For Stalled Growth Curves

Five behaviorally grounded growth experiments,
built from data you already have,
ready to run in two weeks.

Growth leaders are rarely short on data or customer insight. The problem is turning what you already know into new, credible action fast enough to matter. Upstream reviews the research and data you already have, filters it for what actually predicts behavior change, and hands you five ready-to-test growth experiments in two weeks.

Scoped to the outcomes you are accountable for:
Customer Growth

Tested against what drives relevance, and where it drops off.

Deeper Engagement

Tested against what reinforces expected benefits once someone is in the product.

Habitual Use

Tested against what drives repeat engagement instead of one-off use.

Resource Allocation

Scoped to your team's real constraints, so you control executional risk.

The Real Problem

Certainty got expensive.
Waiting is now the risk.

The old model for turning research into growth was built for a world where taking action was expensive: development hours, resource deployment, real implementation risk. Waiting for certainty was the responsible choice. That math has flipped. When testing something live costs almost nothing, certainty becomes the expensive tradeoff and speed becomes the safe one.

Most growth teams already operate this way. What hasn't caught up is how they coordinate action from research: a process still calibrated to the old certainty bar can't keep pace with partners who no longer wait for one.

Pain 01
Another workshop, another list nobody prioritizes
Unstructured brainstorming produces plausible ideas with no behavioral mechanism underneath, so results are inconsistent and hard to learn from.
Pain 02
The team knows what's wrong, no one agrees what to test first
The data and research already exist. What's missing is a fast, disciplined way to turn them into a prioritized set of bets.
Pain 03
Every option needs a research sprint before anyone will greenlight it
A process built for certainty adds months to a decision that speed, not more analysis, is now built to make.
How It Works

From existing insight
to five testable bets

No new primary research required. Upstream reviews the customer and market data you already have, filters it for the elements that actually carry behavior-change logic, and scopes what's left into experiments sized to what your team controls, a single function or a coordinated set.

Diagnose
Review the data and research you already have for what actually predicts behavior change.
Design
Translate what's found into five candidate experiments, scoped to what your team controls.
Deliver
One synthesis, five ready-to-run experiments, two weeks door to door.
Decide
Run what you can right away. Whatever needs more research becomes the next scoped effort.
Proof, by Challenge Type

Three problems.
Three ways this method held up.

Product Adoption & Habitual Use
Bring continuous glucose monitoring to the phone, without the stigma
Dexcom

Upstream reviewed existing user and clinical data to design the G6 app, balancing the functional need to keep blood sugar in range against the social need to avoid stigma and unwanted attention.

OutputApp design shipped inside the FDA-cleared G6 system
First FDA-approved continuous glucose monitoring app, then a Silver 2019 Medical Design Excellence Award winner.
Retention & Churn Reduction
Keep a patient's care team, and trust, connected across a 20-state system
Ascension

Patient and caregiver journey mapping across the health system's facilities showed where continuity, and trust, breaks down between departments, informing a Command Center model built to hold it together at scale.

OutputCommand Center operating model for coordinating care
Scaled beyond the original site to a Command Center model now running across 20 states.
Outcome Alignment & KPI Improvement
Trace a support and sentiment problem back to its emotional root cause
Meta / Facebook

Ethnographic research and journey mapping across four continents, then quantitative segmentation, traced why offensive content was hurting engagement and brand sentiment while overloading the support team.

OutputEmotional-stages framework, translated into a cross-functional design sprint
Concepts covering in-product information, tone, and user control shipped platform-wide after segment testing.

"Upstream's process opened our minds to new ways of addressing our users' needs and approaches for our products to cross from early adopters to the majority."

Dexcom
Director of Research & Development

Also Trusted By Cash App, Pinterest, Rackspace, Tribune Publishing, City of Austin, and the Gates Foundation
Why Upstream

The same method,
applied for years.

This is the same underlying model Upstream has applied for organizations across scale and category for years, not a new offering built speculatively. An interdisciplinary team brings 20+ years of combined industry and consulting experience, with recent work for some of the largest organizations globally.

One measure of whether a method holds up: the Alzheimer's Association has used Upstream's Path to Diagnosis framework for nearly eight years to align leadership on the behavioral journey of people living with the disease and their caregivers.

Recognition
Fast Company, 2021 World Changing Ideas
Recognition
2019 Medical Design Excellence Award, Silver
Start Here

More effective experiments
might be closer than you think.

As you churn through ideas, the solution might not be more brainstorming or opening up for more and more perspectives. That can lead to the opposite result: less impact, more noise, and more confusion even when things work out. Upstream finds teams get the most out of their institutional knowledge with better experimental design, rooted in proven decision-making logic.

There's rarely a bad time to find out which five experiments are worth running. The two-week version just means finding out sooner.