--- title: Causal market research that reveals why customers act | Gradient description: Gradient uses advanced models to uncover why customers act the way they do, so you can move from observation to action with confidence. --- ![bg_default_8](https://www.gradientmetrics.com/hs-fs/hubfs/Brand_Current/default/backgrounds/bg_default_8.png?width=300&name=bg_default_8.png) # We look for the *why*, not just the *what*Reveal your audience's hidden preferences We use advanced statistical models to go beyond surface trends, uncovering the real drivers of behavior so you can make smarter, more effective decisions. [See Our Work](https://www.gradientmetrics.com/case-studies) [Contact Us](https://www.gradientmetrics.com/contact) ## You don’t just want to know what happened. You want to know what to do next. Anyone can report that brand preference dropped 5 points. We want to know *why*. And what you can *do* about it. That’s where Gradient shines: we specialize in statistically-powered research that reveals the underlying motivations, tradeoffs, and tensions behind consumer behavior. Our models are designed to cut through noise and find the signal. Not just correlation, but causation. Not just what consumers say, but what actually drives their choices. From observation to explanation Patterns in data are just the start. We use causal modeling to uncover what’s really influencing behavior. Make better predictions Knowing why people make choices helps us predict future trends and actions more accurately. We give you clear next steps Our approach gives you specific guidance on what actions will make a difference. ## How we make it work **![Icons-Gradient-32x32-WhitexColor-08](https://www.gradientmetrics.com/hs-fs/hubfs/Brand_Current/default/icons/Icons-Gradient-32x32-WhitexColor-08.png?width=100&height=76&name=Icons-Gradient-32x32-WhitexColor-08.png)Align on strategy questions via workshop** Collaboratively define key business questions and research objectives through a facilitated workshop, ensuring alignment with overall business goals. **![Icons-Gradient-32x32-WhitexColor-12](https://www.gradientmetrics.com/hs-fs/hubfs/Brand_Current/default/icons/Icons-Gradient-32x32-WhitexColor-12.png?width=100&height=100&name=Icons-Gradient-32x32-WhitexColor-12.png)Design analysis plan & ghost deliverables** Develop a detailed research plan outlining methodologies, and expected outcomes. Create "ghost" versions of final deliverables to guide the research process. **![Icons-Gradient-32x32-WhitexColor-17](https://www.gradientmetrics.com/hs-fs/hubfs/Brand_Current/default/icons/Icons-Gradient-32x32-WhitexColor-17.png?width=100&height=100&name=Icons-Gradient-32x32-WhitexColor-17.png)Write survey ** Design a survey instrument tailored to the research objectives, incorporating best practices in question formulation and survey design. **![Icons-Gradient-32x32-WhitexColor-42](https://www.gradientmetrics.com/hs-fs/hubfs/Brand_Current/default/icons/Icons-Gradient-32x32-WhitexColor-42.png?width=100&height=100&name=Icons-Gradient-32x32-WhitexColor-42.png)Field survey ** Deploy the survey to the target audience, ensuring appropriate sample size and representation to achieve statistically significant results. **![Icons-Gradient-32x32-WhitexColor-33](https://www.gradientmetrics.com/hs-fs/hubfs/Brand_Current/default/icons/Icons-Gradient-32x32-WhitexColor-33.png?width=100&height=94&name=Icons-Gradient-32x32-WhitexColor-33.png)Data Quality Checks** Rigorously clean and validate data to ensure accuracy and reliability, removing any low-quality responses or outliers. **![Icons-Gradient-32x32-WhitexColor-06](https://www.gradientmetrics.com/hs-fs/hubfs/Brand_Current/default/icons/Icons-Gradient-32x32-WhitexColor-06.png?width=100&height=100&name=Icons-Gradient-32x32-WhitexColor-06.png)Develop statistical models** Build and implement custom statistical models to analyze data, employing techniques such as conjoint analysis, regression, and segmentation. **![Icons-Gradient-32x32-WhitexColor-16](https://www.gradientmetrics.com/hs-fs/hubfs/Brand_Current/default/icons/Icons-Gradient-32x32-WhitexColor-16.png?width=100&height=100&name=Icons-Gradient-32x32-WhitexColor-16.png)Analyze and Interpret modeled results** Execute the analysis plan, using statistical models to uncover insights and relationships within the data. **![Icons-Gradient-32x32-WhitexColor-43](https://www.gradientmetrics.com/hs-fs/hubfs/Brand_Current/default/icons/Icons-Gradient-32x32-WhitexColor-43.png?width=100&height=100&name=Icons-Gradient-32x32-WhitexColor-43.png)Synthesize and debrief** Translate analytical findings into clear, actionable insights and strategic recommendations. Present findings through comprehensive reports and debrief sessions. ## The Gradient difference [![ToddRose](https://www.gradientmetrics.com/hs-fs/hubfs/ToddRose.webp?width=300&name=ToddRose.webp) TODD ROSE - POPULACE Getting it right is Todd Rose's top priority. Hear from the Co-Founder and CEO of Populace on how he views Gradient as an extension to his team to make that happen. ](https://www.gradientmetrics.com/our-research-approach/causal-modeling#)[![heidi_hackemer](https://www.gradientmetrics.com/hs-fs/hubfs/Landing%20Page/heidi_hackemer.jpg?width=300&name=heidi_hackemer.jpg) HEIDI HACKEMER - OATLY Hear from the former Executive Director at Oatly on why she keeps bringing in Gradient when she needs to transform a brand from the ground up using quant research. ](https://www.gradientmetrics.com/our-research-approach/causal-modeling#) "Gradient is our favorite partner to work with. They help us craft approaches that answer our clients' most complicated market research questions and are able to adapt when projects shift. The Gradient team are great partners in helping us visualize and explain complex research findings that are easy for our clients to understand and make data-driven decisions." Aprill Huff Vice President, Research & Insights at Media Monks "The Gradient team gave us a deep understanding of our customers through a robust analysis of our database. The insights validated our marketing strategy and enabled investments to continue growing. Their model’s not only accurate but also transparent, giving us confidence in the data. They’re sharp, extremely easy to work with, and fast. We recommend Gradient for any of your analytics needs." GianKarloVP Ecommerce at Orthofeet "Gradient is a great firm to work with. We collaborate with them for all of our analytic work, and even use them to re-run analyses that other firms have done for us because they have earned our complete trust. We think of the Gradient team as an extension of our own—they are knowledgeable, thoughtful, flexible, honest, and an overall pleasure to work with." Parisa RouhaniCo-founder & Sr. Vice President at Populace "The team at Gradient is top notch. They partnered closely with us to understand the intricacies and unique dimensions of our data science problem and they were able to generate some truly fascinating insights from our data." Michael KaminskyFormer Data Science Director at Harry's "Gradient lives at the bleeding edge of market research, data science, and technology. They are not afraid to push the envelope of data collection and analysis methods. It’s a thrill to see the Gradient team roll up their sleeves to craft a sharp analysis for a rapidly approaching press deadline. If you need a no nonsense research partner, Gradient is exactly that." Nate LubinFormer Director of the Office of Digital Strategy at the Obama Administration "I have one word to describe collaborating with Gradient on our latest research project: Refreshing. Not only did their team help us execute the research in ways we weren't thinking about internally, they came to every session we had with enthusiasm, fresh ideas and proactive insights to share with us. That's not something you typically get with 'traditional' market research partners. I couldn't recommend working with them more. A breath of fresh air, and an incredible extension of any internal market research or marketing team." Mike NallanSenior Director Insights & Analytics at Morning Brew "Joining a project in progress can be a challenge, but we learned quickly that we could trust and rely on the Gradient team. That’s a testament to their patience and professionalism." Ali NooraniExecutive Director at the National Immigration Forum "We’re enormously grateful for our partnership with Element Labs on safety and justice message testing. Our findings challenge the conventional political wisdom that when Americans are concerned about crime, their elected leaders must lean into punishment and retribution, with calls for more police, jails, and prisons. The truth is that people across the political spectrum overwhelmingly favor proven solutions that address the root causes of crime, rather than punitive approaches. There’s real power in these insights; we’ve incorporated them into further testing and we’re now working with ACLU affiliates, allies, and civil liberties-aligned candidates to uplift messaging that focuses on preventing crime, strengthening communities, and leading with justice." Ellen FlennikenDeputy Director at the ACLU ## Trusted by leading organizations ![Reformation](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/reformation.png?width=300&name=reformation.png) ![Lippincott](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/lippincott.png?width=300&name=lippincott.png) ![Populace](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/populace.png?width=300&name=populace.png) ![The Just Trust](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/TJT.png?width=300&name=TJT.png) ![Jonathan Adler](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/jonathan-adler.png?width=300&name=jonathan-adler.png) ![Bacardi](https://www.gradientmetrics.com/hubfs/Landing%20Page/bacardi-limited-seeklogo.com.svg) ![Amy Myers MD](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/amymyersmd.png?width=300&name=amymyersmd.png) ![BreadNButter](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/breadnbutter_transparent.png?width=300&name=breadnbutter_transparent.png) ![Brooks](https://www.gradientmetrics.com/hubfs/Client%20Logos/brooks.svg) ![Co:Collective](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/co_collective.gif?width=300&name=co_collective.gif) ![Cowbell](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/Cowbell.png?width=300&name=Cowbell.png) ![Estée Lauder](https://www.gradientmetrics.com/hubfs/Landing%20Page/estee-lauder.svg) ![FitXR](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/fitxr.png?width=300&name=fitxr.png) ![Framebridge](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/framebridge.png?width=300&name=framebridge.png) ![Godiva](https://www.gradientmetrics.com/hubfs/Landing%20Page/godiva.svg) ![Harry's](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/harrys.png?width=300&name=harrys.png) ![Lumin](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/Lumin.png?width=300&name=Lumin.png) ![Noble People](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/noble_people.png?width=300&name=noble_people.png) ![Stitch Fix](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/stitch_fix.png?width=300&name=stitch_fix.png) ![MasterClass](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/master_class.jpg?width=300&name=master_class.jpg) ![Orthofeet](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/orthofeet.jpg?width=300&name=orthofeet.jpg) ![McKinsey&Company](https://www.gradientmetrics.com/hubfs/Landing%20Page/McKinsey__and__Company.svg) ![Nike](https://www.gradientmetrics.com/hubfs/Client%20Logos/nike.svg) ![C-Space](https://www.gradientmetrics.com/hs-fs/hubfs/c_space_logo.png?width=300&name=c_space_logo.png) ![Redscout](https://www.gradientmetrics.com/hubfs/Client%20Logos/logos_redscout-black.svg) ![Sylvain](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/Sylvain.png?width=300&name=Sylvain.png) ![Riverside Company](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/Riverside.png?width=300&name=Riverside.png) ![AEA](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/AEA.png?width=300&name=AEA.png) ![Solomon Partners](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/Solomon_Partners.png?width=300&name=Solomon_Partners.png) ![Child Poverty Action Lab](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/CPAL.png?width=300&name=CPAL.png) ![Consensus Media](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/Consensus.webp?width=300&name=Consensus.webp) ![Squared Circles](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/Squared%20Circles.png?width=300&name=Squared%20Circles.png) ![Morning Brew](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/morning-brew.png?width=300&name=morning-brew.png) ![City Fund Action](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/City%20Fund.png?width=300&name=City%20Fund.png) ![HelloFresh](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/HelloFresh.png?width=300&name=HelloFresh.png) ![MediaMonks](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/M.m_media._m_onks_b.png?width=300&name=M.m_media._m_onks_b.png) ![Oatly](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/oatly.png?width=300&name=oatly.png) ![SYPartners](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/01_SYPartners_Logo_XLarge_reverse_RGB.png?width=300&name=01_SYPartners_Logo_XLarge_reverse_RGB.png) ![National Immigration Forum](https://www.gradientmetrics.com/hs-fs/hubfs/Client%20Logos/NIF.png?width=300&name=NIF.png) ![crowds-people-walking-street-with-ai-generated](https://www.gradientmetrics.com/hs-fs/hubfs/crowds-people-walking-street-with-ai-generated.jpg?width=300&name=crowds-people-walking-street-with-ai-generated.jpg) ## Intuition alone is never enough We go beyond the numbers to answer your biggest questions with confidence. Our advanced modeling uncovers hidden motivations—turning insights into action. [Contact us now](https://www.gradientmetrics.com/contact) ```json { "@context" : "https://schema.org", "@type" : "Organization", "address" : { "@type" : "PostalAddress", "addressCountry" : "", "addressLocality" : "Brooklyn", "addressRegion" : "New York", "postalCode" : "NY 11249", "streetAddress" : "325 Kent Avenue" }, "knowsLanguage" : "en", "logo" : { "@type" : "ImageObject", "url" : "https://www.gradientmetrics.com/hubfs/branding/default/gradient_default_black.png" }, "name" : "Gradient Metrics", "url" : "https://www.gradientmetrics.com/our-research-approach/causal-modeling" } ```