Written from 16 named sources · Aug 23 · first result Proposal Readiness & Consortium Strategy — FDA RFQ 136111 Social Listening, Cannabinoid Market Surveillance, and Survey Analytics Period of Performance: September 30, 2026–September 29, 2027 Acquisition Method: Single-award, firm-fixed-price (FFP), lowest-priced technically acceptable (LPTA) Executive Summary & Strategic Posture Bottom line FDA RFQ 136111 is a brand-name-or-equal, subscription-led evidence service, not a stand-alone social-listening software procurement. The successful quoter must demonstrate that it can provide—on day one or within the RFQ’s specified onboarding periods—a mature, FDA-usable ecosystem comprising: Existing longitudinal market and consumer datasets for CBD, non-intoxicating cannabis products, general population consumption, legal-cannabis consumers, and women’s health; Daily social-listening coverage of TikTok, Instagram, and Reddit, including FDA-configurable segmentation and click-through to verbatim public posts; Quarterly, semiannual, annual, and ad hoc research outputs in dashboard, Excel, Word, PowerPoint, webinar, and briefing formats; Unlimited FDA-user access, support, and training; A Section 508-compliant reporting and service environment; and A price low enough to prevail in an LPTA evaluation after being found technically acceptable. The key competitive reality is that Brightfield Group likely benefits from pre-existing survey waves, cannabis category panels, taxonomies, social datasets, dashboards, and analyst workflows. A consortium cannot credibly “build equivalent” survey history or an established >90,000-respondent general-population trend series in the time remaining before award. The viable strategy is therefore acquire-and-integrate, not greenfield build. Bare minimum threshold to be competitive under FAR 52.211-6 Under a brand-name-or-equal posture, the quote must affirmatively establish equality to the RFQ’s required performance characteristics—not merely propose comparable technology. The quote should use an FDA requirement-by-requirement crosswalk and make no unsupported claims. Competitive threshold Minimum acceptable evidence in quote Non-competitive position Market intelligence Existing, statistically sound CBD and emerging non-intoxicating cannabinoid market data covering market size, brands, forms, claimed content, and state-regulated products; annual refresh capability. A promise to conduct desk research after award. CBD/non-intoxicating consumer research Existing representative consumer research plus two refreshed waves during the 12-month period. A one-time custom survey with no trendable baseline. Daily social listening Operational TikTok, Instagram, and Reddit data coverage; daily refresh; trained AI segmentation; filters by age group, sex, persona, and FDA-defined attributes; verbatim click-through. Keyword alerts, text-only monitoring, or a dashboard with no platform-specific data-rights proof. General Population Survey A representative, trendable U.S. adult survey program, quarterly fielded, spanning cannabis, kratom, psilocybin, functional ingredients, claimed benefits, and need states. The RFQ specifically calls for a historical sample exceeding 90,000 since 2021. Proposal to start a new panel or create a new trend series after award. U.S. Cannabis Consumer Survey Existing census-balanced legal-cannabis / dispensary-shopper survey history since 2018, plus two new waves with unique respondents. General cannabis surveys that cannot isolate legal dispensary shoppers or provide historic trend analysis. Women’s Health Study One representative U.S. women’s survey of 5,000 respondents across fertility/prime, perimenopause, and post-menopause; product and need-state coverage plus social-conversation analytics. A small qualitative study, social-only analysis, or a survey with insufficient sample size. Reporting and support Quarterly reports, FDA-directed emerging-issue analyses, webinars, Excel/Word/PowerPoint outputs, unlimited users, support during required hours, and two live trainings. Dashboard license only. Federal readiness Active SAM registration and UEI; Section 508 ACR; IPP readiness; commercial-item compliance; no macros/executable files; firm price held through September 30, 2026. A technically capable team without an eligible, compliant prime contractor. Recommended strategic posture: “Evidence Subscription Plus” The recommended offer is a low-risk, modular evidence subscription delivered under one prime contractor: Buy the irreplaceable assets: historical surveys, market data, and platform-authorized or licensable social data. Configure and validate the differentiators: FDA-specific cannabinoid taxonomy, multimodal classification, quality controls, reporting templates, and ad hoc alerting. Avoid overbuilding: FDA did not ask for a bespoke data lake, a new government-hosted platform, or a novel AI research platform. Building those before award adds cost and schedule risk without improving technical acceptability. Lead with auditability over “AI magic”: Every substantive trend, classification, alert, and report finding must be traceable to a source post, survey instrument, market-data methodology, or documented analytic rule. The FDA regulatory-science model is multidisciplinary and evidence-centered; its research programs combine epidemiology, behavioral science, toxicology, addiction research, economics, and marketing expertise. [5] The quote should reflect that orientation through documented methodology, transparent confidence limitations, and human validation rather than unsupported automated conclusions. Bid/no-bid recommendation Bid only if the consortium can secure, before submission, binding letters of commitment or subcontractor commitments for all four critical assets: Existing longitudinal General Population Survey history meeting the RFQ’s >90,000-since-2021 criterion; Existing legal-cannabis / dispensary-shopper survey history dating to 2018; Authorized or legally licensed TikTok, Instagram, and Reddit data access sufficient for daily monitoring and verbatim link-out; and A qualified federal prime with active SAM/UEI, commercial-item contracting experience, IPP readiness, and ability to submit the required Section 508 ACR. Without those assets, the consortium can be technically innovative but is unlikely to be technically acceptable. Because this is LPTA, an elegant but higher-priced custom-build proposal will lose to an acceptable subscription offering. Consortium Operating Model & Governance Recommended prime–subcontractor architecture The team should present to FDA as one accountable prime contractor with four specialist subcontractors. The prime must own contractual performance, invoice submission, customer support, quality assurance, data-rights assurance, Section 508 documentation, and all communications with the Contracting Officer (CO). Entity Contractual role Primary accountability Prime contractor / GovTech lead Sole quoter and awardee Contract, FFP price, SAM/UEI, invoicing, security, QA, dashboard access, support, reporting, subcontract management Social-data aggregator Critical subcontractor Licensed/platform-permitted TikTok, Instagram, and Reddit collection; data provenance; daily updates; platform-change response Survey and market-data provider Critical subcontractor Historic survey series, market sizing, quarterly and periodic waves, weighting, crosstabs, methodology documentation Cannabinoid NLP and regulatory-science SME Specialist subcontractor Taxonomy, cannabinoid and claim coding, model validation, emerging-product watchlist, human adjudication Deliverable and reporting lead Specialist subcontractor or prime function Quarterly reports, ad hoc briefs, FDA meetings, accessible Excel/Word/PowerPoint outputs, webinar support Work-share allocation A practical FFP work-share model is: Workstream Prime Social-data partner Survey / market-data partner Cannabinoid NLP / regulatory SME Reporting lead Contract management, QA, compliance, help desk 100% — — — — Dashboard access and user administration 70% 20% 10% — — Social ingestion and normalization 20% 70% — 10% — Cannabinoid taxonomy and model validation 15% 20% 10% 55% — CBD market data 15% — 85% — — Consumer survey operations 10% — 90% — — Quarterly and ad hoc reports 35% 15% 25% 15% 10% FDA training, webinars, stakeholder briefings 50% 10% 15% 10% 15% Governance model 1. Prime-led Program Management Office (PMO) The prime should establish a weekly delivery review, monthly data-quality review, and quarterly executive review. The PMO holds authority for schedule, release approval, change control, incident escalation, subcontractor performance, and invoice readiness. 2. FDA-facing Integrated Product Team (IPT) A single Program Manager is the operational point of contact. The team should support the COR, while recognizing that only the CO may authorize changes to scope, price, period of performance, or delivery schedule. 3. Scientific and Data Quality Board The board should include the survey methodologist, biostatistician, cannabinoid regulatory SME, social-data lead, and reporting lead. It approves: Taxonomy changes; Classification thresholds; Annotation protocols; Survey weighting and trend-comparison methods; Emerging issue definitions; Publication/reporting language; and Confidence statements and limitations. 4. Data Rights and Platform Continuity Council This council maintains platform-access documentation, tracks API and terms-of-service changes, assesses data-retention constraints, and activates fallback procedures. This is necessary because third-party platform access restrictions can materially alter data availability and granularity. [4] SAM.gov, UEI, and federal-administration controls Before quote submission, the prime must verify: Active SAM.gov registration through the required offer validity period; UEI included in the quote; SAM representations and certifications completed consistent with the RFQ; No exclusions, suspensions, or debarment concerns for the prime or critical subcontractors; Current bank and business-point-of-contact information in SAM; Ability to receive and process IPP registration communications; Treasury Invoice Processing Platform enrollment or a documented plan to complete enrollment immediately after award; Acceptance of Net 30 payment after FDA acceptance; and A process for complete invoice logs and subcontractor support documentation. The prime must flow down applicable commercial-item, debarment, trafficking, accessibility, security, and subcontracting requirements to all subcontractors. The quote should identify each critical subcontractor, its role, and its legal entity name, avoiding vague “partner ecosystem” descriptions. Past Performance leverage Because the RFQ’s stated evaluation is LPTA, past performance is primarily a responsibility and risk-reduction asset, not an overt best-value discriminator. Still, it should be used to make the lowest-priced technically acceptable quotation feel low-risk. Past performance source How to deploy it Federal IT/data-subscription performance Demonstrate user provisioning, secure dashboards, help-desk service, Section 508 delivery, and invoice discipline. Public-health or FDA-adjacent research Demonstrate scientific integrity, survey operations, and defensible analytical reporting. Social-intelligence deployments Demonstrate daily data feeds, high-volume data processing, dashboard uptime, and alerting. Cannabis, hemp, nicotine, supplements, or consumer-health research Demonstrate domain fluency without creating a real or perceived conflict of interest. Survey-panel / longitudinal research experience Demonstrate representative sampling, weighting, historical trend analysis, respondent de-duplication, and data delivery. Independence rule: Exclude, or tightly firewall, partners whose principal business is cannabis brand marketing, product sales, retail promotion, lobbying, advocacy, or investment. A cannabis regulatory adviser may be useful, but it must be conflict-screened and should not control the analytical conclusions. FDA emphasizes independent, objective science as the basis for regulatory decisions. [16] Partner Capability Matrix Mission-critical role Core capabilities required Build vs. procure position Minimum proof required before quote submission Selection criteria Prime / GovTech Lead Federal contracting; FFP estimating; SAM/UEI; IPP; user access; security; 508; subcontract management; help desk; dashboard operations; quality management Build/configure internally; procure specialists Active SAM/UEI; FFP pricing model; Section 508 ACR plan; named Program Manager; commercial-item contracting record Federal past performance; financial capacity; no exclusion; ability to accept full contractual risk; secure SaaS operations Social Data Aggregator TikTok, Instagram, Reddit coverage; lawful collection; daily refresh; data provenance; API/rate-limit management; URL preservation; deletion and outage management Procure Written evidence of platform-authorized, licensed, or otherwise permitted access; data dictionary; daily refresh SLA; platform-change contingency Coverage quality; lawful rights; historical data availability; API resilience; data localization; transparent limitations Cannabinoid NLP / Taxonomy Expert CBD, Delta-8, Delta-9, Delta-10, HHC, THCA, CBN, CBG, hemp-derived intoxicants; product form, ingredients, health claims, adverse-event indicators, youth-appeal signals, coded slang Configure/build on top of licensed tools Versioned taxonomy; sample labeled corpus; model-card template; human-review protocol; SME credentials Regulatory-science knowledge; taxonomy governance; explainability; classification performance; lack of commercial conflicts Survey Research Provider Representative U.S. panels; weighting; longitudinal trend data; quarterly fielding; legal-cannabis consumer panel; women’s health sample; respondent de-duplication; crosstabs and microdata Procure; do not build Documentation of historic samples and waves; methodology; sample source; weighting; ability to meet wave frequencies and 5,000-woman study Existing historical series; statistical rigor; field speed; panel quality; cannabis and women’s-health experience; respondent privacy controls Market Intelligence Provider CBD and emerging-product market sizing; brand, category, form, claimed-content tracking; state-regulated products; annual refresh Procure Existing market data schema; refresh calendar; sources/methodology; sample outputs Established category coverage; transparent market methodology; cannabinoid breadth; trend continuity Deliverable / Reporting Lead Quarterly reports; emerging-issue briefs; webinars; Word, Excel, PowerPoint; accessible reporting; executive synthesis; FDA meeting support Prime-led with specialist support Sample quarterly report; alert template; accessible document-production workflow; named editor and analyst Federal writing quality; Section 508 knowledge; ability to explain uncertainty; turnaround time Public-Health / Epidemiology Advisor Sampling-frame review; survey validity; prevalence estimation; trend methods; inter-rater reliability; uncertainty statements Subcontract or consultant CVs; methodology review plan; deliverable sign-off role Relevant academic/public-health experience; independence; practical availability Privacy, data-rights, and records counsel Terms-of-service review; privacy/PII minimization; retention; copyright; public-record and platform-risk advice Procure targeted advisory support Data-rights memorandum; retention schedule; incident response role Platform-law fluency; federal privacy understanding; availability during award transition Recommended selection order: secure the survey/market-data provider first, then social-data access, then the federal prime. These three are gating dependencies. A sophisticated NLP team cannot cure a missing historic survey series. Technical & Data Architecture Design principle The architecture should be presented as a federated subscription and evidence environment: retain source-system provenance, normalize only what is necessary, and provide FDA with a common dashboard and reporting layer. Do not promise collection of private accounts, direct messages, login-gated information, or personally identifying user records unless expressly authorized by FDA and permitted by the applicable platform and law. FDA’s social-media policy recognizes user-generated videos, photos, microblogs, blogs, and related third-party social-platform content, while requiring compliance with terms of service, privacy, security, Section 508, copyright, and records-management requirements. [6] The architecture separates source collection, normalized analytic data, and human-validated findings. This permits FDA users to distinguish raw platform material, model-inferred tags, and analyst-confirmed findings. A. Platform ingestion pipelines Platform Required ingestion approach Daily processing requirements Evidence retained TikTok Authorized API, platform-approved research access, or commercial licensed provider access; no proposal dependence on prohibited scraping Video metadata, captions, hashtags, comments where permitted, audio transcription, OCR, visual signals, engagement data, URLs Source URL, collection date/time, permitted metadata, content hash, source-access method, deletion/unavailability status Instagram Authorized Meta access or licensed provider data; use platform-permitted public content only Captions, comments where available, hashtags, images, reels/video metadata, OCR, visual product/logo/packaging recognition, URLs Source URL, media type, timestamp, normalized text, image/video analytic outputs, data-rights label Reddit Authorized API or licensed provider access; public posts and comments subject to platform terms Subreddit, title, body, thread structure, comments, flair, engagement, links, deleted/removed status Post/comment ID, permalink, subreddit, timestamps, collection timestamp, revision/deletion flag Required controls Daily collection job health checks; Connector-level service-level objectives; Rate-limit monitoring and retry logic; Platform-specific data dictionaries; Provenance field for every record; Duplicate and cross-post detection; Deleted, edited, inaccessible, and unavailable-content labeling; Separation of raw, derived, and adjudicated fields; Automated alerts when collection volume materially deviates from historical baselines; Quarterly platform-access review; and A manual-source review procedure for high-priority emerging signals. B. Cannabinoid taxonomy and NLP design The taxonomy should be FDA-configurable and version controlled. It must account for changing cannabinoid terminology, product forms, claims, slang, and illicit-market euphemisms. Social-media conversations commonly rely on implication, coded language, and emotional nuance; AI can help detect patterns at scale but should be paired with human interpretation. [13] Core taxonomy domains Domain Required tags Cannabinoid / ingredient CBD, cannabidiol, Delta-8 THC, Delta-9 THC, Delta-10 THC, HHC, HHCP, THCA, THC-P, CBG, CBN, hemp-derived intoxicants, full spectrum, isolate, synthetic or semi-synthetic descriptors Product form Gummies, beverages, vapes, cartridges, disposables, tinctures, capsules, flower, pre-rolls, edibles, topicals, concentrates, pouches, pet products Claim type Disease-treatment, therapeutic, symptom relief, sleep, anxiety, pain, stress, mood, menopause/perimenopause, sexual health, focus, energy, recovery, safety, natural, non-intoxicating, dosage/potency Potential harm / concern Adverse experience, accidental ingestion, pediatric or youth-related concern, impairment, drug interaction, contamination, mislabeling, dosage concern, withdrawal, dependence, emergency-care reference Commercial signal Brand, seller, retailer, price, discount, shipping, direct message prompt, marketplace link, dispensary, interstate-shipping language Audience / appeal indicator Youth-oriented imagery or language, influencer/creator endorsement, cartoon/candy cues, lifestyle segment, age self-disclosure where publicly stated and permitted Sentiment and intent Positive, negative, neutral, uncertainty, seeking advice, reporting experience, purchase intent, cessation/substitution discussion, warning or complaint Regulatory relevance Unsubstantiated claim, possible illegal sales activity, possible state-regulated-product reference, product availability, novel cannabinoid emergence The system should apply NLP, speech-to-text, optical character recognition, image classification, and video-frame analysis. A prior Instagram study showed that combining machine learning with manual annotation can detect suspected illicit drug-selling posts; however, the study’s data collection and platform context were specific to its time and methods and should not be represented as a current permission model. [11] The consortium’s approach must be platform-compliant and independently validated. C. Human-in-the-loop validation and scientific controls The platform should not represent model outputs as regulatory determinations. AI assists triage; trained human analysts adjudicate material findings. Control Minimum implementation Gold-standard labeling Create an FDA-approved labeled sample during inception and refresh it quarterly. Double coding Double-code a statistically meaningful subset of high-risk and ambiguous posts. Inter-rater reliability Track Cohen’s kappa or equivalent agreement measures for key taxonomic categories. Model performance Report precision, recall, F1, false-positive rate, false-negative rate, and calibration by platform and major class. Threshold governance Use different thresholds for routine discovery versus high-priority alerts; human review required for alerts. Drift detection Monitor novel terms, changing slang, model degradation, and class imbalance. Model documentation Maintain model cards, taxonomy versions, training-data provenance, intended use, limitations, and change logs. Auditability Preserve source link, collection timestamp, inference timestamp, model version, reviewer identity/role, and adjudication result. D. Survey execution design The survey provider must demonstrate that the RFQ’s historical requirements already exist. The consortium should not imply that a new survey can substitute for RFQ-mandated historical continuity. Survey program Required delivery model Proposed analytic controls CBD and emerging non-intoxicating consumer data Two updated waves during the period of performance, linked to prior representative data Probability- or high-quality panel sample; weighting to U.S. adult benchmarks; pre-specified questionnaire core; unique respondent controls; documented trend breaks General Population Survey Quarterly waves; U.S.-representative; cross-category adoption, behavior, need states, conditions, and social linkage Stable core questionnaire; rotating FDA module; weighted estimates; age/sex/life-stage cuts; confidence intervals; trend comparability memo U.S. Cannabis Consumer Survey Two census-balanced waves among legal-cannabis / dispensary shoppers; new unique respondents Clear screener; legal-market purchase path; product, symptoms, preferred effects, dosage, occasion, purchase frequency, and history modules Women’s Health Study One 5,000-U.S.-women wave by fertility/prime, perimenopause, and post-menopause Life-stage quota and weighting plan; physical/emotional/financial dimensions; cannabis, supplements, functional CPGs, beverages, topicals, feminine health, and hygiene modules E. FDA dashboard and reporting engine The dashboard should provide unlimited authorized FDA users with: Search and filters by time period, platform, product, cannabinoid, claim, condition, sentiment, demographic proxy, persona, and regulatory relevance; Trend views with denominator/methodology notes; Click-through to permitted verb