Cashew Research targets the $90B insights market: fresh human data + AI summaries, C$1.5M pre-seed and Disrupt win

# Cashew Research targets the $90B insights market: fresh human data + AI summaries, C$1.5M pre-seed and Disrupt win TechCrunch profiled Calgary-based Cashew Research on December 9, 2025. The startup targets time and cost in the $90 billion insights industry. It blends fresh human responses with AI-generated summaries. The result is decision-ready research in days, not weeks. Moroccan teams face similar pressures. Budgets are tight and timelines are short. Multilingual audiences add complexity. A human-plus-AI model fits these realities. ## Key takeaways - Cashew builds fresh, bespoke datasets, then uses AI to synthesize findings fast. - The model addresses a common gap in Morocco: speed without sacrificing data quality. - Multilingual support and human oversight are vital for Morocco's diverse audiences. - SMBs gain access to custom research that was often out of reach. - Governance matters. Data protection, consent, and bias controls must be clear. ## What Cashew Research is building Cashew automates core research workflows. It designs research plans and survey instruments from the client's question. It gathers responses from real people. It then applies AI to produce clear, client-ready reports. The positioning is intentional. Generic LLM desk research often rehashes the public web. Traditional firms offer depth but can be slow and expensive. Cashew aims to sit between those options. The promise is simple. New human data meets fast AI synthesis. The goal is decisions in days. That is attractive to teams under constant launch pressure. ## The tech inflection behind the model Addy Graves is a market-research veteran. Client asks for full primary studies in days used to be unrealistic. Recent AI advances changed the calculus. Automation now accelerates survey design, methodological rigor, and report formatting. These workflows are repeatable and pattern-based. AI can draft, check, and iterate quickly. Humans still set objectives and review quality. The stack compresses timelines without gutting standards. ## Human-in-the-loop, by design Cashew is not pure automation. Every project still collects new responses from people. AI augments the process rather than replacing it. This is crucial for trust and nuance. Human oversight helps with cultural tone and edge cases. It also provides accountability for methodological choices. The blend supports speed with context. It reduces the risk of AI hallucinations. ## A compounding data advantage Over time, anonymized results feed a proprietary database. That corpus can inform future studies. It improves benchmarks and prompts. It also strengthens quality checks. This is a flywheel. More projects create better guidance for the next project. Teams see more consistent outputs. The value increases with scale. ## Company milestones that matter Cashew was founded in 2023 by CEO Addy Graves and COO Rose Wong. The early focus is CPG, especially food and beverage. The company raised C$1.5 million in pre-seed financing. It plans a seed round in early 2026 targeting up to $5 million. Cashew is pushing into the U.S. and deeper B2B penetration. It also earned credibility in 2025. The company was selected for TechCrunch's Startup Battlefield 200. It won the Enterprise Stage pitch competition at Disrupt. These signals suggest product-market momentum. They also reflect a crowded AI-for-marketing field. Differentiation will depend on results and trust. Buyers will test both. ## Why this approach resonates in Morocco Moroccan brands work across languages and regions. Darija, French, Arabic, and Amazigh all appear in campaigns. Consumer behavior also differs by city, income, and channel. One-size-fits-all research rarely works. Speed is a constant constraint. Launch windows are tight in CPG, telecom, and travel. Public agencies face urgent feedback needs. A days-long research cycle is powerful in these contexts. Budgets also matter. Many SMBs could not afford custom studies. Lower production costs open the door. Teams can test ideas earlier and more often. ## Policy and ecosystem context in Morocco Morocco has expanded digital programs in recent years. Public bodies encourage modernization and startup activity. Regulators oversee data protection practices. Organizations are expected to manage consent and security. Events like GITEX Africa have brought global and regional AI players to Morocco. Local incubators and tech parks support founders. Universities grow data science talent. The pipeline of applied AI skills is improving. Enterprises continue to invest in analytics. Sectors include agriculture, mining, retail, and logistics. These teams seek faster market readouts. Practical AI is getting a fair hearing. ## Practical use cases for Moroccan teams - CPG concept testing: flavor, packaging, and price sensitivity across Casablanca, Rabat, Tangier, and Marrakech. - Tourism messaging: A/B test copy for European and Gulf audiences in French, English, and Arabic. - Telecom offers: evaluate prepaid bundles and churn risk signals among youth and rural users. - Financial services: test onboarding flows and trust cues for digital wallets. - Public services: gather citizen feedback on transport, waste, and local service satisfaction. - Agriculture value chains: validate labels and formats for agrifood exports. Each use case benefits from fresh data and quick synthesis. Localization is essential. Mobile-first surveys improve reach. Clear incentives improve response rates. ## Implementation playbook for Morocco Start with a focused question. Define the target audience and required sample size. Decide on languages and reading level. Set a clear timeline and budget. Build for mobile. Keep surveys short and direct. Offer fair incentives. Respect respondent time and privacy. Plan for multilingual quality. Use native translators for Darija, French, Arabic, and Amazigh where needed. Back-translate critical questions. Pilot test before launch. Integrate results with your workflow. Align tags and metadata with your BI tools. Push summaries to dashboards. Close the loop with decision owners. ## Quality and governance guardrails - Sampling: ensure geographic and demographic diversity to reduce bias. - Consent: provide clear explanations of data use and retention. - Privacy: align with Morocco's data protection requirements and internal policies. - Incentives: avoid crowding bad actors; add fraud detection and device checks. - Language: verify that translations preserve meaning and tone. - AI checks: require human review of summaries and recommendations. These guardrails protect credibility. They also support long-term adoption. Trust is the currency in research. It cannot be outsourced entirely to software. ## How it compares to the status quo LLM desk research is fast but shallow. It mostly repackages public information. It struggles with local nuance and proprietary questions. It also risks hallucinated claims. Traditional research is rigorous but slow and costly. It can struggle to meet rapid release cycles. Many SMBs cannot sustain repeated projects. Opportunity costs grow. Cashew's model aims for the middle path. It keeps humans in the loop and data fresh. It uses AI where it adds leverage. It removes time sinks without gutting quality. ## What to ask vendors before you buy - Who are the respondents and how are they verified? - How do you handle consent, privacy, and data retention? - Can you support Darija, French, Arabic, and Amazigh? - How do you detect and remove low-quality responses? - What is your approach to bias identification and mitigation? - Can I audit the AI's chain of reasoning or prompts? - How are benchmarks built and updated over time? - Where is data stored and who has access? Clear answers help teams manage risk. They also set expectations. Transparency beats magic. Every time. ## Risks and limitations to manage AI summaries can miss cultural nuance. Human review is essential. Samples can skew urban without careful outreach. Mobile coverage gaps can limit rural reach. Small base sizes can mislead. Pair fast tests with follow-up qualitative work when needed. Language drift can creep into translations. Litmus test critical phrases. Privacy compliance needs disciplined processes. Consent flows must be clear and logged. Data exports should follow strict controls. De-identification is not optional. Vendor lock-in can happen. Favor open formats and documented APIs. Keep your taxonomy portable. Avoid custom fields that cannot migrate. ## Cashew's roadmap and what it signals Cashew plans a seed round in early 2026 to advance its technology. The team will expand U.S. presence and deepen B2B reach. A proprietary database should strengthen with volume. This is the compounding effect at work. Moroccan buyers can watch for three things. Evidence of multilingual performance. Transparent quality metrics and audits. Clear governance across data collection, storage, and reporting. If those boxes are checked, adoption can scale. Teams can run more tests, more often. Decisions can move faster with less risk. That is the real win. ## Bottom line for Morocco Fresh human data plus AI synthesis is a practical path. It speeds insights without reducing them to web summaries. It also keeps humans accountable. Morocco's mix of languages and markets makes this model relevant. CPG, telecom, tourism, and public services can all benefit. SMBs gain access to custom research. Larger enterprises gain throughput and learning loops. Execution will decide outcomes. Invest in localization, sampling rigor, and governance. Build your feedback muscle. Then scale with confidence.