Where we are, where we're going

Where we are right now (5 days in)

• Engagement started Friday June 5
• Spend pulled back from ~$4,000/day weekday average to ~$1,000/day intentionally
• Two clean campaigns running: Branded (manual CPC, captures demand cheaply) and General Rehab (with 101 negatives applied across 5 waste categories)
• All historical campaigns paused (Addiction Landing Page, Ketamine Therapy, Adolescent Search, etc.)

What the previous setup actually looked like

Pre-engagement non-brand performance (favorable estimates, since half of spend was hidden by Google's privacy threshold):

For context, qualified opportunity means anyone past Missing Info in Salesforce. So VOB on Queue, VOB In Progress, Priority Follow Ups, Follow Ups, Adolescent Follow Ups, Needs Travel/Detox, At Detox, Scheduled for Admission, all the way through Admitted and Alumni.

What we've already changed this week

• Built a dedicated Branded campaign. Manual CPC, low bids, with 53 clean brand keywords (Ambrosia, Neuroscience Institute at Ambrosia, The Academy at Ambrosia). Stops paying premium prices to reach people who were already coming to you.
• 101 negative keywords applied to General Rehab. Across 5 waste categories: 29 competitor facility names, 20 wrong-customer patterns (sober living, halfway house, Medicaid, court-ordered), 16 off-menu services, 25 medication research terms, 11 informational queries.
• Rebuilt the conversion tracking architecture. Salesforce qualified opportunities and admits now feed Google Ads directly via the API. No more optimizing toward 60-second phone calls.
• 98 historical qualified opportunities and admits pushed back to Google as bootstrap. The bidding model starts with real signal instead of cold.

What we found in Salesforce (where admits actually come from)

Pulled the full 365 days of Salesforce admit data. Headline numbers:

• 1,011 admits last year, ~$10.8M in estimated revenue
• 82% from Florida, 18% out of state (174 OOS admits, ~$2.17M revenue)
• Out-of-state patients average $12,454 per case vs $10,580 in Florida

Top out-of-state markets by admit volume (last 365 days)

Revenue and case values are pulled from Salesforce's Estimated Revenue field. Actual Revenue is empty across the dataset, so accuracy of these estimates is unverified. Worth validating against collected revenue at some point.

State preferences (Saturday) overlaid against admit data

States flagged on Saturday: GA, TN, IL, IN, MO, NJ, NY. Removing: TX, SC. Open clarification on the "above 90" comment.

States not on the list with significant admit volume (worth discussing)

MA is the largest OOS market by volume and revenue. Are these intentional omissions or just not top-of-mind?

Current out-of-state spend reality

Proposed OOS launch plan

Phase 1 launch (tier 1): NY, GA, NJ. Highest admit volume, highest case value, and zero current paid presence. NY has the best historical close rate of any out-of-state market.

Phase 2 expansion (tier 2): IL, TN.

Probe markets (low historical signal): IN, MO. Small budgets, watch quality.

Open question for this call: MA, PA, MD — high admit volume but not on the Saturday list. Decision needed on inclusion.

Insurance picture (where the high-value admits come from)

• BCBS-FL: 446 admits, 44% of total volume, $8,874 avg case. The dominant payer.
• United Healthcare: 92 admits at $14,352 avg.
• Out-of-state BCBS plans: BCBS-IL ($14,274), BCBS-PA ($12,634), BCBS-TX ($12,903), BCBS-MA ($23,886). All above $12k/case.
• BCBS-MA averages $23,886/case, the highest of any payer.

Campaign structure proposal — fewer campaigns, more algorithm leverage

Currently running 2 campaigns: Branded + General Rehab. Proposed next state: 3 total non-brand campaigns, max.

Why no more than 3 non-brand campaigns

• At ~$1,000/day total, 3 non-brand campaigns equals ~$333/day each. Adding more dilutes budget below the level Google's algorithm needs to learn.
• Within each campaign we still get program-level flexibility via ad groups. Ad groups don't get separate budgets, but they all optimize toward the same CPA target. With our SF Qualified Lead and Admit conversions now feeding the algorithm, ad-group-level optimization happens automatically.
• Every additional campaign added trades efficiency for control. We want enough control to push budget where it matters (Mental Health vs SUD vs Academy) without sacrificing the algorithm's signal-density advantage.

What this means for Ketamine, Outpatient, Detox

• Ketamine: open question. Doc may want it as a campaign. If so, it becomes the 4th non-brand campaign. Worth pressure-testing whether it's strategic or sentimental.
• Outpatient: Clement's instinct is dedicated campaign + LP, eventually. Not in phase 1. They do "a little outpatient" but no virtual IOP, so volume is limited.
• Detox: lives inside General Rehab / SUD. No separate campaign needed; ad group within General Rehab can handle detox-specific intent.

Timeline for the next 4-6 weeks

This week — current

• Daily negative-keyword cleanup as new search terms surface (Clement is actively doing this)
• Brand bid sweep underway: 9 competitors in auction insights, 26% absolute top impression share at current bid. Bumping max CPC $8 to $15 and reassessing by end of week.
• Launch Mental Health campaign (landing page ready). Targets largest program category by admit volume.
• Teen Academy campaign landing page also ready; launch decision depends on this call's alignment.

Next 2-3 weeks

• Backweight the budget to second half of June once foundation is solid
• Build dedicated landing pages for any additional campaigns greenlit on this call
• Plan out-of-state campaign structure based on Dr. Alam's preferences and this call's alignment
• Monitor signal quality from new conversion tracking; first real read on cost per qualified opp from clean attribution

3-6 weeks out (flexible launch window)

• Launch out-of-state campaign cluster targeting agreed states (NY, GA, NJ tier 1)
• First monthly retrospective with cost per qualified opp + cost per admit on the new tracking

What we need from Ambrosia to keep moving

• Ketamine campaign decision. Doc may want this as a 4th non-brand campaign. Is Ketamine therapy a strategic priority worth dedicated paid acquisition, or sentimental from the previous setup? Volume in Salesforce doesn't justify it as a top-3 priority, but happy to add if it's strategic.
• Confirmation on other services. Ibogaine, Spravato, TMS therapy — do you offer these? The previous setup was bidding heavily on them and we paused that, but we don't want to negate real demand if you do offer them. Virtual IOP confirmed off the menu (Dr. Alam + Natalia confirmed Saturday).
• "States above 90" clarification. Confirm what the 90 metric is.
• TX/SC reasoning. Revenue data looks fine; what's actually going wrong at the operational level?
• MA, PA, MD decision. High admit volume, not on his list. In or out?
• Insurance focus areas. Any payers you want us to lean into or away from in messaging?
• Landing page review cycle. Who is the right reviewer when we ship new LP drafts? How fast should we expect feedback?
• Adolescent program priority. The Academy has the highest case value of any program ($17,919 avg). Is dedicated adolescent acquisition a yes for this year, or is the existing referral pipeline enough?

For Josh and the admissions team

Most of our understanding of the Salesforce funnel and admissions process came from background conversations. Wanted to verify the picture and pick up operational signal that doesn't show in the data.

Funnel reality check

• Walk through the Missing Info → VOB on Queue → Follow Ups → Admitted path. Does the picture we have match how it works day-to-day?
• What pushes an opp from Missing Info to qualified? Manual update by an admissions rep, or automated based on field completion?
• About a third of paid-search opps get stuck at Missing Info. What's the typical reason a lead lands there and doesn't move?

Lead quality signal we can't see in data

• Do paid-search inbound calls feel different from referral or organic? Patterns?
• What does a best-case lead look like? Worst-case? What makes the difference at intake?
• Average time from first contact to insurance verification (VOB)?
• 9 GCLID-attributed Refer Outs in the last 90 days. Where are they typically going, and why? Anything we could be doing differently at the search-term level to filter those out?

Adolescent / Academy specifically

• How do parents typically find their way to the Academy? Same intake line, or separate path?
• $17,919 average case value on Academy admits in our 365d pull. Is that pattern holding currently?

Phone routing and after-hours

• The 888-492-1633 number. Any issues you've noticed with how calls route?
• Dr. Alam mentioned working 24/7. How does after-hours intake flow operationally? Any time-of-day patterns we should know about for bidding?

What would make Josh's job easier

• Anything you wish leads knew before they called?
• Anything we could add to landing page copy that would cut down the time you spend on disqualifying conversations?
• What's the one piece of context you wish was already in Salesforce by the time the call lands?

Open floor

• Anything you're hearing from admissions or sales that we should know about
• Census trajectory and capacity by program. Are some programs over-filling while others have empty beds?
• Anything from the previous reporting that you want clarified or rebuilt