REEL GOODSign in
ABOUT · A PACAYA DIGITAL CONCEPTCase study

Reel Good: A Personalized Guide to Local Movies, Music, and Games

Reel Good helps people answer a deceptively simple question: what is actually worth seeing near me this week?

A subscriber enters their ZIP code and immediately sees well-reviewed movies playing nearby. From there, they can tune the experience around their taste—distance, minimum Metascore, favorite theaters, film genres, and classics or revivals—before receiving a personalized digest every Sunday morning. Optional add-ons extend the same model to newly released games and music.

We designed and built the product end to end: the acquisition experience, live recommendation preview, preference system, accounts, curation engine, recurring email delivery, and engagement tracking.

At a glance

Product
Personalized local entertainment discovery and weekly email digests
Core experience
Enter a ZIP code, preview relevant films, adjust optional preferences, and subscribe
Default film criteria
Metascore 70+, within 25 miles, with favorites and nearby theaters prioritized
Optional add-ons
New games by platform and genre; new albums by music genre, each with its own score threshold
Delivery
Sundays at 9:00 a.m. in each subscriber’s local time
Technology
React, TypeScript, TanStack Start, Tailwind CSS, Lovable Cloud, and Resend

The challenge

Movie discovery is fragmented. Reviews live in one place, local showtimes in another, and personal preferences—favorite theaters, genres, and acceptable travel distance—usually live in someone’s head.

A useful product needed to bring those pieces together without becoming another complicated filtering tool. It also needed to solve several challenges at once:

  • Deliver value before asking for commitment. Visitors should be able to see relevant local films before creating an account.
  • Make personalization optional, not exhausting. Sensible defaults should produce a good digest with minimal input.
  • Keep recommendations trustworthy. A title should only appear when its review score and Metacritic page can be verified.
  • Reduce noise at the point of decision. A long list of every possible theater is less useful than a few well-chosen options.
  • Support different kinds of fandom. Movies, games, and music have different preference models but can share one digest framework.
  • Close the measurement loop. The system needed to connect each digest to delivery, open, and click events so engagement could be measured accurately.

Our approach

1. Lead with the ZIP code, not a registration form

The landing experience begins with the smallest useful question: where are you?

Entering any US ZIP code triggers a live local preview. The results belong to that location; the interface never quietly substitutes sample listings from another ZIP when a lookup fails. That decision makes the preview both more useful and more honest.

The page then introduces optional preferences and email signup after the visitor has already seen the product’s value. This creates a natural progression:

  1. Enter a ZIP code.
  2. Browse real local films and showtimes.
  3. Adjust preferences—or keep the defaults.
  4. Subscribe with an email address.
  5. Verify the email and manage the digest through an account.

The same curation logic powers the preview and the recurring email, so the signup experience accurately represents what subscribers will receive.

2. Use strong defaults and progressive personalization

Every preference is optional. A subscriber can get started with no configuration beyond ZIP code and email, while more engaged users can make the digest highly specific.

The default movie digest includes films with a Metascore of 70 or higher playing within 25 miles. Distance can be adjusted from 1 to 100 miles, and the film score threshold can be tuned across the full 0–99 range. Subscribers can also select film genres, include classics and revivals, and identify favorite theaters.

Games and music are separate, opt-in add-ons. Subscribers can save these preferences and preview matching releases today; their weekly email sections are the next delivery milestone. Each has its own Metascore threshold, also defaulting to 70:

  • Games: subscribers can choose from the platforms Metacritic uses for filtering and narrow releases by genres such as RPG or shooter.
  • Music: subscribers can select genres such as jazz, classical, or rock.
  • Empty selections mean no unnecessary restriction: leaving platforms or genres empty keeps that dimension open.

This model keeps the initial experience simple without limiting the product’s long-term flexibility.

3. Design for fewer, better decisions

A film can play at many venues, but showing every option at once makes the digest harder to scan. Reel Good initially presents three theaters per film, ordered by:

  1. Favorite theaters first.
  2. Then the closest theaters.
  3. A subtle expansion control for all remaining showtimes.

That small interaction rule reflects a broader product principle: surface the most likely choice, preserve access to everything else, and avoid making the interface feel crowded.

Metascores use a consistent green treatment, and movie titles link directly to their verified Metacritic pages. Readers can therefore move from recommendation to source without an intermediate search.

4. Treat the preview as part of the product

The interactive preview is not a decorative marketing mockup. It uses the same preference model and live listing pipeline as the subscriber experience.

As visitors change distance, score thresholds, genres, theaters, or add-ons, they can see how those choices affect their digest. This makes personalization tangible before signup and reduces the gap between expectation and delivery.

The visual design supports that clarity: a restrained editorial layout, Instrument Sans typography, minimal chrome, and content-led hierarchy. On desktop, the preferences panel scrolls independently from the movie results, with a scrollbar that appears only while scrolling and fades away afterward.

5. Add accounts without adding friction

Accounts arrive after the visitor understands the value of the product. Email verification and first sign-in use a magic link; setting a password is optional, and magic-link sign-in remains available.

The account page gives subscribers control over the complete preference set:

  • ZIP code and distance.
  • Film score, genres, classics, and favorite theaters.
  • Game score, platforms, and genres.
  • Music score and genres.
  • Optional password setup and magic-link access.

Saved multi-value choices appear as removable pills. Searchable multi-select controls make it easy to add several new options, and any section containing saved selections opens automatically so those choices are visible immediately. Empty sections offer a small “+ Add” action that opens the section and focuses its search field instead of presenting a dead-end “None selected” label.

Loading was treated as part of the interface rather than an afterthought. A reusable skeleton primitive supports an account-shaped loading layout, which remains visible until both saved preferences and theater names are ready. This prevents internal theater IDs from briefly appearing before their display names load.

6. Build a cost-aware, verification-first title-matching pipeline

Matching a local listing to the correct Metacritic page is not always straightforward. Titles can include punctuation variations, subtitles, re-release labels, or ambiguous names.

Reel Good resolves titles through a cascading pipeline:

  1. Fast programmatic matching handles clear cases at zero AI cost.
  2. Programmatic search investigates titles that need another pass.
  3. AI-assisted resolution is reserved for unresolved titles.
  4. An AI-suggested result is used only when confidence is 98% or higher and the proposed Metacritic page has been fetched and verified against the title.
  5. If no page can be verified, the film is not displayed as a scored recommendation.

Confirmed and missing resolutions are cached for different durations to avoid repeating work unnecessarily. Difficult titles hydrate separately from the main response, so confirmed movies appear quickly while late-arriving matches load in a second phase.

The important distinction is that AI helps establish identity; it does not invent a review score or make an unsupported recommendation.

7. Connect curation, delivery, and measurement

Recurring movie digests are generated per subscriber using the same shared preference and curation rules as the preview. The delivery system schedules emails for Sunday at 9:00 a.m. in the subscriber’s local time, includes duplicate-send protection, and sends through Resend using the Reel Good domain.

Authentication and other recipient-triggered account emails use the managed application email system, separating account messages from the recurring newsletter pipeline.

A signed Resend webhook receives delivery and engagement events. Opens and clicks are connected to the corresponding digest delivery, creating the foundation for metrics such as:

  • Delivery rate.
  • Open rate.
  • Click-through rate.
  • Which films or Metacritic links generate interest.

Webhook signatures are verified before events are accepted, and private delivery and engagement records remain accessible only to trusted server-side jobs.

Under the hood

The application uses a modern full-stack TypeScript architecture:

  • React 19 and TanStack Start for the interface, server functions, and application routes.
  • Tailwind CSS v4 and reusable interface primitives for a consistent visual system.
  • Lovable Cloud for managed PostgreSQL data, authentication, and row-level access control.
  • Resend for recurring digest delivery and engagement webhooks.
  • Shared, pure curation functions so preview and email generation apply the same rules.
  • Server-only resolution caching and privileged operations to keep sensitive work away from anonymous visitors.

Anonymous visitors can create a subscription but cannot read subscriber records. Verified accounts can claim and manage subscriptions matching their own email address. Authorization is enforced at the data layer rather than relying only on interface restrictions.

Email sending, title-resolution cache access, and engagement records are handled through trusted server-side paths. Public webhook routes perform their own caller verification because they are intentionally reachable by an external provider.

The HTML parsing path was also made compatible with the application’s edge runtime, avoiding server-only dependencies that work locally but fail after deployment.

Validation and reliability

The implementation includes automated coverage for the rules that matter most: preference filtering, score thresholds, theater selection, release curation, scheduling, duplicate protection, and webhook behavior. The latest complete test run passed 44 tests.

Key flows were also exercised in a real browser, including:

  • Loading and saving signed-in preferences across movie, theater, game, and music sections.
  • Keeping saved selections expanded after repeated refreshes.
  • Removing pills and adding multiple choices through search.
  • Delaying preference and theater responses to confirm that the skeleton stays visible and raw IDs never flash.
  • Receiving real Resend delivery, open, and click events through the live webhook.

These checks focus on behavior rather than screenshots alone, helping protect the product’s core promises as it evolves.

The result

Reel Good is now a complete, working discovery loop:

local intent → live personalized preview → verified account → curated weekly digest → measurable engagement

The product combines a low-friction consumer experience with a modular backend that can support richer preferences over time. Its most important technical choices—shared curation logic, verification-first matching, progressive hydration, data-layer authorization, local-time delivery, and signed event tracking—directly serve the user experience rather than existing as infrastructure for its own sake.

For subscribers, the result is a quieter way to discover something worth watching, with games and music previews pointing toward a broader entertainment digest. For the business, it is a reusable personalization and retention engine with a clear path to measuring what resonates.

What this project demonstrates

  • Turning an ambiguous consumer problem into a focused, usable product journey.
  • Balancing immediate value with progressive onboarding and account creation.
  • Designing preference systems that are powerful without being overwhelming.
  • Integrating live external data, verification workflows, and AI selectively and cost-consciously.
  • Building recurring, localized email automation with duplicate protection.
  • Instrumenting the full delivery and engagement loop.
  • Iterating on small interaction details—loading, scrolling, empty states, and saved selections—that make the product feel considered.
  • Validating business rules with automated tests and critical journeys with real browser checks.

Budget and timeline

Reel Good went from an idea to a live, branded app in about 6 working hours. A project like this one costs $1,000–$2,000.

Timeline
2–3 days from kickoff to a working app you can share
Investment
$1,000–$2,000
What moves it
How many rounds of feedback you want

The budget is primarily influenced by the initial spec and number of feedback iterations. One focused build pass sits at the lower end of the range. Each extra round of try it, tweak it, try it again loops adds about a day and moves you toward the upper end. Most projects land in the middle after two rounds.

That range covers everything this case study describes: the landing page and signup flow, live data wired to real sources, personalized email delivery, and the account area — plus the back-and-forth it takes to get the design right.

Talk to Pacaya Digital about a project like this one.