---
title: "Quantitative Research Agency for Consumer and Tech Brands | antedote"
description: "Quantitative research for consumer and tech brands: segmentation, conjoint and choice modeling, MaxDiff and TURF, and pricing research you can act on."
url: "https://antedote.com/what-we-do/quantitative-research/"
---

# Quantitative research that shows who to put first, and what they’d choose.

We size the groups in your market, show which option people would choose and at what price, and say how far each answer can be trusted.

## Which method answers your question.

- ### Which customers should we put first?

    [Customer segmentation](https://antedote.com/what-we-do/customer-segmentation/)

    We group your market by need and behavior, size each group, and find it again in later surveys, or in your customer records where those have the right variables.

- ### Which features, range and price would people choose?

    [Conjoint and choice modeling](https://antedote.com/what-we-do/conjoint-and-choice-modeling/)

    People choose between whole products, screen after screen. One choice doesn’t say why; a model of many estimates what each feature and each step in price is worth.

- ### Which ideas on a long list matter most?

    [MaxDiff and TURF](https://antedote.com/what-we-do/conjoint-and-choice-modeling/#maxdiff-and-turf)

    Picking the most and least important from small sets, again and again, sorts a long list. TURF then finds the set that would reach the most people.

- ### What should we charge?

    [Pricing research](https://antedote.com/what-we-do/pricing-research/)

    When the decision involves features, tiers or competing offers, a choice model can help. For other price questions, a direct method or a live test may be more useful.

When the question is why people choose as they do, start with [qualitative research](https://antedote.com/what-we-do/qualitative-research/). When your own data can size the opportunity, add [data science and analytics](https://antedote.com/what-we-do/data-science/).

## Which differences matter?

We read the size of a difference, the uncertainty around it and the consequences of acting on it. A significance marker helps interpret the evidence; it doesn’t make the decision for you.

- A letter: the gap meets the test’s threshold under its assumptions. Whether it matters commercially is a separate question.

- No letter: the gap doesn’t meet that threshold. Read the estimate, its uncertainty and the other evidence before deciding what to do.

- Too few people: the estimate is imprecise, however big the gap looks. We make that uncertainty visible.

*An illustration, not a client project.*

A drawing: one page of a table printout comparing five groups of people, A to E, on five measures, with a base row showing how many people stand behind each group. Printed letters mark the differences that meet the test’s threshold, and a pen rings one of them. A gap with no letter is underlined in pen with a question mark. Group E, with very few people behind it, is boxed and hatched out in pen, although its bars are the longest on the page.

## How far an answer can go.

Quant measures what people choose under the conditions studied. It doesn’t predict exactly what people will do, so each answer comes with its limits.

### A fair comparison, read for what it is.

[Choosing which home-heating customers to prioritize](https://antedote.com/explore-our-work/home-heating-customer-segmentation/)

For a home services company in Europe, we built and sized a segmentation, then tested versions of its existing service proposition. Each person saw one version, allocated at random, and interest was read as conditional on buying a new heating system.

The segments went into use: the client’s media agency planned around the priority segment, and the company read its next proposition test through them.

### Enough people in every group you’ll read.

[Choosing which personal care buyers to target in each market](https://antedote.com/explore-our-work/personal-care-buyers-in-four-markets/)

For a personal care company, we surveyed category buyers in four European markets, at a scale chosen so that smaller groups could still be read within every market.

The central and local teams came away with one segmentation they could all use.

### Evidence two teams agreed to trust.

[Deciding what goes into a consumer technology roadmap](https://antedote.com/explore-our-work/conjoint-and-simulation-for-a-product-roadmap/)

For a consumer technology company, engineering and marketing agreed before fieldwork began what evidence they would trust. A conjoint model and simulator then tested portfolio options, every output a preference modeled under the conditions tested, and never a market forecast.

Engineering and marketing had one shared set of evidence to plan the roadmap from.

[EXPLOR Award winner](https://www.quirks.com/articles/8-top-segmentation-research-companies), Insights Association, for a segmentation and market model, voted by the Corporate Researchers Conference.

## How many, which one, and how sure?

Bring the decision and what you already know. We’ll suggest the method that would answer it, and say how sure the answer could be.

[Talk to us](https://antedote.com/talk-to-us/) · [hello@antedote.com](mailto:hello@antedote.com)

A partner or director will reply.