Startup Spotlight #18: A Product That Users Want
To create a truly useful product, you need to know your users. Today's startup helps you make sense of user feedback, using a little AI magic.
Project Overview
Lancey is a platform for product developers that leverages AI to help them “build products that users want”.
They do it through a process called “micro-experiments”. But what is it, exactly? A micro-experiment targets a small group of product users, selected not just by characteristics but also by prior behavior.
This sets micro-experiments apart from traditional product experiments, which typically focus on all users or large segments, such as subscribers of a specific pricing plan.
Why are micro-experiments necessary? Because every user is different — they behave differently and use products in varied ways. Consequently, there’s no universal solution to simultaneously enhance activity, engagement, or retention across all users.
On the other hand, manually analyzing user behavior, segmenting them into small groups, running experiments for each group, and summarizing results is highly cumbersome. Lancey streamlines this process.
To start, all user behavior data sources—both within and around the product—are integrated into the Lancey platform.
This data can come from Segment, Snowflake, MailChimp, payment systems, and various other sources. Lancey already has pre-built integration modules for many popular platforms.
Next, you define the goal of the experiments — whether it’s increasing product usage frequency, engagement, retention, or another objective.
Lancey’s AI then identifies the user characteristics and behavior patterns that most impact the chosen metric.
It segments users into groups based on these characteristics and patterns.
Finally, the AI generates a list of experiments for a small sample within each group. These tests measure their impact on the target metric by comparing results with a control group of unaffected users from the same segment.
To conduct an experiment, the core idea must be programmed as a procedure—e.g., displaying a banner with a specific call-to-action. Lancey executes the experiment for the selected user samples.
Over time, these experiments build a library of procedures for “nudging” users. This enables Lancey to operate in autopilot mode.
In this mode, the platform’s AI regularly analyzes user behavior, creates groups, designs and runs experiments using the existing procedure library, evaluates results, and scales successful procedures across user groups to achieve set goals.
Lancey participated in Y Combinator in the summer of 2022, securing $500,000 in initial funding.
What’s the Gist?
Most businesses still operate on a “one-size-fits-all” approach — not out of ill intent, but due to constraints in resources, time, patience, and budget.
The same is true for digital products. Analyzing individual user behavior, understanding what drives it, and how it impacts product outcomes is tedious. Product teams lack the resources to handle this complexity, so they make broad decisions aimed at a generalized user base, hoping for overall success.
But this approach overlooks potential gains from the multitude of unique users who don’t fit the generalized mold. The broader a product’s functionality, the longer this tail becomes — different users use different features for different purposes, making personalization even harder.
Lancey isn’t the only startup doing this. Subsets, a startup that reduces subscriber churn for digital publications by analyzing reader behavior to uncover different usage patterns and tailoring actions for each group, can be described as similar.
Today’s Lancey conceptually resembles Subsets but supports a wider variety of products beyond digital media.
Another example is OfferFit. It customizes offers for shoppers based on individual preferences, like purchasing habits, responsiveness to promotions, and optimal communication times.
Lancey, Subsets, and OfferFit share a common theme: “scalable personalization” — tailoring experiences to individual users without excessive manual effort, using AI-driven automation.
This concept isn’t limited to subscriptions or retail. Studio, a startup offering online courses, uses AI to personalize lessons and support for students, ensuring each achieves their goal of producing one quality song per month.
AI-powered scalable personalization enables businesses to adjust to diverse user needs at scale, saving resources and enhancing outcomes.
Key Takeaways
Scalable personalization seems to be the way of the future. The challenge faced by product creators is clear, and the AI-driven solution to this problem has only recently become available.
The first approach is to use AI for meticulous analysis and customization of your product to suit different user types, improving activation, engagement, and retention metrics.
The second approach is to create platforms that enable other product developers to do the same.
Even the startups mentioned today demonstrate that this concept can be applied across various industries in multiple ways.
Subscription services, retail, and education are already obvious candidates. However, there’s still room to focus on these sectors by creating alternatives to existing platforms. These markets are large enough to accommodate plenty of players.
But here’s a thought — what other markets and purposes could benefit from scalable personalization? Where else do users, clients, or customers exhibit diverse behavior patterns while using the same products? And how can businesses adapt to these patterns to boost revenue for product creators?
Company info
Lancey
Website: https://www.lancey.ai/
Last funding round: $500k, 01.08.2022
Total funds raised: $500k over 1 round