800+ Qualified Sales Meetings in 9 Months for Sully.ai

The Challenge

Most outbound playbooks assume you can find a buyer, email them, and get a reply. Healthcare breaks that assumption in four places.

• Clinical staff do not live in their inbox. A medical director is on a ward, in theatre, or in clinic, not refreshing email between meetings. Cold email into clinical roles gets low engagement no matter how good the copy is. The channel does not match the working day.
• Gatekeeping is heavier than almost any other sector. Hospital switchboards are built to protect clinical time. Direct dials are rare, reception is trained, and getting through takes a specific skill set and a lot of attempts, not a better subject line.
• The buying committee is large and split. A clinical AI purchase can touch the medical director, the CMIO or clinical informatics lead, IT, procurement, compliance, information governance, and finance. Sell only to IT and you get benchmarked to death. Sell only to clinicians and you stall at procurement.
• Trust is the constraint, not interest. Every hospital leader has sat through a software pitch that promised to fix everything. Interest in AI is high; willingness to take a first meeting from a stranger is not.

What that meant in practice

Sully.ai needed volume, because the committee structure means a high number of first meetings are needed to produce a small number of deals. Quality, because a wasted meeting with the wrong title costs their team more than it costs ours. Reach into large accounts, because the economics of their product work best at scale. And a channel that actually reaches clinicians, not one that theoretically should.

And they needed it without hiring, training and managing an in-house SDR floor with healthcare-specific skills, which takes six months before it produces anything.

The Approach

We tested the channel before we scaled it

Most agencies pick a channel on day one based on what they are set up to run. A-Sales did not assume: cold calling and cold email ran in parallel against the same account list, the same personas, and the same offer. Same market, same week, same targets. The only variable was the channel.

The result was decisive. Calling produced the overwhelming majority of booked meetings; email produced a small fraction of the same volume against the same list. So the budget moved. Calling became the primary channel and got the headcount, the script iteration and the daily management. Email stayed on in a supporting role: follow-up, re-engaging accounts that had gone quiet, and accounts where a phone route did not exist.

We targeted clinical operations, not IT

The obvious target for an AI software product is the IT department. In healthcare that is usually the wrong door: IT will take the meeting, put you in a queue behind twelve other vendors, ask for a security review before there is any clinical sponsor, and never champion the purchase internally because it is not their pain. A-Sales built the target list around clinical operations instead: titles that own the workflow the product improves and carry weight in the room.

• Medical directors and clinical directors
• Directors of clinical informatics
• Quality improvement and patient safety leads
• Perioperative services directors
• Case management directors
• Programme directors and department heads
• Practice managers and operations directors
• Chief-level clinical and information roles where the account was small enough for direct access

Roughly half of every meeting booked went to a director-level title. The rest sat across heads of department, chiefs, vice presidents, presidents and managers with real operational ownership. IT and information roles stayed in the mix, but as a secondary persona, not the entry point.

We built the lists for accuracy, not volume

Healthcare data ages fast. Clinicians move, departments restructure, and titles are non-standard across systems. Every list was researched and validated before it went to the floor, and validated again with Sully.ai where the account mattered. Bad data on a calling floor is expensive twice over: it wastes the dial and it wastes the rep's confidence.

We moved up market deliberately

Early meetings covered a wide size range. As the data came in it was clear where the value sat, and the targeting weight shifted upward into large systems. Three quarters of meetings ended up with organisations of over 1,000 staff, and 42 were with health systems of over 10,000 staff: multi-site systems with real budget, real clinical admin burden, and a reason to buy at scale.

We opened a second region when the first one matured

Once US pipeline was running predictably, A-Sales opened Saudi Arabia and the wider Gulf. The logic was specific, not opportunistic: Gulf health systems are consolidated into clusters, so one conversation can cover many hospitals; national health transformation programmes have created active budget for clinical technology; and the market is significantly less saturated with outbound than the US. Both regions ran in parallel with separate scripts, separate calling windows, and separate qualification notes.

“We tested calling against email on the same list, in the same week, with the same offer. Calling won by a distance. So we moved the money.”

The Process

Phase 1: Foundations

• ICP definition sessions with Sully.ai
• Persona mapping across clinical, informatics and IT
• Lead research and list build for the first US segments
• Cold calling infrastructure and dialler setup
• Cold email infrastructure: dedicated sending domains, inbox provisioning, full warmup before any send
• Qualification criteria agreed in writing so both sides score a meeting the same way

Phase 2: Channel test

• Cold calling and cold email run against matched lists: same personas, same offer, same period
• Performance tracked per channel, per persona, per account size
• Decision taken to lead with calling

Phase 3: Scale the winner

• Dedicated calling floor assigned to the account
• Scripts rewritten monthly against live objection data
• Objection handling library built from real recorded calls
• Daily call volume targets, weekly quality review
• Email repositioned to a support role

Phase 4: Move up market

• Targeting weight shifted toward 1,000+ and 10,000+ staff systems
• Multi-site groups, academic medical centres and national provider networks prioritised
• Deeper account research per target, since the value per meeting is higher

Phase 5: Regional expansion

• Saudi Arabia and Qatar opened as a second region
• Separate scripts written for Gulf health clusters
• Calling windows adjusted to local working week and prayer times
• Meeting times booked in the lead's timezone, not the client's

Ongoing: iteration

• Weekly reporting on booked, held, and rejected meetings
• Rejected meetings fed straight back into targeting and scripts
• Continuous list refresh as titles move

How We Define a Qualified Meeting

A meeting only counts when it meets the criteria agreed with Sully.ai up front. Right company profile, right title, real interest, confirmed slot, and the prospect knows who they are meeting and why.

If a meeting does not meet the standard, it does not get billed. Sully.ai can reject a meeting, and rejected meetings are reviewed rather than argued about. That feedback is the fastest way to improve targeting, so A-Sales wants it.

This is the discipline that makes a performance-based model work. A-Sales earns the overwhelming majority of its revenue from meetings that actually happen and actually qualify. If we book badly, we do not get paid. There is no version of this where we win and the client does not.

“If the meeting does not meet the standard, it does not get billed. That is the whole model.”

Results

• 800+ qualified sales meetings booked in 9 months
• 724 booked by phone, proving calling beat email on this ICP
• 75% of meetings with organisations of 1,000+ staff
• 42 meetings with health systems of 10,000+ staff
• 93% show rate on tracked meetings
• Meetings with academic medical centres, multi-state hospital groups, national dialysis networks and government health clusters
• Two regions run in parallel, United States and the Gulf
• Engagement ongoing

The accounts we reached

Individual prospect accounts stay unnamed, because that is Sully.ai's pipeline. By category, meetings were booked across:

• Academic medical centres and university health systems
• Multi-state hospital operators
• Regional not-for-profit health systems
• National dialysis and specialty care networks
• Government health clusters and ministry-level health authorities in the Gulf
• Specialist hospitals and research centres
• Behavioural and mental health providers
• Urgent care, occupational health and physical therapy groups
• Multi-site dental and primary care groups

What Made the Difference

Test the channel, do not assume it

The entire result rests on a decision made in the first weeks. Email is the default for outbound agencies because it scales cheaply. On this ICP it was the wrong default. The test cost a few weeks and changed the outcome of the account.

Pick the persona that owns the pain, not the one that owns the software

Clinical operations titles took the meeting, understood the problem in one sentence, and carried it internally. IT-led entry produced slower, colder conversations.

Volume and quality are not opposites if the qualification bar is written down

Agreeing the standard in advance, in writing, removed every argument about what counts. It also gave the calling floor a clear target that was about the right meeting, not just a meeting.

Market selection beats message optimisation

Opening the Gulf added more volume than any script rewrite did. A less saturated market with consolidated buying is worth more than another round of A/B testing on a subject line.

Rejected meetings are the most valuable data in the account

Every rejection told us something about targeting. We asked for them.