Introducing august

Welcome to august, a human prediction platform powered by Digital Twins of the people your business needs to understand. We build Digital Twins of your users, customers, visitors, buyers and prospects and allow you to run focus groups, surveys or one-on-one interviews with the Twins. Each Twin corresponds to a specific real individual. It is not a generic persona and is designed to predict the actions and behaviors of that person. august supports providing screenshots, Figma mock links and public URLs as part of focus groups, surveys or one-on-one interviews so you can directly leverage the Twins to predict reactions to potential new features and designs. Our Figma Plugin, Slack app and MCP also allow you to leverage august for user feedback directly where you're already working.

Why we built this

FERMÀT is primarily an ecommerce company, and as a result we were exposed to two things:

1. A large amount of commerce data
2. Novel agentic usages of commerce data to create "synthetic shoppers" or "synthetic customers"

Most of the "synthetic shopper" approaches we saw took the approach of training the shoppers on aggregate category-level data across generic traffic. This resulted in very high-level personas that were interesting ("shopper that browses quickly and churns", "shopper that cares a lot about reviews", "shopper that heavily researches") but too broad to accurately predict behavior or deliver feedback at a high enough level of fidelity to be useful.

We started august to take the opposite approach: instead of aggregate personas, what if we used semantic and behavioral data to train Digital Twins for each individual customer? We could then allow you to converse directly with a Twin or roll up a segment of Twins into a focus group or survey while still preserving the fidelity of the real individual's behavior and actions. This model directly replicates traditional user research, but with the ability to conduct the research at a much higher velocity and with less friction.

We saw initial traction from product teams within commerce brands and retailers we work with. Product Managers, Product Engineers and UI/UX Designers would use august to get quick, direct user feedback on what they were working on:

• PMs used august to validate and improve PRDs directly in Claude
• Designers used august for quick actionable feedback in Figma
• UX researchers used august to pressure-test ideas

We quickly realized that any product team in any org could get value from august and decided to offer august as a standalone product that can be leveraged by anyone with users, customers, visitors, buyers, prospects or any group of people they want to better understand.

Why this matters

Without august, conducting user research or soliciting user feedback involves one of two options:

1. Conduct traditional manual research on your own user or customer base
2. Use a platform like Kantar or UserTesting to understand the opinion of generic individuals outside of your customer base

AI is fundamentally changing how product teams operate and compressing cycles and delivery timelines. Conducting traditional research on your own user or customer base is now too time intensive to be done regularly, leading modern product teams to often forgo what is valuable feedback. Platforms like Kantar or UserTesting solve this time problem, but are expensive and do not offer predicted actions and feedback from individuals in your own user or customer base.

august directly solves both of these problems. Quickly solicit feedback or gauge expected responses where you already work, on Digital Twins trained directly on your own data.

What's next

The team behind august is shipping rapidly. Here's what's coming up:

1. PostHog integration to allow you to quickly train Twins on your existing data
2. Segments that allow you to describe a cohort ("Subscribers in the US who cancel after the second order", "Users who used our new feature once but have not come back") and use directly in focus groups and surveys
3. Guidance directly in the product on how to best get feedback from your Digital Twins

and much more!

The team will regularly be publishing posts on this blog about what we're working on. Coming up next will be a post that dives into our training pipeline and how a Digital Twin is constructed.