Finding Signal Without Infrastructure
The growth team had no model of whom they were building for. I built one from three data sources that had never been used together and found that personalization loses precision when it treats all users as if they have the same relationship to risk and trust.
TL;DR
Overview
When I joined Kikoff, the growth team was optimizing without any model of who they were optimizing for. That is a strategic risk at any company — but at a high-growth fintech building credit and financial wellness products for consumers largely underserved by traditional financial infrastructure, it meant features were shipping without a behavioral foundation and retention strategy had no consumer model to build from. There was no research infrastructure and no user segmentation. The gap was structural, not incidental.
I identified that gap and made the case for user segmentation research before the personalization work moved further. This wasn't on the roadmap. I pitched it, framed it as a retention risk, and ran it.
Research Approach
Because there was no established research infrastructure and the roadmap was already in motion, data collection was naturalistic across three channels rather than structured by a fixed protocol.
- Live call observation: I listened to customer calls in real time without a fixed observation framework. The organizational conditions didn't allow for structured interference, so the approach was unstructured but purposive: designed to surface unfiltered sentiment and behavioral context as it emerged.
- Analytics data: Direct pulls from Amplitude to map where users were going, where they dropped, and what the journey indicated about unmet needs.
- Marketing insight synthesis: Examining acquisition channels to assess whether incoming messaging reflected the consumers the product was actually serving.
This was convergent data collection: not accuracy-checking across sources, but layering them to build depth and surface the human context that no single channel could provide alone.
Findings
Three distinct behavioral patterns emerged across the data, each reflecting a different relationship with the product.
- Non-Returners: Signed up but never came back. Trust was not established at the onboarding moment, and the product gave them no reason to return.
- Occasional Checkers: Returned sporadically. Passive interest existed, but without enough clarity or motivation to sustain engagement.
- Active Builders: Consistently engaged, using Kikoff as an active tool toward a specific financial goal.
What live call observation added beneath the behavioral data was a cross-segment finding: every group, regardless of engagement level, was navigating some form of financial stress.
Framework
Financial stress reframed the strategic question. The research had started as a segmentation problem. Financial stress doesn't just shape behavior, it shapes how much risk a person is willing to take with something unfamiliar. That made it a trust problem, not an engagement problem. The question was no longer how to increase engagement in the abstract. It shifted to how personalization efforts might account for the emotional conditions underneath the behavioral patterns. Without that context, optimization could have been targeting the wrong variable.
The segmentation framework changed how product and design approached personalization strategy. Before it existed, decisions were driven largely by competitive observation: borrowing patterns from other platforms rather than building from an understanding of Kikoff's own consumers. The organization was looking outward when the behavioral data was internal.
The framework gave product a foundation for personalization targeting built on behavioral understanding. It gave design consumer context for onboarding grounded in actual user goals rather than industry convention. Two teams now had a shared model for who they were building for.
Impact
Within four months of implementation, retention improved 12% against baseline. The segmentation framework gave the personalization strategy a behavioral target it had not previously had.
A Parting Perspective
Personalization is less a technical challenge than a behavioral one. It is about probability and context: understanding the conditions under which a consumer is most likely to engage, trust, and return. Without that, it is a feature with good intentions and no foundation. What this research kept returning to wasn't behavioral segments. It was the conditions underneath them. Every group was navigating financial stress, and that stress was shaping how much trust users were willing to extend to a product they had just met. Recognizing that required a different kind of looking.
“Expanding perspectives into deliberate insight for bolder moves”
Made and designed by nayeri
© 2026 All Rights Reserved
Finding Signal Without Infrastructure
The growth team had no model of whom they were building for. I built one from three data sources that had never been used together and found that personalization loses precision when it treats all users as if they have the same relationship to risk and trust.
TL;DR
Overview
When I joined Kikoff, the growth team was optimizing without any model of who they were optimizing for. That is a strategic risk at any company — but at a high-growth fintech building credit and financial wellness products for consumers largely underserved by traditional financial infrastructure, it meant features were shipping without a behavioral foundation and retention strategy had no consumer model to build from. There was no research infrastructure and no user segmentation. The gap was structural, not incidental.
I identified that gap and made the case for user segmentation research before the personalization work moved further. This wasn't on the roadmap. I pitched it, framed it as a retention risk, and ran it.
Research Approach
Because there was no established research infrastructure and the roadmap was already in motion, data collection was naturalistic across three channels rather than structured by a fixed protocol.
- Live call observation: I listened to customer calls in real time without a fixed observation framework. The organizational conditions didn't allow for structured interference, so the approach was unstructured but purposive: designed to surface unfiltered sentiment and behavioral context as it emerged.
- Analytics data: Direct pulls from Amplitude to map where users were going, where they dropped, and what the journey indicated about unmet needs.
- Marketing insight synthesis: Examining acquisition channels to assess whether incoming messaging reflected the consumers the product was actually serving.
This was convergent data collection: not accuracy-checking across sources, but layering them to build depth and surface the human context that no single channel could provide alone.
Findings
Three distinct behavioral patterns emerged across the data, each reflecting a different relationship with the product.
- Non-Returners: Signed up but never came back. Trust was not established at the onboarding moment, and the product gave them no reason to return.
- Occasional Checkers: Returned sporadically. Passive interest existed, but without enough clarity or motivation to sustain engagement.
- Active Builders: Consistently engaged, using Kikoff as an active tool toward a specific financial goal.
What live call observation added beneath the behavioral data was a cross-segment finding: every group, regardless of engagement level, was navigating some form of financial stress.
Framework
Financial stress reframed the strategic question. The research had started as a segmentation problem. Financial stress doesn't just shape behavior, it shapes how much risk a person is willing to take with something unfamiliar. That made it a trust problem, not an engagement problem. The question was no longer how to increase engagement in the abstract. It shifted to how personalization efforts might account for the emotional conditions underneath the behavioral patterns. Without that context, optimization could have been targeting the wrong variable.
The segmentation framework changed how product and design approached personalization strategy. Before it existed, decisions were driven largely by competitive observation: borrowing patterns from other platforms rather than building from an understanding of Kikoff's own consumers. The organization was looking outward when the behavioral data was internal.
The framework gave product a foundation for personalization targeting built on behavioral understanding. It gave design consumer context for onboarding grounded in actual user goals rather than industry convention. Two teams now had a shared model for who they were building for.
Impact
Within four months of implementation, retention improved 12% against baseline. The segmentation framework gave the personalization strategy a behavioral target it had not previously had.
A Parting Perspective
Personalization is less a technical challenge than a behavioral one. It is about probability and context: understanding the conditions under which a consumer is most likely to engage, trust, and return. Without that, it is a feature with good intentions and no foundation. What this research kept returning to wasn't behavioral segments. It was the conditions underneath them. Every group was navigating financial stress, and that stress was shaping how much trust users were willing to extend to a product they had just met. Recognizing that required a different kind of looking.
What would you like to Read next?
Mapping the Spectrum of Engagement
Integrating the Monetization Funnel
“Expanding perspectives into deliberate insight for bolder moves”
Made and designed by nayeri
© 2026 All Rights Reserved
Finding Signal Without Infrastructure
The growth team had no model of whom they were building for. I built one from three data sources that had never been used together and found that personalization loses precision when it treats all users as if they have the same relationship to risk and trust.
TL;DR
Overview
When I joined Kikoff, the growth team was optimizing without any model of who they were optimizing for. That is a strategic risk at any company — but at a high-growth fintech building credit and financial wellness products for consumers largely underserved by traditional financial infrastructure, it meant features were shipping without a behavioral foundation and retention strategy had no consumer model to build from. There was no research infrastructure and no user segmentation. The gap was structural, not incidental.
I identified that gap and made the case for user segmentation research before the personalization work moved further. This wasn't on the roadmap. I pitched it, framed it as a retention risk, and ran it.
Research Approach
Because there was no established research infrastructure and the roadmap was already in motion, data collection was naturalistic across three channels rather than structured by a fixed protocol.
- Live call observation: I listened to customer calls in real time without a fixed observation framework. The organizational conditions didn't allow for structured interference, so the approach was unstructured but purposive: designed to surface unfiltered sentiment and behavioral context as it emerged.
- Analytics data: Direct pulls from Amplitude to map where users were going, where they dropped, and what the journey indicated about unmet needs.
- Marketing insight synthesis: Examining acquisition channels to assess whether incoming messaging reflected the consumers the product was actually serving.
This was convergent data collection: not accuracy-checking across sources, but layering them to build depth and surface the human context that no single channel could provide alone.
Findings
Three distinct behavioral patterns emerged across the data, each reflecting a different relationship with the product.
- Non-Returners: Signed up but never came back. Trust was not established at the onboarding moment, and the product gave them no reason to return.
- Occasional Checkers: Returned sporadically. Passive interest existed, but without enough clarity or motivation to sustain engagement.
- Active Builders: Consistently engaged, using Kikoff as an active tool toward a specific financial goal.
What live call observation added beneath the behavioral data was a cross-segment finding: every group, regardless of engagement level, was navigating some form of financial stress.
Framework
Financial stress reframed the strategic question. The research had started as a segmentation problem. Financial stress doesn't just shape behavior, it shapes how much risk a person is willing to take with something unfamiliar. That made it a trust problem, not an engagement problem. The question was no longer how to increase engagement in the abstract. It shifted to how personalization efforts might account for the emotional conditions underneath the behavioral patterns. Without that context, optimization could have been targeting the wrong variable.
The segmentation framework changed how product and design approached personalization strategy. Before it existed, decisions were driven largely by competitive observation: borrowing patterns from other platforms rather than building from an understanding of Kikoff's own consumers. The organization was looking outward when the behavioral data was internal.
The framework gave product a foundation for personalization targeting built on behavioral understanding. It gave design consumer context for onboarding grounded in actual user goals rather than industry convention. Two teams now had a shared model for who they were building for.
Impact
Within four months of implementation, retention improved 12% against baseline. The segmentation framework gave the personalization strategy a behavioral target it had not previously had.
A Parting Perspective
Personalization is less a technical challenge than a behavioral one. It is about probability and context: understanding the conditions under which a consumer is most likely to engage, trust, and return. Without that, it is a feature with good intentions and no foundation. What this research kept returning to wasn't behavioral segments. It was the conditions underneath them. Every group was navigating financial stress, and that stress was shaping how much trust users were willing to extend to a product they had just met. Recognizing that required a different kind of looking.
“Expanding perspectives into deliberate insight for bolder moves”
Made and designed by nayeri
© 2026 All Rights Reserved