OverviewThe artifactPrinciplesHow to read itIn practice
Back to home
Research Shareout

The deliverable, not the deck.

Findings only matter if a room full of people can act on them. This is a real shareout from a Lamps Plus Voice of Customer study — published as a live, navigable proposal rather than forty slides nobody reopens.

Artifact
Interactive proposal
Evidence base
128 surveyed · 7 interviewed
Audience
Merchandising, ecommerce, IT, exec
Format
Live site, not slides
stirring-seahorse-4fa4b6.netlify.appOpen ↗
Interactive ProposalLamps Plus
Rewards
One loyalty program for consumer and Pro, online and in store — built on our own customer research.
Open the live shareoutOpens in a new tab · roughly a 12-minute read · scenario model is illustrative, no client financials included
Overview

Why a site instead of a deck.

The study behind this shareout was ordinary mixed-methods work: a customer survey of 128 respondents, seven follow-up interviews, and a read of the competitive landscape. What made it land was the delivery. A loyalty program touches merchandising, ecommerce, IT, finance, and the showroom floor — five audiences who never sit in the same meeting and never need the same level of detail.

A deck forces one linear path through the argument, at one depth, for whoever happens to be in the room. A site lets the finance lead go straight to the revenue model, the merchandiser read only their segment, and the engineering lead filter benefits by build effort — from the same source of truth, without a follow-up meeting to re-explain it.

It also survives the handoff. Six weeks later the artifact is still a URL that renders on a phone, not a 40MB attachment in a thread nobody can find. That durability is the point: research that cannot be re-found cannot be re-used.

A shareout is a product.
Its users are a team about to make a decision.

Principles

Six rules I apply to every shareout.

The same UX thinking I bring to a product, turned on my own deliverable. The audience is a stakeholder under time pressure — design for them accordingly.

01

Open with the decision, not the method

Stakeholders do not arrive curious about sample sizes. The first screen states the tension and the stakes; sample, instrument, and limitations sit where the people who audit them will look.

02

One claim per screen

Each section carries a single argument and the evidence for it. If a section needs two claims, it is two sections. This is what keeps a long document readable in a 30-minute meeting.

03

Put the stated and the behavioral side by side

Self-report and behavior disagree, and the disagreement is usually the finding. Showing both in one view stops the room from arguing about which number is real.

04

Let the audience test the numbers

A static forecast invites a debate about the forecast. An interactive model with visible assumptions moves the debate to the assumptions — which is the conversation worth having.

05

Progressive disclosure over completeness

The main line stays short. Journey walkthroughs, tier architecture, stacking rules, and sources live in an appendix — reachable in one click, invisible until wanted.

06

End with what would change your mind

The closing section is a validation plan, not a victory lap. Naming the tests that could still kill the idea is what makes the recommendation credible.

How to Read It

A guided path through the artifact.

Six stops, in the order the argument builds. Each one opens the live shareout at that section.

01The problemOpens on the gap between what customers said and what they did. The tension is the argument — everything after it is support.02Who they areFour segments sized from the survey. Clickable, so a merchandiser can read one and skip three.03The marketA positioning map plus a benchmark table. The map makes the whitespace obvious; the table is there for the person who wants to check it.04The revenue caseAn interactive model with conservative, base, and upside scenarios. Assumptions are exposed and labelled illustrative — you can move the inputs and watch the case hold or break.05The recommendationOne platform, two doors. Benefits filter by build effort — ship now, some dev, or with the replatform — so the conversation stays honest about cost.06The askA phased plan and four validation steps before anyone greenlights spend. Research that names its own next test.
In Practice

What the format is good for — and what it isn't.

Worth building

Cross-functional decisions with real money attached, where the audience is distributed, the detail requirements differ by role, and the work needs to stay referenceable for months. Anything with a model in it — letting people move the inputs themselves is faster than three rounds of "what if we assumed…".

Not worth building

Weekly usability readouts, single-team tactical findings, anything with a 48-hour shelf life. Those go in the channel where the team already works, as a short written summary with clips. The format follows the decision, the same way method follows question.