NUSAWAN Request access

Cognitive agents / Social media

Connect emerging conversations to observed sales outcomes.

Nusawan is developing a pilot workflow to help marketing teams discover permitted signals, form a testable sales hypothesis, and assess whether human-approved content is associated with leads, conversions, and recognized revenue. Engagement explains the journey; established sales remain the ultimate outcome.

Request early access See the workflow

Coverage with clear boundaries

Different platforms.
Different permissions.

Nusawan’s coverage model respects what each platform officially makes available. It combines three practical modes instead of pretending every network offers the same access.

01

Official public signals

Live only where the platform supports it.

Bluesky provides public-network streaming through AT Protocol. YouTube provides quota-limited discovery data, and X offers paid public search subject to access, caps, and policy. Neither YouTube nor X is presented as a universal live firehose.

Bluesky / live AT ProtocolYouTube / discovery APIX / paid API
02

Authorized operations

Business context only where permission allows.

Approved or owner-authorized access can support Pinterest Trends, LinkedIn Pages, Facebook Pages, Instagram Business or Creator accounts, WhatsApp Business, Google Business Profile, and installed Discord servers. Eligibility, review, consent, and scope still apply.

PinterestLinkedInFacebookInstagramWhatsApp BusinessGoogle Business ProfileDiscord
03

Permission-scoped pilots

Open protocols do not erase product boundaries.

DiVine may support a Nostr/Funnelcake discovery pilot, and Skylight public signals may be assessed through AT Protocol. Noplace requires explicit permission. These labels do not claim a mature contracted API or autonomous publishing route.

DiVine / protocol pilotSkylight / AT ProtocolNoplace / permission required
04

Guided human research

Curated inputs, not automated extraction.

Commercial research can use TikTok Creative Center and approved human observations. Lemon8, UpScrolled, and Fanbase remain human-curated under their current terms. Pinterest also falls back to this mode when approved Trends access is unavailable.

TikTokLemon8UpScrolledFanbase

Telegram is limited to owned or consented bot and channel operations and is not enabled as an AI trend-ingestion source under current API terms. Discord is limited to installed servers and approved intents. LinkedIn, Facebook, and Instagram do not imply platform-wide listening. WhatsApp Business is an opt-in messaging surface, while Google Business Profile supplies owner-location context rather than market-wide trends. Availability remains subject to permissions, account eligibility, regional rules, supported integrations, and revalidation of current platform terms. Nusawan does not claim official partnership with any listed platform.

Sales evidence chain

Every test stays traceable to its observed outcome.

The outcome may be a recognized sale, no sale observed, an unattributed sale, or insufficient evidence. The ledger keeps each result visible.

  1. 01 / SIGNAL

    Register the hypothesis

    Record the permitted source, topic, audience, market, product, timestamp, expected customer action, and the reason a sale may follow.

  2. 02 / CONTENT

    Approve the experiment

    Give each channel-specific draft a campaign and content ID. A named human reviews the claim, brand fit, risk, offer, budget, and publishing route.

  3. 03 / RESPONSE

    Capture the journey

    Keep impressions and engagement as diagnostic evidence, then connect attributable clicks, messages, calls, QR scans, coupon uses, and store visits where consent and systems allow.

  4. 04 / LEAD

    Qualify real intent

    Deduplicate enquiries and preserve source, market, product, offer, and status in the authorized CRM or sales workflow. A reaction alone is not a lead.

  5. 05 / CONVERSION

    Join the transaction

    Link an order, booking, or point-of-sale transaction through approved platform attribution, UTM or deep-link IDs, referral codes, QR codes, coupons, or privacy-safe first-party keys.

  6. 06 / OUTCOME

    Record the result

    Confirm the outcome in the system of record, account for cancellations, returns, and refunds, and label the evidence directly attributable, incrementally supported, directional, or unattributed. “Assisted” remains a touch role, not an evidence level.

Evidence contract

Measure the path before the content goes live.

Each experiment declares its attribution window, touch model, deduplication rule, conversion event, revenue source, and refund treatment before activation. The ledger minimizes copied personal data: authorization alone is not a reason to retain it. Source permission, allowed purpose, expiry, revocation, deletion, and derived provenance are recorded with the observed outcome.

Decision record01

Why this market, message, and offer?

Hypothesis, source, audience, product, market, channel, expected action, owner, approval, version, and activation time.

Attribution record02

Which approved paths could connect to a sale?

Execution and touch IDs, touch role and weight, model and version, tagged link or deterministic key, declared window, deduplication decision, and permitted join method.

Outcome record03

What did the business recognize?

Qualified lead, order or POS reference, product, market, sale time, net revenue state, return status, and directly attributable, incrementally supported, directional, or unattributed evidence.

Market focus / Continuous expansion

Establish the sale.
Then test the next market.

Observed, established sales provide commercial evidence about where demand may be concentrated. Nusawan is designed to use that evidence to propose the next bounded experiment—not to confuse attention, attribution, or sales contribution with causality or market share.

01

Establish

Build a comparable sales baseline.

Group recognized sales by audience, market, product, offer, channel, and time period. Compare conversion, net revenue, acquisition cost, margin, repeat purchase, and returns only where the underlying business data is reliable.

Recognized saleNet outcomeConsistent definitions
02

Focus

Concentrate where sales evidence repeats.

Prioritize combinations associated with repeated, acceptable sales quality. Market share requires total brand sales and qualified total category sales for the same category, geography, channel, period, currency, gross-or-net treatment, and sell-in or sell-through basis. Incomplete brand coverage is sales contribution, not market share.

AudienceGeographyProductChannel
03

Expand

Earn each adjacent market test.

Recommend testing one meaningful boundary at a time—audience, geography, channel, offer, or product—with a defined budget, comparison baseline, success threshold, stop condition, observation window, and named human approval.

Bounded testHoldout where practicalHuman gate

Sales establish commercial evidence, not automatic causality. Seasonality, promotion, availability, paid media, pricing, and offline activity remain possible explanations. The evidence design preserves these factors in the decision record and treats expansion as a controlled test.

Social intelligence / Sales evidence

Make the sale visible.
Make expansion deliberate.

Tell us which products, markets, sales systems, and authorized channels should form your first evidence chain.

Request early access