The Tactical Guide to GTM Automation

ShiftOS / Holly — multi-channel outbound for healthcare & hospitality workforce buyers

This is the playbook for getting Holly in front of the people who can actually buy her: pharmacy directors, DONs, ops leaders, and multi-site operators in healthcare and hospitality. It assumes you already have product-market signal from ShiftRx and are now scaling Holly as a standalone SaaS motion. Everything here is tactical — no positioning theory, no deck reviews. If a step doesn't move a meeting onto Samantha's calendar, it isn't in here.

The motion, in one paragraph

Holly is mid-market B2B SaaS sold into regulated industries. Cycles are 30–90 days, deals are multi-stakeholder (operator + finance + sometimes IT/compliance), and pilot-before-rollout is the norm. That means cold outbound alone won't carry it — you need a layered motion: (1) warm wedge from the ShiftRx installed base, (2) cold outbound on LinkedIn + email to look-alike accounts, (3) event presence at the 3–4 conferences that actually matter, and (4) a partner channel through scheduling-adjacent vendors. The automation stack underneath all four layers is the same: Clay for list building, HeyReach for LinkedIn, Smartlead for email, HubSpot (or Attio) for the CRM, and a thin agent layer that routes replies to the right human within minutes.

The four layers, in priority order

Layer 1 — The ShiftRx warm wedge (highest ROI, do first)

ShiftRx already has relationships with pharmacy facilities buying PRN shifts. Every one of those facilities has a scheduling problem — that's literally why they're using ShiftRx. Holly's first 20 customers should come from this list, not from cold outbound. The economics: warm intros from an existing vendor relationship close at 20–30%, cold outbound at 1–2%. You'd need 20x the volume to match it.

Tactical steps:

Pull every ShiftRx facility account from the ShiftRx CRM. Segment by: number of shifts booked in the last 90 days (proxy for scheduling pain), number of locations, and whether they have a multi-site ops contact. The top quintile by shift volume is your Tier 1.
Build a one-page internal brief on Holly tailored to the pharmacy use case (license tracking, PRN coverage gaps, call-off automation). The ShiftRx CSM/sales rep needs to be able to pitch it in 90 seconds on a regular check-in.
Add a Holly mention to every existing ShiftRx QBR and renewal call. Soft intro, not a hard pitch — "by the way, the team that built ShiftRx also built Holly, an AI scheduler — want a 15-min look?"
Track conversion in HubSpot with a source = shiftrx_warm field. This becomes the highest-converting cohort and you'll want the data to defend the channel internally.

This layer doesn't need automation tooling. It needs discipline: every ShiftRx customer touch is also a Holly touch, every week.

Layer 2 — Cold outbound to look-alikes (HeyReach + Smartlead)

Once Layer 1 is moving, build the cold motion against accounts that look like your warm wins. This is where the automation stack lives. Detailed below.

Layer 3 — Events (3 per year, no more)

Healthcare ops people buy at conferences. Pick three:

ASHP Midyear (December) — pharmacy directors, the core ICP. Non-negotiable.
HIMSS (March) — health system IT and ops, where the bigger deals start.
NCPA Annual (October) — independent and regional pharmacy chains, perfect Holly profile.

For hospitality: HITEC (June) for hotel groups, NRA Show (May) for restaurant operators. Pick one, not both, until hospitality is a proven segment.

Tactical: don't take a booth in year one. Walk the floor with a list of 50 pre-booked 15-min coffees set up via Clay + Smartlead 6 weeks ahead of the show. A booth costs $40K and gets you 30 leads. 50 pre-booked coffees costs$ 0 and gets you 50 qualified conversations.

Layer 4 — Partnerships (slow burn, asymmetric upside)

Holly slots next to but doesn't replace existing pharmacy software (PioneerRx, Liberty, BestRx) and HRIS systems (Paycor, Paylocity, Gusto). Each of those vendors has a partner program. One distribution deal with a pharmacy software vendor that has 2,000 customers can outperform 12 months of outbound. Don't run this in Q1 — it takes 6+ months to close — but get the first conversation started in month one.

The cold outbound stack (Layer 2 in detail)

The tool stack and what each costs

Tool	Job	Cost
Clay	List building, enrichment, AI personalization	$349/mo (Explorer plan, scales with volume)
HeyReach	LinkedIn outbound automation	~ $79–$ 200/mo per sender seat
Smartlead	Email outbound + warmup + reply detection	~$94/mo (50 mailbox plan)
HubSpot Sales Hub	CRM, dedupe, reporting	$90/seat/mo (Starter is fine for now)
Apollo (or Cognism)	Contact data + verified emails	$99/mo
Calendly	Booking	already paid
Slack	Reply routing channel	already paid

All-in: ~$1,000/mo for the stack. One closed Holly deal pays for it for the year.

Sender infrastructure (do this first, takes calendar time)

You cannot send cold email from @shiftos.com or @shiftrx.io without burning your primary domain reputation. Buy 3–5 lookalike domains and send from those:

getshiftos.com
tryshiftos.com
shiftos.team
meetholly.ai
holly-shiftos.com

For each:

Register through Cloudflare or Porkbun (cheaper than GoDaddy, cleaner DNS).
Set up Google Workspace ($6/user/mo) — one mailbox per domain, two if you want to push volume. Real names: samantha@, noah@, team@. Never info@, sales@, hello@ — those get filtered.
Configure SPF, DKIM, and DMARC on each domain. (See the appendix below for the exact records — same as the editorbunny.com setup, just with shiftos parameters.)
Set up a forwarding rule: any reply to a lookalike domain forwards to a shared replies@shiftos.com inbox so nothing gets lost.
Connect every mailbox to Smartlead. Turn on warmup immediately and let it run for 14 days minimum before sending a real cold email. Settings: ramp from 5 to 40 emails/day over 14 days, 30% reply rate, weekday-only.

You're shooting for an aggregate sending capacity of ~150–200 cold emails/day across all mailboxes once warmed. That's enough to work a 3,000-account list in ~6 weeks.

Email authentication (the appendix, but critical)

For each lookalike domain, in Cloudflare/Porkbun DNS:

SPF — TXT record at root:

Host: @
Value: v=spf1 include:_spf.google.com ~all
Only one SPF per domain. If one already exists, edit it instead of duplicating.

DKIM — generate inside Google Workspace first:

admin.google.com → Apps → Google Workspace → Gmail → Authenticate email
Select the domain → Generate new record → 2048-bit
Copy the host (e.g. google._domainkey) and the long v=DKIM1; k=rsa; p=... value
Add as a TXT record in DNS, wait an hour
Go back to that Google page and click Start authentication. This is the step everyone forgets — until you click it, DKIM is configured but inactive.

DMARC — TXT record:

Host: _dmarc
Value: v=DMARC1; p=none; rua=mailto:postmaster@shiftos.com
Start at p=none (monitor only). After 3 weeks of clean sending, tighten to p=quarantine.

Verify each domain by sending to a personal Gmail → Show original → confirm SPF, DKIM, DMARC all PASS. Or run each domain through mail-tester.com and don't launch until you score 9/10 or higher.

ICP and list building (Clay)

Healthcare ICP — Tier 1

The fastest segment to close is multi-site outpatient pharmacy and long-term care, because they have enough scheduling pain to feel it daily but not so much bureaucracy that procurement takes 6 months.

Account filters in Clay:

Industry: Retail Pharmacy, Long-Term Care Pharmacy, Specialty Pharmacy, Compounding Pharmacy, Surgery Centers, Urgent Care, Dental Service Organizations
Size: 3–50 locations (under 3 = no budget, over 50 = enterprise procurement hell)
Headcount: 50–1,000 employees
Geography: US only, prioritize states with strict labor law (CA, NY, IL, WA, MA — they feel compliance pain hardest)
Tech stack signal: currently using a legacy scheduler (When I Work, Deputy, Homebase, paper, Excel) — Clay can pull this from BuiltWith and job postings

Title filters for the contact:

Pharmacy Director, Director of Pharmacy, Pharmacy Operations Manager
Director of Operations, COO, VP Operations
Director of Nursing (DON), Chief Nursing Officer
Scheduling Manager, Workforce Manager
Owner / Principal (for independents)

Avoid: HR generalists, recruiters, IT directors. They're not the buyer — they're the blocker.

Hospitality ICP — Tier 2

Don't run this until healthcare is producing pipeline reliably. When you do:

Multi-unit restaurant groups (5–50 locations)
Boutique hotel groups (3–20 properties)
Coffee chains, fast-casual concepts

Targets: VP Operations, Director of Restaurant/Hotel Ops, GM (for smaller groups).

Clay enrichment columns (the difference between 1% and 12% reply rate)

In every Clay table, add these columns before exporting to Smartlead/HeyReach:

Verified email — Clay waterfall (Findymail → Datagma → ZeroBounce). Drop anything that isn't Valid.
LinkedIn URL — required for HeyReach.
Number of locations — pulled from website or Google Maps API, used as a personalization variable.
Recent news — funding, acquisitions, new location openings (Clay's Google News integration). This becomes a personalized opener.
Current scheduling tool (if detectable) — turns into a "ripping out X" angle.
One-sentence company description — generated by Clay's Claude integration from the company website. Used as the personalized hook.
Pain trigger — a Clay AI prompt that scans recent Glassdoor reviews for the company looking for words like "scheduling," "shifts," "called off," "overworked." If it finds them, that's gold for the opener.

The personalization rule: never send a message where the first sentence could have been sent to anyone else on the list. Use Clay's AI columns to compute a unique opener for every lead, stored as {{personalized_opener}} in Smartlead.

Example pre-computed openers:

"Saw Mercy Pharmacy just opened your fourth location in Sacramento — congrats on the expansion. Curious how you're handling scheduling across all four now."
"Noticed three Glassdoor reviews from your techs in the last quarter mentioning short-staffing and last-minute call-ins. That's the exact problem Holly was built for."
"You've been on When I Work for ~3 years from what I can tell — most pharmacy directors we talk to outgrow it around the 4-location mark. Worth a 15-min look at what's next?"

The HeyReach sequence (LinkedIn)

Sender setup

You need at least 2 LinkedIn senders to hit volume:

Samantha (or whoever owns AE-level conversations) — primary
A second team member with a real, healthcare-credible profile — secondary

Both profiles need to look like operators, not marketers. Headlines should be specific: "Helping pharmacy operators automate scheduling with Holly | ShiftOS" — not "Growth at ShiftOS." Pin one demo video and one customer logo wall to the featured section.

Sending limits per profile:

Week 1: 15 connection requests/day
Week 2: 25/day
Week 3+: 50/day, max 100/day

Going faster gets profiles restricted. A restricted profile is dead for 30+ days. Don't do it.

The sequence skeleton

Day 0  → Connection request + note (<300 chars)
         "Hi {first} — saw {personalized_trigger}. Building an AI scheduler 
         for pharmacy ops, would love to connect."

Day 2  → If accepted: warm intro message
         "Thanks for connecting. Quick context — Holly handles call-offs, 
         PTO, swaps, and license tracking for multi-site pharmacy ops. 
         Open to a 15-min look?"

Day 6  → If no reply: value drop
         Link to a 90-second demo video of Holly handling a shift swap.
         "Easier to show than tell — this is what Holly does in 90 seconds."

Day 11 → If no reply: case study
         "We just helped {peer_company} cut 18 hrs/week of scheduling work. 
         Same setup as yours. Worth a quick look?"

Day 17 → If no reply: breakup
         "Last note — should I close the loop, or is there a better time 
         to circle back?"

Day 30 → If no reply: warm touch (no message)
         Profile view + post like

On reply (any positive signal): HeyReach pauses the sequence automatically and webhook-pings #linkedin-replies in Slack. A human (Samantha or AE) responds within 30 minutes during business hours. Speed-to-reply is the single biggest lever on conversion — go from 4 hours to 30 minutes and you'll roughly double meetings booked.

The Smartlead sequence (Email)

Email runs in parallel with HeyReach against the same list, with deliberate timing offset.

The orchestration rule

LinkedIn connection request fires Day 0. First email fires Day 2. Reason: by the time the email lands, your name is already familiar from the LinkedIn notification — open rate jumps 15–25%.

The sequence skeleton

Day 2  → Email #1 — short, personalized, single CTA
         Subject: "{first}, quick Q on {company}'s scheduling"
         Body: 4 sentences. Personalized opener → problem statement → 
         one-line Holly pitch → ask for 15 min.

Day 5  → Email #2 — different angle (peer proof)
         Subject: "how {peer_company} cut scheduling time 60%"
         Body: case study, 5 sentences, same CTA.

Day 9  → Email #3 — concrete offer
         Subject: "free schedule audit?"
         Body: offer to run a free audit on their current scheduling 
         workflow and show where Holly would save time. Low-friction ask.

Day 14 → Email #4 — breakup
         Subject: "should I close the loop?"
         Body: 2 sentences. "Haven't heard back — should I check in 
         next quarter, or is now not the right time?"

Reply branch: pause sequence, route to Slack

Smartlead config that matters

Volume: 30/day per mailbox during steady state. With 5 mailboxes that's 150/day = ~750/week of cold sends.
Sending hours: 8am–5pm in the recipient's timezone (Smartlead supports this — turn it on).
Stop on reply: yes, across all sequences for that lead.
Stop on auto-reply (OOO): yes, resume after 7 days.
Lead categories: Smartlead auto-classifies replies into Interested / Not Interested / OOO / Wrong Person / Do Not Contact. Triage in Slack.
Webhook: EMAIL_REPLY and LEAD_CATEGORY_UPDATED → #email-replies Slack channel. Same SLA as LinkedIn: 30 min response during business hours.

Subject line rules for healthcare buyers

Lowercase first letter (looks like a real person, not a marketing blast)
Question marks beat statements
Keep it under 50 characters
Reference their company name when possible
Never use: "free," "guaranteed," "limited time," "act now," exclamation marks. Healthcare spam filters are aggressive on these.

CRM and reply routing

Why you need a CRM on day one (unlike Editor Bunny)

For a single founder running outbound to 1,000 leads, Slack channels are enough. ShiftOS is different: you have multiple senders, multi-stakeholder deals, 30–90 day cycles, warm + cold + event + partner sources merging together, and you'll need clean attribution to know what's working. Start with a CRM from day one.

HubSpot Sales Hub Starter is the right call. $90/seat/mo. Native Smartlead and HeyReach integrations exist (Smartlead has a direct one, HeyReach pipes through Zapier/n8n). Avoid Salesforce — too heavy for this stage. Avoid Attio if you don't have an in-house engineer to maintain the schema.

Object model (keep it simple)

Companies — one per facility group, deduped on domain
Contacts — one per person, deduped on email + LinkedIn URL
Deals — one per opportunity, with a custom field source = shiftrx_warm | linkedin_outbound | email_outbound | event | partner | inbound
Activities — every email send, open, reply, LinkedIn touch, call, demo

Pipeline stages

New (Clay-sourced, not yet contacted)
Contacted (sequence active)
Engaged (replied, qualifying)
Demo Scheduled
Demo Completed
Pilot (30-day)
Negotiation
Closed Won / Closed Lost

Track conversion rate stage-to-stage weekly. The leak point is almost always 4→5 (no-shows on demos) or 6→7 (pilots that don't convert). Different fixes for each.

Reply routing flow

Reply hits HeyReach or Smartlead
Webhook fires → Zapier/n8n
Zapier creates/updates HubSpot contact + activity
Zapier posts to #replies Slack channel with: contact name, company, channel, the actual reply text, and a one-click link to the HubSpot record
AE has 30 minutes to respond
Once replied, AE manually moves the deal to "Engaged" stage in HubSpot

This is a 2-hour Zapier build. Do it before launch, not after — retrofitting attribution after you've sent 2,000 emails is miserable.

What "working" looks like — the metrics

Healthcare B2B benchmarks, calibrated for ~750 cold sends/week and 100 LinkedIn connects/week per sender:

Metric	Bad	Target	Good
LinkedIn connect accept rate	<25%	35–45%	>50%
LinkedIn reply rate (of accepts)	<8%	12–18%	>20%
Email open rate	<40%	55–65%	>70%
Email reply rate	<4%	6–10%	>12%
Positive reply rate (combined)	<1%	2–3%	>4%
Demos booked / positive replies	<40%	60%	>75%
Demo show rate	<60%	75%	>85%
Demo → Pilot conversion	<20%	30–40%	>50%
Pilot → Closed Won	<40%	50–60%	>70%

On a 3,000-account cold list over 8 weeks: ~60–90 positive replies → ~45 demos booked → ~35 demos completed → ~12 pilots → ~6 closed deals. At a $3K/mo ACV that's$ 216K of new ARR from one cold cycle. The warm wedge from ShiftRx should produce another 6–10 deals on top of that with way less effort.

Reading the numbers

Connect rate low → LinkedIn profile or connection note is broken. Fix profile first, message second.
Email opens fine, replies low → message body is generic. Rewrite with sharper personalization.
Opens low → deliverability problem. Run mxtoolbox + mail-tester, check warmup status.
Demos no-showing → calendar friction. Cut to 15 min, add SMS reminders via Calendly, send a "looking forward" note 1 hr before.
Pilots not converting → product, success, or pricing issue, not a GTM issue. Different conversation.

Weekly operating cadence

Monday — 60 min, full team

Pull last week's metrics from HubSpot dashboards
Review every reply in Slack channels — anything not yet responded to
Review pipeline movement stage-by-stage
Identify the bottleneck stage and pick one fix to test this week

Wednesday — 30 min, growth lead

A/B test review in Smartlead and HeyReach
Kill the worst variant, double the best
Check warmup status on all mailboxes
Spot-check Clay enrichment quality on this week's new leads

Friday — 30 min, growth lead

Run mail-tester.com on each sending domain
Confirm no mailboxes are throttled or blacklisted
Pull next week's lead list from Clay → Smartlead + HeyReach
Send pipeline summary to leadership

Monthly — 90 min, leadership

Source attribution review (warm vs cold vs event vs partner)
CAC and CAC payback by source
Decide which sources to lean into / cut
Decide whether to add a sender, add a domain, or add a sequence

Two-week launch plan

Week 1 — Infrastructure

Mon: Buy 5 lookalike domains, set up Google Workspace, start DNS records
Tue: SPF/DKIM/DMARC on all domains, click "Start authentication" in Google Admin
Wed: Connect mailboxes to Smartlead, start warmup. Set up HubSpot, build pipeline stages, install Smartlead and Zapier integrations
Thu: Clay account, build first ICP filter for Tier 1 healthcare
Fri: HeyReach account, connect 2 LinkedIn profiles, profile cleanup pass

Week 2 — Build and stage

Mon: Clay enrichment columns built, first 500 leads cleaned
Tue: HeyReach sequence skeleton built (no copy yet), first 100 leads imported
Wed: Smartlead sequences built (no copy yet), same 100 leads imported
Thu: Copy session — write all sequence copy with a real human who knows the buyer
Fri: Slack reply routing live, internal test send to team Gmail accounts, fix anything that doesn't route correctly

Week 3 — Soft launch

Mon: First 50 leads enter LinkedIn sequence
Wed: First 50 leads enter email sequence
Fri: Review reply quality, message tone, deliverability. Adjust before scaling.

Week 4 — Scale

Open the floodgates: 100/day LinkedIn, 150/day email
Weekly cadence kicks in

By end of Month 2: first demos. By end of Month 3: first closed deals from the cold motion. The warm wedge from ShiftRx should produce demos in Week 1 of the program — that's why you do it first.

What's deliberately not in this guide

AI-generated reply drafting. Not yet. The first 200 conversations need to be handled by humans so you learn what the actual objections are. Once the playbook is written, then automate the easy categories (OOO, wrong person, "send more info").
Paid ads. Not the right channel for this buyer. Healthcare ops people don't click LinkedIn ads. Save the budget.
A second ICP segment. Hospitality waits until healthcare is producing pipeline reliably. Don't split focus.
A booth at HIMSS in year one. Walk the floor with pre-booked coffees instead. Booth comes in year two.
Building your own outbound tool. Smartlead and HeyReach exist. Don't waste eng cycles.

The bet, in one sentence

Holly wins because she's a category-of-one product (the only AI scheduler that actually handles the full loop end-to-end), and ShiftOS already has a warm wedge nobody else has (the ShiftRx installed base). The job of GTM is to compound those two advantages — not to invent a new motion. This guide is just the operating manual for compounding them.