September 2024 ARVIC

Across five sessions, the September 2024 Accelerant Research Virtual Insights Conference returned to a single, persistent tension: insights teams produce more data than ever, yet too much of it sits unused. Speakers from Morning Brew, The Fresh Market, United Health Group, and beyond offered concrete responses to that problem, covering the professional skills, research methods, personalization infrastructure, proprietary community design, and internal influence strategies that separate teams that drive decisions from teams that generate reports.

The Evolving Role of the Insights Professional

Two talks addressed the profession's identity directly, and their conclusions overlapped. Insights professionals who position themselves as data couriers face displacement. Those who interpret findings, connect them to business outcomes, and maintain ongoing relationships with stakeholders in marketing, IT, legal, and product earn a seat at the strategy table.

• AI removes drudgery; human judgment and narrative remain the differentiating output.
• Data literacy does not require a data science degree, but it does require comfort with varied data sets, AI tools, and business-outcome framing.
• Meta-analysis of past projects within a brand or vertical reduces duplicate research and demonstrates cumulative value, a practical first step toward a connected insight stack.
• The profession is shifting toward decision intelligence. Lagging indicators still matter, but stakeholders increasingly expect a view of what is likely to happen next.

Getting Insights Acted Upon

A recurring theme across sessions was that a research readout is not the finish line. Speakers offered tactical and structural approaches to closing the gap between delivery and action.

• Match format and depth to audience maturity. Skeptical stakeholders may need to see a prototype before the research narrative lands.
• Involve stakeholders before the readout through workshops, midpoint debriefs, and interview observation. Surprises at the end reduce ownership of findings.
• Distribute the same finding across multiple channels. Newsletters, lunch-and-learns, and recorded forums each reach a different type of stakeholder.
• Find internal champions who can advocate for research when you are not in the room.
• Organizational memory is short. Even in research-mature organizations, teams must continue demonstrating value to maintain credibility.

Building Proprietary Audience Relationships

Morning Brew's session offered a detailed case for why owned research infrastructure produces better and faster insights than third-party sample. A single newsletter survey returned over 53,000 responses in one day, with 81 percent of respondents expressing interest in further participation. That signal led to the Break Room, a 100,000-member proprietary insights community now spanning 18 brands.

• Progressive profiling stores past survey responses against member profiles, allowing shorter surveys and more precise targeting without losing analytical depth.
• Audience research has supported over $7 million in revenue through custom client projects and category studies.
• Combining behavioral data with survey feedback produced a concrete product change: moving a show's publish time grew it to the number one business news show on Spotify.
• The same finding often matters to multiple internal teams but needs to be framed differently for each. Distribution and shelf life are as important as the finding itself.

Personalization at Scale

The Fresh Market's session grounded personalization in operational reality rather than aspiration. Effective personalization rests on three pillars: quality data collection, AI and machine learning at scale, and a genuine customer-centric mindset. Weakness in any one area limits the others.

• McKinsey data cited in the talk puts personalization ROI at five to eight times marketing spend, with top performers generating 40 percent more revenue than average players.
• Stated preferences do not always match behavior. The Fresh Market's highest-identified vegan and vegetarian customers were most frequently purchasing meat and seafood.
• A/B testing is the practical method for measuring personalization's incremental value: generic messaging as control, personalized messaging as variant.
• Data privacy is a trust issue, not only a compliance requirement. Collect only what customers consent to share and respect stated communication preferences.
• A phased approach, starting with audience understanding and high-impact use cases, builds a foundation that can be scaled using a personalization maturity framework.

Methodological Rigor Across Qualitative and Quantitative Research

One session tackled a persistent professional friction: the perception that qualitative research is less rigorous than quantitative research. The argument was direct. Applying quantitative evaluation standards to qualitative work is a category error rooted in mismatched philosophical paradigms, not a finding about qualitative rigor.

• Qualitative research has its own rigorous evaluation criteria, including reflexivity, triangulation, thick description, transparency, and transferability.
• Saturation signals completeness of understanding of a phenomenon, not generalizability to a population. Confusing the two is a consequential mistake.
• The five-participant usability testing heuristic is grounded in binomial probability and speaks to problem occurrence, not population-level prevalence.
• Mixed methods designs, particularly explanatory and exploratory sequential approaches, leverage the strengths of both paradigms to produce fuller, more actionable insights.
• Researchers have a professional responsibility to evaluate work using the standards appropriate to its paradigm.