Through

Through has an unusual success condition for a digital product: brief use followed by meaningful action away from the phone.

The intended experience is a small private space where a young person can choose something they want, make a manageable commitment and look back honestly without entering another feed.

The early working title explained the product direction, but sounded more like positioning than something a young person might own. Through carries the action at the centre of the idea: choose something real and see it through, then get back to your life.

I treated the naming decision as more than visual branding. Public copy, consent language, email display identity, Figma, tests and production moved together before the first campaign signup. The canonical public site has since moved to through.brendan-flynn.com, with its governed signup flow remaining the public entry point; provider resources, campaign constants and dated evidence remain unchanged where migration would add risk rather than clarity.

The working thesis connects a chosen real-world action, a light-touch commitment and non-punitive reflection. The product should never turn missed intentions into shame or optimise for time spent inside the app.

That principle is firm; mechanics such as goal limits, notification cadence, coaching tone, streaks and rewards remain open to evidence.

The production refinement retained the orbit as the landing page's main visual motif: three restrained rings and three points, exported directly from Figma rather than approximated in CSS.

It scales from 340px on mobile to 480px on desktop, moves slowly and becomes static when reduced motion is requested.

The live experiment is a Republic-of-Ireland, 18+ signup that gives immediate access to a fixture-only concept experience. It is not a beta, youth research cohort or durable application, and it asks for no personal goal, reflection, voice recording or open-text response.

One transactional access email is sent, with at most one separate availability message later. An optional four-question fixed-choice feedback form has separate consent and no link to signup identity. Fixed dates, stated deletion controls and no analytics or profiling keep the experiment real without presenting it as product validation.

The proposed trust model treats personal reflection as sensitive data. It prioritises clear consent, data minimisation, age-appropriate safeguarding and careful escalation, while keeping AI support bounded in tone and purpose.

The adult access flow and optional unlinked feedback implement a narrow privacy boundary, but no youth-facing collection or safety system exists. Those broader safeguards remain requirements to investigate and validate.

The public pitch, coherent identity and bounded adult concept access make the direction discussable, but the fixture interaction and optional feedback are not participant research or evidence that the youth-facing premise works.

The next gate is research with intended users and age-appropriate safeguarding expertise to test the problem, language, trust expectations and smallest credible experience before application development begins.