runs continuously, independent of when a particular tenant's brief fires.
No tenant ever has to enable it, configure it, or maintain it. It is present in the system before the first customer is connected.
Signal categories
Cultural / festival signals
The full commercial calendar for every market BHASM serves. US & Global: Black Friday, Cyber Monday, holiday gifting, back-to-school, Mother's Day, Father's Day, Memorial Day, Labor Day, regional bank-holiday windows. EU: Boxing Day, Sinterklaas, regional Easter and Christmas windows. India: pre-Diwali gifting intent, post-Eid windows, Akshaya Tritiya for jewellery, Onam, regional wedding seasons. SEA & LATAM: Lunar New Year, Singles Day, Ramadan-end windows, Carnival, regional independence days.
Each signal has a known effect: pre-festival amplifies retention urgency, post-festival suppresses cold outreach for 3–5 days while customers finish their purchase clusters.
Economic signals
Salary credit windows in India are a documented pattern in SMB customer behaviour — spending capacity rises around month-end. Quarter-end B2B purchase patterns matter for SaaS and services.
Macroeconomic sentiment by market — inflation prints, currency moves, fuel prices. A premium send during a fuel-price shock day reads tone-deaf even if the message is technically correct.
Disruption signals
Political events, natural events, economic shocks. BHASM monitors disruption signals at three tiers:
- Global — major events with cross-market effect.
- Country — national-level political or economic shock.
- City — localised event (flood, riot, strike, transit failure).
When a disruption signal is active for a customer's geography, all promotional sends are suppressed for 24–72 hours depending on event severity. The hold is automatic and logged. The send queue resumes when the signal clears.
Industry signals
Some industries have non-obvious seasonal patterns BHASM tracks specifically:
- Major sport finals / cup windows (Super Bowl, World Cup, IPL) — attention fragments, consumer-app engagement drops.
- Exam / back-to-school window — parents distracted; education-adjacent products see lower response.
- Wedding season (regional) — gifting + premium fashion intent rises sharply.
- Severe-weather window (monsoon, hurricane season, winter storm) — delivery delays affect message tone (avoid "ships today" framing).
- Fiscal year-end (varies by market) — B2B services receive lower attention during audit + close periods.
How signals modify scoring
World signals apply as a multiplier on the base urgency score. The modifier is bounded so it can amplify a real signal or suppress a wrong-moment send — but it never invents urgency from thin air, and it never silences a critical alert into nothing.
| Context | Modifier | Effect |
|---|---|---|
| Pre-festival high intent | × 1.3 | Surface earlier, higher in brief |
| Pay-cycle window (regional) | × 1.2 | Mild amplification for purchase-likely customers around bi-weekly / monthly / month-end pay dates |
| Normal context | × 1.0 | No effect — the default |
| Off-season for seasonal customer | × 0.7 | Suppress until window opens |
| Post-disruption | × 0.6 | Hold most sends; allow only critical retention |
The multiplier is applied to the score after the five base inputs sum. A customer at urgency 80 in pre-festival becomes 104 — clamped to 100, surfaces at the top. The same customer at urgency 80 in post-disruption becomes 48 — drops below threshold, no send.
Data sources and refresh cadence
| Source type | What it provides | Refresh |
|---|---|---|
| RSS feeds (50+) | Industry, regional, national news streams | Every 30 minutes |
| NewsAPI / GNews | Targeted industry × region news | Every 6 hours (cached) |
| OpenWeatherMap | Weather + disruption signals by city | Every 3 hours |
| Cultural calendar | Internal static calendar of festivals + windows | Annual review |
| Economic indicators | Salary windows, quarter ends, fiscal events | Internal rules |
All signals are cached at 6-hour intervals to avoid API rate limits and to keep cost bounded. The cache is fresh enough for the brief cadence and stable enough for repeat calls within a single brief generation.
When no signal is available: default modifier is 1.0. The system never blocks a send because a world signal API failed — it defaults to neutral and continues. This is a deliberate design call: world intelligence amplifies and suppresses, but never silently halts an otherwise-valid send because the signal source went down.
A worked example
A customer in a fashion-adjacent category, urgency score 75 from base inputs. Send window: Tuesday morning. The example below is shown for the two largest calendars BHASM serves — the principle is identical in every market.
Scenario: A customer in Chicago, fashion category. Today is mid-November, 10 days before US Thanksgiving and the Black Friday window.
World signals consulted:
- Cultural calendar: pre-Black-Friday window opens in 10 days → modifier × 1.3
- Economic: bi-weekly pay-cycle window active → modifier 1.2× (half-weight when combined)
- Disruption: no active event for Chicago
- Industry: seasonal launch window confirmed → delivery confidence restored
Scenario: A customer in Mumbai, fashion category. Today is mid-October, 12 days before Diwali.
World signals consulted:
- Cultural calendar: pre-Diwali window opens in 12 days → modifier × 1.3
- Economic: regional payday window active → modifier 1.2× (half-weight when combined)
- Disruption: no active event for Mumbai
- Industry: monsoon ended last week → delivery confidence restored
Final modifier (both scenarios): × 1.35 (capped at 1.3 base + 0.05 economic). Score becomes 75 × 1.35 = 101 — clamped to 100. Surfaces top of brief. Claude crafts the message. Send approved.
Two weeks later, post-festival or post-Black-Friday, same customer, same base signal — world context suppressed by the post-event cool-off. The score drops below the brief threshold. Hold-out is the right call: customer just made a major seasonal purchase elsewhere; a re-engagement send right now would feel out of touch.