From Product Data to Product Decisions
Using behavioral data, AI-assisted analysis, and in-product guidance
to inform product strategy and adoption.
I established a more intentional analytics practice around MAX Completions so product and business leaders could better understand adoption, usage trends, emerging pain points, and opportunities for improvement.
By combining behavioral data from Pendo and Datadog/RUM with AI-assisted analysis and recurring reporting, I helped turn product usage into actionable insights for UX priorities, product strategy, and broader business decisions. I also used Pendo as an in-product adoption channel, building a Knowledge Center and targeted guidance for onboarding, support, release communication, and new-feature discovery.
Product Analytics | UX Strategy | AI-Assisted Analysis | Leadership Reporting | Product Adoption
Feature Impact Summary example
From Signal to Decision
Connecting behavioral data, analysis, leadership discussion, and action.
I connected analytics, reporting, product discussions, in-product guidance, and UX improvements into a broader decision-making workflow.
A behavioral signal could identify an area worth investigating. AI-assisted analysis helped me compare larger sets of usage data over time and surface patterns or anomalies worth deeper attention.
I then brought those findings into weekly leadership workshops, where they became inputs into product priorities, UX improvements, adoption efforts, and broader business decisions.
The goal wasn’t simply to understand what had happened. It was to turn product behavior into something the organization could act on.
DATA → ANALYSIS → INSIGHT → DECISION → ACTION → MEASUREMENT
Seeing What Users Actually Do
Using behavioral data to identify patterns worth investigating.
I used Pendo and Datadog/RUM to monitor product behavior across areas such as visitors, sessions, feature usage, engagement, and interaction patterns.
This helped me investigate questions such as:
Which parts of the product were gaining or losing engagement?
Which features were being discovered or overlooked?
Where did behavior suggest possible friction?
What changed after a product update or in-product guide?
Analytics could show what was happening. Understanding why often required customer feedback, support input, stakeholder context, and knowledge of the product.
Using AI to Find Patterns and Trends
Accelerating analysis without replacing product judgment.
I used AI as an analysis partner to help synthesize usage data, compare time periods, identify trends, and surface questions worth deeper investigation.
This was particularly useful when working across larger sets of product metrics and looking for changes in engagement, feature adoption, or behavior that weren’t immediately obvious in an individual dashboard or report.
I treated AI-generated observations as analytical starting points rather than conclusions. Findings were checked against the underlying data and interpreted in the context of the product, users, and business before being shared or acted on.
Turning Data Into Shared Understanding
Giving leadership a clearer view of product behavior.
I created recurring reports that translated behavioral data and analysis into a more consistent view of how MAX Completions was being used.
I shared these findings during weekly leadership workshops, where they helped create a common evidence base for discussions about product priorities, customer behavior, feature investment, and business direction.
The reports were not simply status updates. They helped connect what was happening in the product to decisions about what deserved attention next.
From Insight to Product Strategy
The findings helped guide where to focus design and product effort, which workflows needed refinement, and where users needed more guidance.
Not every signal pointed to a redesign. Some informed onboarding, feature discovery, roadmap discussions, or deeper investigation before deciding what action made sense.
Using evidence to guide where the team focused next.
SIGNAL ↗︎
Growing dashboard adoption
Wireline Job Dashboard views increased 522%, from 9 to 56 views in seven days.
What it suggested
Usage was expanding in a workflow that had previously seen relatively little activity.
Product response
Use the increase as a signal to understand which users were adopting the workflow, what was driving the change, and whether additional onboarding or product investment was warranted.
SIGNAL ↗︎
Increased use of configuration tools
Thresholds and Notifications configuration increased from 2 to 24 clicks, while Time-Series controls increased from 2 to 55 clicks.
What it suggested
Users were engaging more deeply with configuration and analysis capabilities.
Product response
Give these workflows greater attention in usability reviews, guidance, and future prioritization rather than treating them as secondary functionality.
SIGNAL ↘︎
Declining administrative activity
Client Users page views fell 56%, from 50 to 22 views.
What it suggested
A meaningful behavior change had occurred in an administrative workflow.
Product response
iInvestigate whether the decline reflected reduced need, changes in account-management behavior, another workflow replacing the task, or a discoverability issue.
SIGNAL ↘︎
Declining feature engagement
3D Wellbore toggle usage decreased 59%, from 17 to 7 clicks.
What it suggested
The decline warranted investigation, but did not by itself indicate that the feature had become less valuable.
Product response
Look for supporting evidence around discoverability, workflow changes, user need, and feature quality before deciding whether the experience required redesign, guidance, or reduced investment.
Turning Insight Into Action
Using targeted changes instead of defaulting to larger redesigns.
When evidence pointed to a discovery, onboarding, or comprehension problem, I used in-app guidance as a targeted response.
New-feature announcements, contextual walkthroughs, and onboarding experiences helped address specific points in the user journey without requiring a larger product change.
Because these experiences lived inside the product, I could also measure engagement and downstream behavior to understand what happened after launch.
Examples of in-app guides and feature education
Supporting Ongoing Adoption
Giving users guidance beyond the initial product experience.
I also built a more persistent in-app support layer through the Knowledge Center, bringing onboarding, tutorials, release notes, feature updates, and support resources into one place.
This gave users a place to discover new features, access guidance, and self-serve help without leaving the MAX Completions application, while giving the product team another channel for supporting adoption after launch.
Together, contextual guides and ongoing resources made product education part of the experience rather than something users had to seek out elsewhere.
Examples of the Knowledge Center
Outcomes
Measuring behavior while keeping attribution honest.
Across this work, internal reporting documented changes in activation, feature adoption, engagement, onboarding behavior, and product usage.
The broader impact extended beyond individual metrics. The analytics practice created a more consistent way to bring behavioral evidence into product and leadership discussions, helping teams identify opportunities that might otherwise have relied primarily on anecdotal feedback or assumptions.
I treated these metrics carefully. Changes in behavior showed that something moved, but did not automatically prove that a single design or guidance change caused it. Where multiple product or business changes occurred at the same time, I avoided making causal claims without supporting analysis.
Data becomes valuable when it helps people make better decisions.
What I Took Away
This work changed how I approached product problems. I became more deliberate about starting with what users were actually doing, using AI to surface patterns worth investigating, and bringing those findings into conversations where product and business decisions were being made.
The broader lesson was that analytics are most useful when they create a shared understanding of what is happening, raise better questions, and help teams decide where to act next.
Good analytics work isn’t about producing more dashboards.
It’s about turning data into better-informed decisions.