A build plan for Alder Behavioral Health

Part one repairs and launches the site Alder already has, and it stands on its own. This page is the other path: a new Alder site on my own engine, with the schema, speed, and booking work built into its bones instead of patched onto a page builder from the outside. Your nine pages of copy took years to get right, and every word of it comes along. What changes is who holds the keys.

I want to be plain about the recommendation, because part one already made it: the repair path is the default for a practice your size, and this document exists because you asked the ownership question on our call, not because I am steering you here. The build costs $2,000 more than the repair path and buys three things the repair path cannot: a site you edit yourselves by clicking the text, machinery nobody rents you, and repairs that stop being a category of work you ever pay for again. Whether those are worth $2,000 to Alder is your call to make with all the cards face up. This page is the cards.

Dana sent three questions after reading part one, Sam sent two, and Sam added one aside. They are good questions, the kind that separate a real decision from a sales pitch, so each gets a plain answer here. Every answer holds for both parts of this plan.

Half of part one's repair phase is fixing machinery that belongs to other people: a page builder on a $23-a-month subscription, templates on its terms, image handling you cannot reach. The repair path works around each limit carefully, and it is real work that holds. The build removes the limits instead. Line by line, from the audit's own findings:

What does not change either way: the listings cleanup, the Business Profile claim and build-out, the review engine inside SimplePractice, and the measurement baseline. That work is platform-independent, it lives in phase 1 below on this path, and choosing part one first wastes none of it if you upgrade later.

Alder already owns its raw material: the leaf mark from 2020, the practice's plain, warm voice, and the alders along Millrace Creek behind the building that gave it its name. These are the two directions I would explore with you in the design pass, described here so you can see the thinking. Real type tests and full-page mockups come first in phase 2, and you choose the direction before any production work begins.

Direction A · Creek light

Paper-warm ground, alder green carrying the brand, river-stone grays doing the quiet work. A warm text serif for the practice's voice, a plain sans for wayfinding, both tested against the leaf mark before anything is committed. The feel is the practice at ten in the morning: unhurried, daylight, nothing clinical and nothing precious. This is the direction closest to what Alder already is, and the furthest from the teal-and-stock-photo counseling sites in the county.

Direction B · The late appointment

The same bones on a deep evergreen ground with lamp-warm light. This one is designed for the person searching at ten at night after a hard week, and it reads like the room they are hoping to find: low-lit, steady, honest. Bolder than A, and probably too much for the whole site.

The likely outcome: A leads the site, B is reserved for the depth pages: anxiety, depression, grief. A practice that mixes daylight and lamplight is telling the truth about what counseling is. Photography direction: the actual waiting room, the actual creek path, Ashford light. No stock photos of hands on windows, no models pretending to feel better.

Twelve pages, each with one job, plus the privacy and consent pages that ride along in the footer. Your existing copy seeds every page marked as carrying over; the new pages are the exact gaps the audit found. Each of the nine current addresses gets its written redirect to the closest new home.

Fixes 7 and 8 from part one's list (the service pages and the location page) live here on this path. Nobody pays for the same page twice; that is the whole point of choosing before the work starts.

The engine is UlricCMS, the same one running ulric.studio itself. There is no dashboard maze and no ticket to a developer for a typo. Signed in, the site becomes editable in place:

• Click the actual text on the actual page. Not a form field that vaguely corresponds to it. The paragraph a client reads is the paragraph you edit.
• Type the change. A new group start date, a fee update, a note that Sam has Tuesday openings. The design system keeps the formatting steady underneath you.
• Save. It is live. No cache to clear, no plugin to update, no waiting. The change is on the page before you close the tab.

Photos work the same way: click one, upload from your phone, done. The engine converts phone photos automatically and serves the right size everywhere, which is how the 6.4 MB problem stays solved after I am gone.

No client data, by design. Scheduling, notes, and records stay in SimplePractice where they belong. The site is marketing only; there is nothing clinical on it to protect, which is exactly how a counseling practice's website should be built.

Owned, not rented. The code is yours at final payment and the engine is licensed to you free, forever. If you and I part ways, everything keeps working, and any developer can pick it up. None is required to run it.

Flat mockups flatter everyone. A page that looks composed in a screenshot can feel wrong the moment it scrolls, and most people cannot judge a design until they watch it move. So at each checkpoint you get a short rendered video of the actual screens, built as code and filmed by the same pipeline I use for the studio's own motion work. You watch a one-minute clip on your phone between sessions, reply with what feels right and what does not, and the design moves. No design software, no screen-share appointments, no guessing from stills.

• Checkpoint one, the direction pass. Both directions from section 4 as scrolling homepage walkthroughs, side by side. You pick the lead before anything else is designed.
• Checkpoint two, the key pages. Home, one service page, and the Ashford page, read top to bottom on phone and desktop, so the rhythm is visible instead of imagined.
• Checkpoint three, the pre-launch pass. The built site itself: every page, the booking path click by click, and the editor demonstrated, so launch day holds zero surprises.

Pippin is the studio's own assistant product, not a third-party widget. For Alder she would be trained only on the twelve pages above, answering the questions people type at midnight and walking them gently toward the booking portal. She runs on your own accounts, on inference that costs nothing at a practice's scale, and there is no license and no subscription: the $2,000 is my work training and fencing her, not rent.

• Grounded in your words only. She draws from your pages, cites which page an answer came from, and never touches the open internet. If she does not know, she says so and offers the portal.
• Fenced hard, in writing. Information, never advice. She does not assess anyone, does not discuss a visitor's situation clinically, defers every such question to you, and stores no personal information. You approve her guardrails before she answers a single visitor, and you can pause or remove her anytime.
• She earns her keep twice. The questions visitors ask become anonymized fuel for the FAQ page, in the exact words other people search with.

My honest counsel, consistent with part one: do not buy her on day one. Launch, watch the first scorecards, and add her when the traffic justifies a night shift. She installs as cleanly at month six as at launch.

Two phases instead of part one's two, rearranged so no dollar buys the same work twice. Everything platform-independent starts first; everything the new site makes unnecessary is deliberately absent. Each phase is deliverable-based at a fixed price: half to begin, half when you have seen it working, and you can stop after either phase and keep everything built.

After cutover, leg four of part one applies unchanged: the Tempo floor at $400 a month runs the upkeep and reports to the same scorecard, month to month, cancel anytime.

* The $500 you paid for the audit is credited against phase 1's closing balance, which is why the path bills $5,500 rather than $6,000. Saying yes costs $1,000 today, half of phase 1. Every other dollar waits for its own milestone, and stopping at any line of the table keeps everything above it.

One row a prescribing practice would see here is missing on purpose: certification fees. Alder is a counseling practice and does not prescribe, so nothing like LegitScript applies, and no line pretends otherwise. If your plan ever includes a fee, it will be a real one, payable to the party that charges it.

The same thirty-minute call covers both documents; you do not need to pick a lane before it. What is worth knowing going in:

1. The first work is nearly identical either way. The listings cleanup, the Business Profile, the review engine, and the measurement start first on both paths. The only real fork is whether the schema and speed work happens inside the old site or the new one, and that decision can wait for the design conversation.
2. Choosing part one first costs nothing later. Every dollar of the repair path except the old site's plumbing work carries forward if you upgrade to the build in a year. I will tell you on the call exactly which dollars those are: about $1,000 of the $3,500.
3. If it is the build, the phase 1 agreement lists everything in section 9, the first invoice is $1,000, and checkpoint one's walkthrough videos arrive before the design goes any further than section 4's descriptions.

And the close from part one holds here too: if neither path goes ahead, both documents are yours, and any competent web person can execute either one. Back to part one: the growth plan.