Method

We start with the problem

Most research companies start with the method: should we do a quantitative or qualitative study? We start with the business question. What do you need to know to make a better decision? The answer determines which data sources and methods we use. This means we sometimes recommend not conducting a survey at all — because the answer already exists in your own data.

Data first, survey second

Reflect uses observed data as the starting point: sales data, transaction data, pricing data, GIS data and customer registries. These data sources show what is actually happening in the market — not what respondents think is happening. Surveys fill gaps that observed data cannot cover: attitudes, drivers, future intentions and price acceptance. But the survey is the last step, not the first.

• Sales and transaction data reveal actual behavior
• Pricing data uncovers market structure and thresholds
• GIS data provides geographic context and market geometry
• Surveys measure what cannot be observed

Calibration against reality

A model that has not been tested against actual behavior is a guess with mathematics. We calibrate our models against observed data: does the price acceptance in the study match actual sales at different price levels? Does the segmentation match how customers actually behave? This is a fundamental methodological requirement that most research companies skip — because they lack access to behavioral data, or do not know how to connect them.

Simulation and modeling

We do not stop at describing the current state. Our analyses include scenario modeling: what happens if you raise the price by 10%? What happens if you launch a new product variant? How is your market share affected if a competitor changes their assortment? Simulation requires models calibrated against reality — not just curve-fitting to survey data. That is why we invest in building our own models rather than using off-the-shelf tools.

• Price simulation with demand elasticity
• Assortment optimization beyond reach
• Brand scenarios with business impact
• Volume prediction based on calibrated models

Decision support, not just reports

A report describing what happened is not decision support. Decision support means delivering a framework that makes it possible to weigh alternatives against each other — with clear consequences. Our deliverables always include a recommendation with rationale, a description of uncertainty and a structure for making the decision. We never leave the client with a report and a 'good luck'.

Our own analytics platform

Reflect has built its own analytics platform for price simulation, assortment optimization (TURF and volume models), cross-tabulation and decision support. The platform enables rapid iteration and real-time scenario testing — something that is not possible with standard tools. The platform is not a product we sell — it is a tool we use to deliver better analysis faster. Clients working with us on a recurring basis get access to interactive reports and dashboards directly in the platform.