/blockstream workspace

▸

Current recommendation lives in the Launch Plan. This page is the deeper rationale for choosing HubSpot. The plan refines the rollout: pages ship on a validation subdomain first and HubSpot runs the lead layer, rather than HubSpot CMS rendering everything via reverse proxy.

The recommendation

IN ONE LINE

Use HubSpot Content Hub as the system for the enterprise landing pages and lead lifecycle, build and deploy them from Claude Code via HubSpot’s official Developer MCP server, and serve them on the main domain at blockstream.com/enterprise/… through a reverse proxy. One platform owns pages, forms, CRM, nurture, and reporting; the SEO accrues to blockstream.com; and the build stays in the workflow we already use.

It is Claude-Code-native now.

HubSpot shipped an official Developer MCP server (GA Feb 19 2026) that builds CMS themes, templates, and modules directly from Claude Code. The thing that used to make HubSpot “a closed tool” is gone.

The lead lifecycle is built in.

A form on each page creates a CRM contact and enrolls them in a nurture workflow — email sequence, lead scoring, LinkedIn retargeting, and sales handoff — with no custom integration code to maintain.

The domain placement is the SEO.

Pages at blockstream.com/enterprise inherit the main domain's authority and rank far faster than anything on a subdomain or a separate host.

It is fast, and it persists.

We skip building forms, CRM, nurture, scoring, and analytics from scratch — weeks, not months — on a platform with a large talent pool and vendor support, so the knowledge survives turnover.

1 · Programmatic deploy from Claude Code

The blocker we worried about — that a marketing SaaS can’t be driven from our existing Claude Code workflow — no longer applies. As of February 19 2026, HubSpot’s Developer MCP server is generally available and explicitly supports Claude Code. It exposes HubSpot’s CMS tooling to the agent: create and manage themes, templates, and modules in natural language. Combined with HubSpot’s HubDB and CMS APIs, the same data-to-pages pipeline we run today lands on HubSpot instead of a one-off host.

01 · CLAUDE CODE

Build the template

HubSpot Developer MCP (hs mcp setup, CLI 8.2+) scaffolds the page template & modules from a prompt.

02 · DATA

Push rows to HubDB

Our page data (persona, keyword, H1, copy, CTA) goes to a HubDB table via the HubDB API — one row per page.

03 · HUBSPOT

Pages generate

One dynamic template renders one live page per row. 200+ pages from a dataset, not 200 hand-built pages.

04 · PUBLISH

Deploy & serve

hs upload ships the template; new rows appear automatically at blockstream.com/enterprise/…

▸

This is the workflow we already run, repointed. We generate templated pages from data today. HubSpot adds the official MCP for Claude Code, plus the CMS Pages API if we ever want individually-tracked Landing Pages instead of dynamic ones. Either way, deployment stays scriptable and in-repo.

2 · How it feeds the nurture campaign

This is the half that is painful to build custom and that HubSpot gives us for free. Every page carries a native HubSpot form. A submission isn’t a webhook we have to wire — it creates a CRM contact and triggers a workflow automatically.

CAPTURE

Form → CRM

Submission creates/updates a contact, stamped with the page, persona, and campaign source.

NURTURE

Email sequence

Enrolled in a segmented track — typically 5–8 emails over 4–6 weeks for mid-funnel B2B.

SCORE

Lead scoring

Behavior adds points (pricing visit, repeat sessions, email clicks); threshold = MQL.

ROUTE

Multi-channel + sales

At high intent: add to LinkedIn retargeting audience, create an SDR task, notify the account owner.

Triggers are native. Form submission, specific page visits, persona, or a lead-score threshold can each start or branch a workflow — no middleware.

LinkedIn is a built-in action. A workflow can sync a contact into a LinkedIn ad audience for account-specific retargeting the moment they hit an intent threshold, alongside the email track.

One record, every channel. Email, LinkedIn, and sales outreach all read and write the same CRM contact — layering channels is where response rates compound, and it requires zero glue code.

Reporting is closed-loop. Page → form → nurture → MQL → deal is one attribution chain in one tool, so we can show which landing pages actually source pipeline, not just traffic.

3 · Why blockstream.com/enterprise must point to HubSpot

This is the crucial one, and it’s the reason a HubSpot subdomain (go.blockstream.com) is not good enough. The whole point of publishing these pages is to earn organic rankings, and ranking power follows domain authority.

Pages on the main domain inherit its authority.

blockstream.com has years of established domain authority and backlinks. A page at blockstream.com/enterprise/asset-tokenization rides that equity and can rank in weeks. The same page on a subdomain or separate host starts near zero — Google treats a subdomain as a semi-separate site.

/enterprise is already an authoritative path.

We are not creating a new property; we are extending one Google already trusts, and interlinking the new pages with existing high-authority pages — itself a ranking lever.

The mechanism is a reverse proxy.

HubSpot can't natively host a subdirectory of a domain whose root lives elsewhere. An edge layer (e.g. Cloudflare) routes blockstream.com/enterprise/* to HubSpot and everything else to the current origin. To Google and users it is one site. HubSpot officially supports this pattern.

It is the same requirement for any tool.

Anything we want at blockstream.com/enterprise — HubSpot or our own pages — needs this routing. So the SEO requirement isn't a HubSpot cost; it's a prerequisite for ranking at all, and HubSpot rides it cleanly.

▸

Get the hygiene right and the SEO lands: canonical tags point to the blockstream.com/enterprise URL, the underlying HubSpot URLs are set to noindex, and the main sitemap lists the proxied pages. With that, Google indexes only the main domain and consolidates all ranking signal there.

4 · Speed of implementation

HubSpot collapses the timeline because the slow parts — the lead lifecycle, not the pages — are already built and supported. We configure rather than engineer.

CAPABILITY	BUILD IT CUSTOM	HUBSPOT
Landing pages at scale	Page pipeline + hosting (we have this)	Dynamic templates from HubDB — 200+ from a dataset
Forms & lead capture	Build, store, secure, GDPR	Native, on every page
CRM	Buy/build + integrate	Included, same platform
Nurture, scoring, email	Sequencer + scoring engine + deliverability	Workflows, out of the box
LinkedIn / ad routing	Custom API integrations to maintain	Native audience sync + workflow actions
Closed-loop reporting	Stitch analytics across tools	One attribution chain
Time to live	Months	Weeks

The one-time cost is the reverse proxy,

owned by whoever runs the blockstream.com edge. After that, every new page appears under /enterprise automatically — no per-page ops.

It persists after any one of us leaves.

HubSpot is a standard platform: a large hiring pool, certified partners, vendor support, and documentation. The org can staff this without depending on the person who set it up.

▸

Continuity note for Head of Enterprise: the only bespoke piece is the edge reverse proxy. We document it in a one-page runbook and assign it an owner on the web/infra team. Everything else — pages, forms, CRM, nurture — lives in a platform the market already knows how to operate.

What it requires (so there are no surprises)

HubSpot Content Hub, Professional or Enterprise. Dynamic pages, HubDB, serverless, and the reverse-proxy pattern need a paid CMS tier. Budgeted line item.

A reverse proxy at blockstream.com's edge. One-time setup; easiest if blockstream.com is already behind Cloudflare or a CDN we control. Requires the web/infra team's sign-off — it sits in front of the main marketing site.

Canonical + noindex hygiene. Non-negotiable for the SEO to work; standard, documented, done once in the template.

Open question to resolve first: what currently fronts blockstream.com (Cloudflare? a specific CDN? raw origin?). That answer determines how fast the subdirectory routing goes in.

▸

Bottom line: HubSpot served at blockstream.com/enterprise gives us programmatic pages from Claude Code, a built-in nurture and lead-routing engine, main-domain SEO, and a weeks-not-months timeline — on a platform the organization can own for the long run. The single dependency to clear is the edge reverse proxy.

Sources: HubSpot Developer MCP server (GA Feb 19 2026) · HubSpot CMS dynamic pages / HubDB & CMS Pages API · HubSpot reverse-proxy support docs · HubSpot workflows / lead scoring / LinkedIn audience sync. Companion: Content calendar (the pages this powers).