New Jersey Lead Acquisition / Channel Strategy
Facebook Groups as an
NJ Therapy Lead Channel
A repeatable system for turning New Jersey communities into booked therapy sessions — without getting our accounts banned.
If you read one slide / 02
The recommendation, in one breath
Manual-first,
relationship-led
Real people posting in real communities. No posting bots. No scraping-to-post. A light human workflow + a simple database.
Borrow · Build · Own
Borrow reach in others' NJ groups. Build reputation as the trusted clinical voice. Own a group we control — the compounding asset.
Meta closed the shortcuts
Groups API killed (Apr 2024) — no auto-posting, no bulk member reads. A 100+ person anti-scraping team. Automation loses accounts.
Finding 30–50 groups is a day or two of work. What we're really building is a machine for showing up well, repeatedly — and a home audience that's ours.
The reframe / 03
Codrut asked the right question — the wrong way
“How do we discover & scrape a database of groups and their metadata?”
“How do we turn NJ communities into booked sessions without losing our accounts?”
The database is a means, not the moat. A list of groups with member counts wins nothing on its own — competitors can build the same list. What compounds is distribution, reputation, and an audience we own. So we optimize for those and treat the group list as a lightweight CRM of where we show up.
The landscape shifted under this play / 04
Meta spent 2024–25 killing every Groups automation shortcut
APRIL 22, 2024
The Groups API was deprecated
No cloud scheduler can auto-post to groups anymore — not Buffer, Hootsuite, SocialBee, Later, or Zapier. They downgraded to “reminder posting” (ping your phone, paste it yourself). Only Meta Business Suite schedules into a group — and only one you own.
APRIL 2024
The member-data API was cut too
The groups_access_member_info permission was removed — so no app can harvest member data programmatically. A person can still eyeball a public group's members by hand; a bot at scale can't, and trips the anti-scraping team.
ONGOING
Enforcement is industrial
Meta's 100+ person External Data Misuse team (its own 2021 figure) detects & blocks scraping — cease-and-desist letters, account bans, lawsuits. Their Terms forbid automated collection. The real risk isn't legality, it's losing the accounts we post from.
Sources: Meta Graph API changelog — removal of publish_to_groups & groups_access_member_info (Apr 2024); TechCrunch, “Meta cuts off third-party access to Facebook Groups” (Feb 2024); Meta Newsroom, “How We Combat Scraping” (Apr 2021, EDM team); Meta v. Bright Data (2024).
The recommendation / 05
Manual-first, built on three pillars
Borrowed audience
Source, vet & post — manually and phased — inside others' NJ groups. Reach without owning. This is the immediate-traffic layer.
Reputation
Become the recognizable, helpful clinical voice. Every touch clearly tied to the PRE / SW clinic brand, so it's obvious who's helping.
Owned audience
Grow our own NJ therapy / wellness group. The only ban-proof, schedulable, compounding asset in this channel.
Underneath: a disciplined human workflow and a plain database. Zero posting bots. At most, a one-time, read-only discovery assist. That's the entire technical footprint.
Answering “is scraping viable?” / 06
The alternatives we evaluated
Scraping + auto-posting bots
PhantomBuster / Apify to harvest data, bots to post. Fatal problems: posting API is dead, member lists hidden, direct ToS violation, fast account bans, legal exposure. Highest risk, lowest durability.
Cloud schedulers / reminder tools
Buffer, Publer, SocialBee, Later. They can no longer auto-post to others' groups — only reminder-post. Genuinely useful in exactly one place: scheduling into our own group via Meta Business Suite.
Manual sourcing + light assist
Humans discover, assess & score each group in one pass, then post manually and phased; a Notion/Sheet database. No scraping — the judgment (spam? relevant? active?) is the actual work, and only a person can do it. Safest, most durable, compounds.
Pillar 01 · Borrowed audience / 07
Find the right groups — vetting beats volume
Find — Facebook search by audience segment × NJ geography (full query bank later in this deck). Target segments:
- Caregivers & family support
- Postpartum & new moms
- Older adults / seniors & their families
- Grief, loss & divorce support
- Chronic-illness & health communities
- NJ town / county “community” groups
Vet — a group earns a slot only if it clears every box:
- Real US / NJ members, active in the last 7 days — not spam-farm or bot-inflated
- Allows relevant posts (many allow user posts; fewer allow Page posts)
- Rules permit our content — read the pinned rules; some ban promo or links
- Real engagement — comments on posts, not just a big member number
- Admins are active & reachable
Pillar 02 · Reputation / 08
Become the voice, not the ad
Educational & helpful
Answer questions, share coping tips, normalize therapy, show up in threads. Mostly no link — just value. This is what builds recognition and earns the right to promote.
Signposted to the brand
Post as — or consistently signed by — the PRE / SW clinician identity, so it's obvious who's being this helpful. Familiarity is the point.
Direct announcements
“We have openings at our NJ clinics — here's how to book.” Immediate-income posts that convert once trust exists. Sparingly, and where rules allow.
The opinion-leader play: the same 2–3 real clinician identities showing up repeatedly → familiarity → inbound DMs and profile visits. Reputation compounds. Ads don't.
Pillar 03 · Owned audience / 09
Grow a group we control — the real prize
- Automation is legitimate here. Meta Business Suite lets us schedule posts into a group we own — the one place tooling still works.
- No gatekeepers, no ban risk. We set the rules, we own the members, no admin queue holding our posts.
- Seed it deliberately. Invite people we meet in other groups (DMs), cross-promote from clinician profiles, run it as a genuine NJ mental-health support & resource community.
- It becomes a durable asset. An owned lead list and a retargeting audience that appreciates over time.
Borrowed audience funds the wait.
Owned audience is the payoff.
Slow to grow — think months, not weeks — which is exactly why we start it now, not later.
The repeatable machine / 10
One loop, run every week
Discover
FB search × the segment query bank.
Vet & score
Spam / relevance / rules checklist → tier A/B/C.
Log
Add to the CRM with segment tag & score.
Warm up
Join & engage 1–2 weeks before posting.
Post
Phased, staggered, non-duplicate copy; respect rules & admin approval.
Track
UTM link → HubSpot; log the post & response.
Maintain
Monthly re-check activity & rules; prune dead groups.
Repeat
The loop feeds itself and the database gets smarter.
Guardrails / 11
How we don't get banned
- Use aged, real, warmed profiles — never brand-new accounts. Ideally a few clinician-branded personal profiles plus our Page.
- Volume is fine — pace is what matters. ~10–15 groups/day per profile is proven safe; just stagger across the day instead of firing all at once.
- Never post identical copy across groups — duplicate text is the #1 spam signal. Vary the wording every time.
- Page vs user reality: most groups allow user posts, fewer allow Page posts. Plan identities to match.
- Admin approval: most groups queue posts for review — build admin relationships and ask permission to post or promote.
- No VPN hopping, no automation tools on any posting account. Keep them clean and human.
The stack / 12
Deliberately lightweight tooling
| Layer | Tool | Notes |
|---|---|---|
| Discovery & scoring | Facebook native search + our query bank | Free. Find, assess & score each group in one manual pass. |
| Database (CRM) | Notion (or Google Sheet) | The group CRM. Schema on the next slides. |
| Scheduling — our group | Meta Business Suite | The only legitimate group scheduler that remains. |
| Posting — others' groups | Manual, from team members' profiles | No true auto-post exists anymore. People post. |
| Tracking | UTM links → GA4 + HubSpot | Ties every post to booked sessions. |
Direct answers / 13
Codrut's seven questions, answered
| 1Best discovery method? | Manual FB search by segment × geo with a maintained query bank; find, assess & score each group in one pass. ~1–2 days for 30–50 groups. No scrape needed. |
| 2Is scraping viable? Tools? | For posting — no, the API is gone. For discovery — PhantomBuster/Apify can dump names+URLs, but they can't judge spam, relevance or engagement (the real work), so they're not worth the ToS/account risk. We do it by hand. |
| 3Limitations / risks / compliance? | ToS bans automated collection; account bans & IP blocks; a 100+ person enforcement team; legal exposure; posting rate caps; admin gatekeeping. |
| 4Automate, or an alternative? | Alternative: a human workflow + our own group (schedulable via Meta Business Suite). No posting bots on borrowed groups. |
| 5Keeping the DB updated? | Monthly re-vet (activity, rules, admins); prune dead groups. Vlad owns the CRM. It's a living CRM, not a one-off export. |
| 6Data fields realistically capturable? | Name, URL, segment, rules, post-permission, admin, last-active, score/tier — manual & reliable. Member count read manually. “Engagement” is a judgment, not a clean number. (Next slide.) |
| 7Effort & infrastructure? | Internal, no VA: Vlad builds the system (discovery + CRM); team members post from their own profiles. Notion + UTM; near-zero software cost. Ramp over 30 / 60 / 90 days. |
Setting expectations / 14
What we can realistically capture
- Group name & URL
- Audience segment tag
- Member count (read manually)
- Post permission — user / page / none
- Admin approval required?
- Posting-rules summary
- Admin names & reachability
- Last-active date
- Score / tier — A · B · C
- Our status & notes
- Precise engagement rate — no public number; approximate from recent comment volume
- Automated member harvesting — API access removed 2024; blocked at scale (viewing public-group members by hand is fine)
- Historical growth curves
- Demographic breakdowns
Measurement / 15
From a post to a booked session
A single UTM convention makes every post traceable all the way to revenue:
| utm_source | |
| utm_medium | group |
| utm_campaign | nj-therapy |
| utm_content | {group-slug}--{segment} — which group & audience |
| utm_term | {post-theme} — educational / announcement / etc. |
KPIs: groups active · posts/week · clicks · leads · cost-per-lead (≈ labor) · booked sessions · conversion by segment. Each group carries its own slug — so we learn which segments actually book, and double down.
Starter toolkit ① / 16
The group database schema
| Field | Type | Purpose |
|---|---|---|
| Group name | Text | Identify at a glance |
| URL | Link | Jump straight to the group |
| Segment | Select | Caregivers · Postpartum · Seniors · Grief · Chronic-illness · NJ-local |
| Members | Number | Rough scale (read manually) |
| Post permission | Select | User · Page · None |
| Admin approval | Checkbox | Posts queued for review? |
| Rules summary | Text | Promo allowed? Links allowed? |
| Admins | Text | Names + whether reachable |
| Engagement | Select | Low · Medium · High (judgment) |
| Score / tier | Select | A · B · C — relevance + activity + size + access |
| Last active | Date | Prune anything gone quiet |
| Our status | Select | Not joined · Joined · Warming · Posting · Blocked |
| UTM slug | Text | This group's tracking id |
| Notes | Text | Admin contacts, post history, gotchas |
Starter toolkit ② / 17
NJ search-query bank
Run each in Facebook search → Groups. Swap in NJ towns/counties (Newark, Jersey City, Montclair, Princeton, Morristown, Toms River, Bergen, Monmouth…).
Caregivers
NJ caregivers supportcaregiver support New Jerseydementia caregivers NJcaring for aging parents NJPostpartum / new moms
New Jersey momsNJ postpartum support{town} moms groupfirst time moms NJOlder adults / seniors
NJ seniors community{county} 55+ groupNew Jersey retireessenior living NJ familiesGrief / loss / divorce
grief support New JerseyNJ divorce supportwidow support NJloss & healing New JerseyChronic illness / health
NJ chronic illness supportanxiety support New Jersey{condition} support NJmental health New JerseyNJ local community
{town} community group{town} NJ residents{county} moms & familiesliving in {town} NJStarter toolkit ③ / 18
Posting cadence (per profile)
| Stage | When | Activity | Volume |
|---|---|---|---|
| New-group courtesy | First few days in a freshly joined group | React & comment before your first post there — don't join and immediately promote. | 0 promo in that group |
| Steady posting | Ongoing, per profile | Educational cadence + occasional announcements, staggered across the day. | 10–15 groups / day |
| Our own group | Ongoing | Scheduled posts via Meta Business Suite. | No cap (it's ours) |
Never the same copy twice — vary every post.
Stagger across hours, not a single burst.
Roughly 80% value : 20% ask.
Rollout / 19
30 / 60 / 90-day plan
- Vlad vets, scores & logs 30–50 groups (~1–2 days)
- Build the CRM + posting system
- Stand up UTM + HubSpot tracking
- Launch our own NJ group
- Team starts educational posting in the top 10
- Phase into all vetted groups
- Begin announcements where trusted
- Grow own group via DMs & cross-promo
- First conversion read by segment
- Double down on top segments
- Prune dead groups, formalize admin ties
- Own group becomes a real audience
- Report CPL & booked sessions
The ask / 20
What we need to agree on today
- Approve the approach — manual-first, no scraping/posting bots — as our direction.
- Confirm the posters — which team members will post from their own profiles.
- Green-light our own NJ group now — it has the longest lead time.
- Confirm UTM → HubSpot wiring reuses the existing booking form.
Approve it and I'll have the CRM, query bank, scoring rubric and post templates ready — we can start posting within two weeks.