Growing commercial-connect.net on the park research

SEO architecture, lead conversion inside the TREC rules, compliance as code, and a data pipeline. Planned 15 Nov 2026.

127park pages
0suite pages
8build steps
105-155hours total
Read section 2 first. I framed the TREC carve-out too loosely in chat and the plan corrected me. A form fill does not make someone your client, so the availability survey cannot be automated. It waits for your call.
Verified, not taken on trust: the footprint detector that would not work is working. I re-ran it myself. It found Brownstone and Grogan's Ridge by building shape alone, plus 13 candidate clusters in one small box that are not in our roster.

Two page types, two rule sets

Park record pagesThird party property facts. All 11 TREC rules. No economics, no availability, non-representation statement, everything dated and sourced.
Service and intent pagesYour services and your own listings. Only 535.155(a) and (d)(1). May say "office condos for lease in Katy" freely.
Capture block on bothClean advertisement. Auto-sends sourced public fact only.
After your callRepresentation established. The availability survey is now a client communication and outside the rule entirely.
!

What is broken right now

Five things are wrong on the live site today. Three of them make SEO spend worthless until fixed.

Also needs doing: grep the live pages and JSON-LD for jacob@dowreg.com, and prove the new alias actually receives a test message before it goes in any capture copy.

2

The legal hinge, and a correction to what I told you

I framed the TREC carve-out too loosely earlier. The plan pushed back and it is right.

This is the sentence the whole design rests on, so it is worth being conservative about it.

3

SEO architecture

Fold the submarket site into commercial-connect.net, build 127 park pages on the open named-entity terms, and a small set of city service pages on the open broker terms.

Roughly 36 to 54 hours for the whole SEO workstream.

4

Lead conversion

Replace the dead form with a Pages Function that writes to Pace and a D1 backup, and give each page type a value exchange that is legal to auto-send.

16 to 25 hours. This ships before any new page exists.

5

Compliance and brand as code

Extend compliance.py with a page_type flag so each page type gets the right rule set, and add rules for the things that have actually bitten.

11 to 15 hours. Blocks nothing shipping today and unblocks every new page type.

6

Data: detector 4 is solved, and I verified it myself

The footprint detector that would not work is working. I re-ran it independently rather than take the agent's word for it.

28 to 42 hours. Storage moves to D1 (dow-hub-db) so the site and the dataset cannot drift apart.

7

Build order

Roughly 105 to 155 hours total. The order is not negotiable in the first three steps.

#StepHours
1Lead capture Function, Pace + D1, retrofit the 7 existing pages8 to 12
2compliance.py page_type flag, rule matrix, prove-gates11 to 15
3Fold submarket site into /submarkets/, 301s6 to 10
4Office condo templates, city indexes, 106 park pages, broker sign-off22 to 32
56 to 8 city service pages, Market Report and fact sheet generators12 to 18
6D1 schema, loaders, Harris quarterly diff with owner-change digest10 to 14
7Detector 4 job, TDLR weekly watcher, candidate review queue12 to 18
8County expansion: bulk requests, Montgomery clerk procedure6 to 10 plus waiting

Steps 1 to 3 gate everything after them. Step 1 because traffic currently lands on a form that captures nothing; step 2 because no service page can pass the gate until it exists; step 3 because every week it waits is authority spent on the wrong domain.

Decisions I need from you

Fold the submarket site into commercial-connect.net?

Which PaceCRM location receives website leads?

Who is the backstop when you cannot answer in 15 minutes?

Gold text token on the website

Auto-send the Market Report on form fill?

CB&A review granularity

Fort Bend unit square footage

SMS automation to prospects

What do you want started first?