Promising Beer — Site Strategy & Roadmap #

A working document capturing an objective read of promisingbeer.com as of April 2026, the chosen direction for the site, and a roadmap of what to build next.

Direction: a photo book, not a magazine #

Promising is built and maintained by one person. The instinct is small, personal, ownable. Long-form essays aren't the voice. Buyer's-guide listicles feel like a trap. The right reference point isn't a magazine or a blog — it's a photo book or a curated guidebook. Think Wallpaper* City Guides, Phaidon monographs, Cereal, Drift, Standart. Compact entries, considered sequencing, photography co-equal to text, and an identity resistant to engagement-grabbing cruft.

This frame resolves several earlier tensions:

• Cadence isn't a problem to solve, it's part of the artifact. Photo books don't apologize for being slow. The cadence of reviews is the cadence of the site.
• Editorial authority comes from taste applied consistently, not from volume of long-form. The existing three-paragraph review form already is the voice. Don't force essays.
• "Best of" lists are out. "Issues" and "collections" are in. Six German lagers as a curated spread is a photo book chapter. The same content as a listicle, but the framing, sequencing, and visual treatment make it an object rather than a trap.
• Photography becomes load-bearing. The lifestyle / environment shot (img_asset) carries the page; the product shot (img_beer) supports. A missing lifestyle photo is a quality bar, not an afterthought.
• The news section goes. It was a workaround for "I can't review fast enough" but it pulls the site toward blog-feel. Existing news posts stay at their URLs (don't 404 them — preserve SEO and inbound links), but no new posts are added and the homepage drops the "Latest news" block.

What's working #

• Voice and structure. The disciplined three-paragraph format with a Parker-style integer score is genuinely differentiated. Most NA review sites are either Untappd-style noise or affiliate-stuffed listicles. Promising is editorial in a category that lacks editorial.
• Technical foundation is unusually clean. Schema.org Review + BreadcrumbList + ItemList + FAQPage markup, Atom feed, sitemap with image sitemap, llms.txt + llms-full.txt. The site is set up to be cited by both Google and LLMs in a way most beer blogs aren't. That's a moat.
• Anti-social positioning. "Promising is anti-social, it's a website" is on-brand and scarce. Worth preserving.
• Color-as-identity per beer. The hex per review threading through the pipe / score / tags is small but tasteful — already photo-book-coded.
• Scoring page + scoring.json. The "modified Parker" framing is the kind of detail that builds trust.

What changes under the photo-book frame #

Cut or deprioritize:

• "Best of" / Wirecutter-style ranked listicles
• Long-form quarterly essays
• "Promising Awards" framed as awards (replaced by an annual issue, see below)
• Drinkability vs. quality split scores (over-engineering)
• News section (archive existing posts, stop producing new)
• Newsletter archive page (only relevant if the newsletter survives this reframe — open question)
• Heavy "buyer's guide" framing on the homepage

Keep and strengthen:

• Reviews — same form, slower cadence is fine
• Brewery profiles, treated as chapters, not exhaustive profiles
• Country / region pages, treated as travelogue spreads
• Issues / collections (curated groupings — "The German Issue", "The Bottle Issue", "Summer 2026")
• Filtering on /beers/ (navigational utility, not content)
• On-site search
• Retailer / "where to buy" links (utility, doesn't compromise curation)
• ABV truthfulness as a small standing reference page, not a recurring beat
• Per-review auto-generated OG images
• Revisit / re-review logic (fits photo-book "second edition" thinking)
• Public structured dataset at /api/beers.json (low cost, high LLM-citation value)
• Photography quality bar — environment, light, composition

Roadmap #

Quick wins (a weekend each) #

• Remove the news section from the homepage and nav. Keep existing posts at their URLs, drop the "Latest news" block, drop news from primary nav. Add a quiet /news/ archive link in the footer if anywhere.
• Add Pagefind search to /beers/.
• Add filtering on /beers/: style, origin, <= calories, true 0.0%, low-sugar, gluten-free. Data is already in frontmatter.
• Generate per-review OG images at build time (satori or build-time canvas). Beer name + score + brewery + color band.
• Add a revisedDate field and a small "Last reviewed Feb 2021 — due for a revisit" footer per old review.
• Add brewery-direct + one major retailer link per review.

Medium-term builds #

• Brewery chapter pages at /breweries/[brewery]/ — short brand portrait, full lineup of reviews, brewery score average. Built from existing brewery tags + a small _data/breweries.json.
• Country / region spreads — /places/germany/, /places/uk/, /places/usa/. Travelogue treatment, not exhaustive listings.
• Issues / collections at /collections/[slug]/ — curated sets of 5–8 reviews with a short editor's note. First three could be drawn from existing reviews and require no new tasting. Examples: "The German Issue", "Light & Crisp", "When You Want a Stout".
• Public structured dataset at /api/beers.json (CC-licensed). Mention it in llms.txt.
• ABV truthfulness reference page — a single standing page listing every "0.0%" beer with its actual declared / lab ABV and sourcing. Linked from individual reviews where relevant.
• Homepage redesign toward "cover spread" — see below.

Annual / slow #

• Promising 2026 — a year-end issue. Curated retrospective, not awards. One page, photo-book treatment, sequenced.

Homepage shift #

Today: hero, featured review, latest news, recently reviewed, brewery logos, scoring teaser, About / Scoring CTAs. Reads as a blog with utility blocks below the fold.

Photo-book direction: hero / cover, one hero spread (the most recent or most notable review treated like a cover story with strong photography), a quiet way into "All beers" / "Issues", and not much else above the fold. Less density, more confidence. The brewery logo row and the "zero percent" graphic block can stay but should feel like back-matter, not feature blocks.

This is the most visually impactful change on the list and worth mocking up before building.

Photography as a quality gate #

If photo book is the model, photography is no longer decorative. Practical implications:

• The lifestyle / environment shot (img_asset) becomes the primary image on the review page. The product shot (img_beer) is supporting.
• A review without a strong lifestyle photo is incomplete — the existing pending-review queue (reviews_pending/) should hold for photo, not just for prose.
• Treat the photo as part of the review, not as illustration. Composition, light, environment, the can or bottle in context.
• Long-term: a consistent visual language across reviews (similar lighting, similar framing, similar palette discipline) will make the site read as a single body of work rather than a stack of posts.

Open questions #

• Newsletter. Does it fit the artifact? A photo-book site can have a newsletter, but a Mailchimp embed with no archive and no editorial cadence is closer to blog-feel. Decide whether to lean in (a quarterly "issue letter") or drop.
• Tags vs. issues. Tags are a blog primitive. Once "issues / collections" exist, are tag pages still pulling their weight, or do they become a back-matter index?
• Featured quote / featured_post. Currently a homepage mechanism. Under the cover-spread homepage, this becomes the cover-story selector — worth formalizing.
• Single-reviewer framing. Eventually worth an explicit one-line statement somewhere ("Reviewed by one palate, on purpose") to make the constraint into the point. Not urgent, but better than silence.

Appendix: Stack question — Eleventy vs. Next.js / React #

Recommendation: stay on Eleventy.

Of every item on the roadmap above, none requires a JavaScript framework. The list is content, structure, and SEO work — exactly what static-site generators are best at:

What Eleventy currently buys:

• Clean SEO output and schema markup that's already paying off
• Free / trivial hosting, instant deploys
• Markdown reviews — content-first, no framework lock-in
• LLM-friendly output (llms.txt, llms-full.txt) is straightforward
• Zero JS runtime cost, perfect Lighthouse profile is achievable
• Tiny mental overhead per review — important for an army of one

What a Next.js / React migration would cost:

• A multi-month rebuild with high SEO regression risk on a site whose SEO is actually working
• Bundle size, hydration cost, and build complexity for no user-visible benefit
• Hosting moves from "static file dump anywhere" to runtime
• Engineering bandwidth diverted from content — the actual bottleneck

The only things that would genuinely justify a framework move are server-rendered or stateful features the roadmap doesn't currently call for:

• User accounts, comments, community ratings, submission forms
• Personalized recommendations
• Real-time data (live availability, price tracking)

If those become priorities, the right move is probably Astro, not Next.js — Astro is static-first like Eleventy but supports React islands where needed. That's a much smaller leap than a full Next.js port and preserves the SEO posture.

Verdict: the framework isn't the bottleneck. Photography quality, on-site search, filtering, brewery chapters, and curated issues are. Spend the engineering budget there. Reconsider the stack only when a planned feature genuinely requires server-side state.