Skip to content
Scott Farrar
Home Portfolio General

Double diamond case study

Designing KidLinks: Connecting with the Parents of Your Kids' Friends

Parents do not need another social network. They need a lightweight view into the connections they already have through their kids.

A case study of the relationship map designed to help families keep track of the parents behind children's friendships.

Product Design

Open the working prototype

KidLinks double diamond A hand-drawn double diamond showing discover, define, develop, and deliver phases.

Discover

My kid is making lots of friends, but I do not have good ways to keep up.

Define

Need a way to wrangle the contacts for connections made by your kids.

Develop

Simple flows: add a new friend before knowing the parent, link a parent after meeting them, and keep it as a view of contacts, not a social network.

Deliver

PM as creator: use AI coding to turn the product thesis into a working prototype.

Discover

The problem appears at birthday-party speed.

The spark was simple: inviting kids to my child's birthday party, while realizing that I did not have all the parent contact info. Some kids were from playdates, some from last year's class, and others were new friends I had only heard about at the dinner table.

I could visualize the web of adults around these kids: parents saved as “Ivy's friend Sam's Dad,” parents connected through old classroom lists, and families that had not quite exchanged contact information yet.

Contacts

I

Ivy's friend
Sam's Dad

messagecallmail
mobile(555) 013-4288
companyIvy — school / Sam's Dad
notesMet at pickup. Maybe parent of Sam from last year's class?
The current workaround bends a contact record into a memory aid.

Define

A lightweight relationship layer, not a new network.

The core user is any parent trying to navigate third-hand connections: parent to kid, kid to friend, friend to parent. The job is not to build a social graph for its own sake. It is to help a parent answer, “How do I know this person, and how can I contact them?”

  • Keep the app simple enough to use during real parent life.
  • Avoid accounts and heavy contact permissions.
  • Make unknown information representable instead of blocking entry.
  • Hand off to the phone's default contact app once the right person is found.

Develop

The alternatives were either too ad hoc or too heavy.

Today, parents often solve this with notes in their contacts app, hacked company fields, names like “Ivy's friend Sam's Dad,” or memory. I considered writing KidLinks data directly into contact notes, but that would require more permissions and would be less flexible.

School directories used to fill some of this space, but they are less common now. Facebook can technically connect adults, but it carries too much social baggage for a quick, practical parent-to-parent connection.

  • Relationship-first, not contact-first.
  • Support the dinner-table clue before the full contact record exists.
  • Remember how we know someone without creating a dossier.

Product principles

Lightweight:
No account, no social feed, no heavy setup.
Respectful:
Not surveillance on children, and not a dossier on families.
Flexible:
Add a new friend quickly, then fill in parent or guardian details later.
Connected:
Link to real contact records when the parent gets the info at pickup or dropoff.

KidLinks

Search kids, friends, parents
Ivy12 connections
+ child+ friend
SamLast year's class
MayaPlaydate
LeoParent unknown
Home screen concept: quick entry first, details later.

Ivy's network

school
playdates
sports
IvySamNoraBenMayaAriLeo
Graph concept: zones make messy parent knowledge easier to scan.

Ivy / friends / Sam

Sam's Dad

Sam is Ivy's friend from school.

known throughIvy → Sam → Sam's Dad
contact statusLinked to iPhone Contacts
Open in Contacts
Detail concept: context first, then a clear handoff to the phone's contact app.

Deliver

PM as creator: shaping the product and building the proof.

I wanted this case study to show more than prioritization and framing. It is also an example of product management that moves directly into creation: using AI coding to translate a product thesis into a working web-based mockup prototype.

The prototype lets the PM work stay concrete. The user need becomes data structure, navigation, empty states, graph views, and deliberate handoffs to the phone's contact app. That loop is the point: think like a PM, then make enough of the product to test the thinking.

hand-drawn sketches

→

paper prototype

→

ChatGPT Codex

→

GitHub

Try the KidLinks prototype