---
title: "How we closed 12 sponsor studies worth $3,713,760 for a clinical research site network in eight months. | Stone Haven Capital Group"
description: "An average of $309,480 per study, from a standing start of $1,375,467. Eight months. A network of owned and affiliated trial sites in North America. No new business development hires, no conference booth, no waiting for a CRO to remember the network existed."
canonical: "https://stonehaven.capital/showcase/clinical-research-site-network"
last-updated: "2026-08-22"
---

> An average of $309,480 per study, from a standing start of $1,375,467. Eight months. A network of owned and affiliated trial sites in North America. No new business development hires, no conference booth, no waiting for a CRO to remember the network existed.

Stone Haven Capital Group

Schedule a Call

Case study · Clinical Research Site Network

# How we closed 12 sponsor studies worth $3,713,760 for a clinical research site network in eight months.

An average of $309,480 per study, from a standing start of $1,375,467. Eight months. A network of owned and affiliated trial sites in North America. No new business development hires, no conference booth, no waiting for a CRO to remember the network existed.

- $3,713,760 generated in 8 months, from 12 sponsored studies at an average of $309,480.

- 406,435 emails to 81,287 companies, producing 732 replies at 0.9%.

- 161 interested conversations became 50 held meetings and 12 signed studies.

- What did most of the work: we kept the network's own domain out of the programme completely.

- What went wrong: the sequence went to sponsors when the CRO controlled site selection on roughly two thirds of the list.

01

## Did they get a good result?

They did, and the figures are below with nothing smoothed over. Before the programme started, study awards came from sponsors the medical director already knew and from CROs who had used the sites before, which produced $1,375,467 of study revenue across the comparable prior period. Eight months later the network had 12 studies signed worth $3,713,760, an average of $309,480 each, and 2.7 times what the relationship route had produced on its own.

What happenedThe number

Studies signed in eight months12, worth $3,713,760

Average value per study$309,480

Growth against the $1,375,467 baseline2.7x

Qualified opportunities behind the closes50, worth $11,762,500

Interested conversations at an earlier stage161, worth $25,438,000

Emails sent406,435

Contacts reached81,287

Replies732 (0.9%)

Positive replies161 (22.0% of replies)

Bounces6,503 (1.6%)

The eight-month programme as it ran in the sending platform, with the sponsor and CRO segments split.

02

## Did you get them in front of the right people?

This is the only question worth asking of an outbound programme, and a send total answers none of it. Four hundred thousand emails to people with no study to place is four hundred thousand pieces of nothing.

12 studies were signed carrying $3,713,760 between them, at an average of $309,480. Behind those sit 50 qualified opportunities worth $11,762,500 where a protocol and a start-up window were both confirmed, and 161 interested conversations worth $25,438,000 at an earlier stage.

Targeting shows itself in the conversion and never in the volume. 732 replies produced 161 positive ones, a 22.0% positive rate, and 50 of those cleared qualification. Close to a third of every positive reply became a real conversation about placing a study, which does not happen if you are reaching people who do not choose sites.

The study pipeline in the client's CRM, from first interest through to signed.

03

## What did not work, and what we did about it

Three failures, and they come first here because of what they have in common. Not one of them was a problem with the message. All three were about who was receiving it. In this market the copy is the easy part and the routing is the whole difficulty, so putting the corrections up front is the most honest way to describe how the programme actually got built.

The sequence went to sponsors when the CRO controlled site selection on roughly two thirds of the list.

The programme opened by targeting sponsors directly, on the assumption that the company running the trial chooses the sites. For small biotechs that is true. For everyone who has outsourced the trial, site selection sits with the CRO, and the sponsor contact receiving our email had no authority to place anything with anyone. It surfaced in the replies rather than in any list report. Positive answers came back warm and then died at the same point every time, on a variation of the sentence that the CRO handles site selection. Six weeks in, the pattern was clear enough to count, and it held on roughly two thirds of the list.

What we changed: the list split into two segments with separate sequences. Sponsor contacts kept the enrolment claim. CRO contacts got a sequence written for a different job, because a study start-up lead at a CRO is judged on filling sites against a timeline they did not set. Same network, same evidence, two entirely different reasons to reply.

A reply thread from the CRO segment, from first answer through to a scheduled call.

### The therapeutic-area token was generic enough to read as a mail merge.

The opener carried a therapeutic-area token, which looked like personalisation in the sequence builder and was not. The wording around it was written to fit any indication, so an oncology sponsor and a cardiology sponsor received the same sentence with one word swapped. A clinical operations director recognises that instantly, and several replies said so directly.

What we changed: the token was replaced by four written variants, one per therapeutic cluster, each naming the enrolment problem specific to that cluster. Screening burden in oncology reads nothing like comorbidity exclusions in cardiology. The volume dropped slightly because fewer contacts fit a written variant cleanly, and the positive rate went up enough to make that trade obvious.

### The first cadence assumed a single decision maker who does not exist in this market.

The opener was built to reach one person, get one answer and move to a call. Site selection is not made that way. A feasibility committee meets on its own schedule, reviews site packets in a batch, and the person who replied to us was one voice in a room that had not yet met. A cadence that ended before that meeting ended before the decision.

What we changed: the cadence was rebuilt around the committee rather than the contact. The gaps widened at the back so the sequence stayed live across a meeting cycle, and the follow-ups gave the contact something to bring into the room instead of asking again for a call. The 14 day gaps in the final cadence come directly from this correction.

No programme of this length runs clean, and a case study with nothing in this section has had it taken out. The corrections are the useful part, because they are what recurs on the next engagement.

04

## What worked, and why

### We kept the network's own domain out of the programme completely.

A site network sends investigator correspondence, regulatory documents and patient-facing material from its primary domain. That domain cannot be put at risk to run a cold campaign, so nothing cold ever left it. We stood up 77 dedicated sending domains carrying 230 mailboxes, each authenticated on SPF, DKIM, DMARC and MX before it sent a single message. Warmup ran on all 230 for the full eight months while only part of the estate carried live volume at a time, with the rest held in reserve and rotated in as the older domains aged.

The estate moved 406,435 sends and finished at a 1.6% bounce rate, 6,503 bounces in total, with inbox placement holding at 93.0% across the programme. Volume was held at 20 messages per inbox per day, which is the ceiling that keeps a mailbox looking like a person. Six alerts were raised across eight months and none of them became an incident.

The sending estate for the network: 77 domains, 230 mailboxes, all authenticated, none of it on the network's own domain.

What that buys is separation. The domain the sites use for investigator and regulatory correspondence was never exposed, and every deliverability risk in the programme sat on infrastructure built to be replaced.

We built the list around who actually places a study, and cut everyone who only influences it.

The people who choose sites are clinical operations directors, VPs of clinical development and study start-up leads at biotech sponsors and mid-size CROs. The first pull on that profile returned 270,956 contacts. Applying therapeutic area, trial phase, active pipeline stage and the exclusions took it to 93,433, and 81,287 of those were contacted once verification and deduplication had run.

The removals did more work than the inclusions. Preclinical companies with nothing to place for two years, device firms outside the network's capability, academic centres running their own sites, and the large global CROs with captive site networks were removed outright. Each of those groups replies at a normal rate and none of them can award a study.

The reason that works: a therapeutic area tells you who might use the network one day. A protocol entering start-up tells you who is choosing sites this quarter, and only the second group has anything to say to a site network.

### We put the enrolment number in the first line and nothing in front of it.

A sponsor reading a site network's email is looking for one thing, which is evidence that these sites enrol faster than the ones already failing them. The opener named the network's enrolment performance as a figure, named the therapeutic area the sponsor works in, and asked whether it was worth a look. No description of the network, no list of sites, no capability paragraph.

Every message went out as plain text with spintax through every line, so no two emails left the estate identical, and the tokens sat inside sentences rather than at the start of them.

The five-touch sequence and the opener that carried the programme.

The opener, built on a single enrolment number.

The reason it holds: a clinical operations director gives the message one line before deciding. Spending that line on who the network is wastes it, because the only thing that changes their week is a site that fills.

### We ran a five-touch cadence and let only start-up-ready conversations through to the network.

Five touches on gaps of 7, 7, 14 and 14 days, four variations on the opener and two on each follow-up, all A/B tested. The wider gaps sat at the back of the sequence deliberately, because a sponsor who is not selecting sites in March may well be selecting them in May, and a message arriving two weeks later catches that turn.

A qualification layer sat between the inbox and the business development team, passing through only conversations where a protocol, a therapeutic area within the network's capability and a start-up window were all present. Of 732 replies, 571 stopped there. The 161 that came through are the only ones anyone at the network spent time on.

### We reviewed the angles weekly and cut the losing ones inside days.

Angles were judged on how many conversations reached a named protocol, never on how many replies they produced. Anything generating polite interest that stalled before a study came off within a week and its share of the list went to the survivors. Across eight months that is roughly thirty reviews, and the compounding across them is most of the result.

05

## The six angles we test, in every market

Which message a market responds to is not something anyone works out in advance, and confidence about it before launch is the most expensive habit in outbound. So six variations of the opener go out at once, one per angle, and the reply data settles the question. These six travel across sectors because each rests on a different reason a person answers an email rather than on anything about what is being sold.

06

## 1. The named problem

State the constraint they live inside using the words they would use internally, then ask whether it applies to them right now. Strongest where the problem is well known and nobody says it out loud in a first conversation.

Most of the sponsors we talk to are carrying two sites that have not randomised anyone since activation. Is that where this study is?

07

## 2. The benchmark

Give them a number for how their peers perform and let them place themselves against it. It works because the recipient checks the comparison before they decide whether to reply, and by then they are engaged.

Median time to first patient in this indication is running near four months across the sites we see. Where did yours land?

08

## 3. Value first

Open with something useful whether or not they ever answer. A read on the market, a comparison, a piece of data they would otherwise have to assemble themselves. Usefulness first and interest second.

We pulled enrolment timelines on the last nine trials in this indication at your phase. Want them?

09

## 4. The soft ask

Lower the bar until agreeing costs nothing at all. No meeting, no commitment, only permission to send something across. It converts the recipient who is interested but has no time this week.

Would it be useful to see our enrolment record in this indication? Happy to send it across, no call needed.

10

## 5. The timing hook

Attach the ask to something that has just changed for them. It only works when the change is public and recent enough that mentioning it proves you looked rather than guessed.

Saw the phase II registration went up last month. Is site selection still open on it?

11

## 6. Referral sideways

Ask who owns the decision instead of asking for the decision. The recipient who cannot help you can usually name the person who can, and answering costs them one line. It also gets you a warm internal introduction rather than a cold second attempt.

You are probably not the one choosing sites for this. Who runs feasibility on your side?

12

## How we use them

All six go out together across a split list, matched on size and segment so the comparison is fair. Within a few weeks the reply data has named the one or two the market wants and the rest come off before they consume more contacts. Which of them wins is unknowable in advance. That one or two will win is close to certain, and running six at once is what converts a guess into a result. The same six port to LinkedIn with tighter wording and the same underlying logic.

13

## Why this works for clinical research sites specifically

Every site network that comes to us describes the same bottleneck in almost the same words. The sites perform, the enrolment record is genuinely good, and none of that reaches a sponsor who has never heard of them. Study awards arrive through people the medical director already knows, which is a real channel with a hard ceiling on it.

Three things make outbound work unusually well in this market:

- The buyer has a problem they cannot hide. Enrolment timelines slip publicly, trial registries show recruitment status, and a sponsor whose study is behind knows it before you tell them. That makes the opening question easy to ask and difficult to ignore.

- One study covers a large share of the programme. At an average of $309,480 per study, a handful of awards pays for the entire engagement several times over. Very few channels have arithmetic that forgiving.

- The ask costs them nothing to answer. Whether a protocol is entering start-up is a question the recipient can answer in four words, and the ones who answer it are the ones worth talking to.

14

## Why would this work for your business?

Possibly it will not, and establishing that now costs neither of us anything. Read down the table.

This works ifThis does not work if

A study, contract or account is worth six figures to youYour average order value is small and the volume has to be enormous

You can name the job titles that decide, and there are thousands of themYour buyer is a handful of named accounts you already know personally

You have one claim your buyer measures themselves on, with a number behind itYour advantage is real but only becomes visible after a long evaluation

Someone can answer a qualified conversation within a dayReplies would sit unanswered for a week while people are travelling

You can commit to a programme across a full buying cycleYou need the pipeline to arrive inside the first six weeks

If the left column describes you, what carries across is the method rather than anything particular to this network. Good sites, a real enrolment record, and no route to a sponsor outside the medical director's address book. All of it was produced by a process, and that process arrives at your engagement unchanged.

15

## One more thing worth understanding

In clinical research there is exactly one claim a sponsor will read, and it is enrolment speed. Everything else in a site network's pitch is treated as noise, and understanding why saves you a quarter of guessing.

The reason is that a sponsor's cost of delay is enormous and entirely predictable. A study that runs long burns budget on a fixed monthly rate, pushes the readout past a funding milestone, and does it for reasons the sponsor cannot control from a distance. The overwhelming cause is that sites cannot find patients. So a site network that can show it enrols faster is answering the only question the sponsor is actually being measured on.

What happens to every other message is worth stating plainly. Quality systems, monitoring performance, staff experience, geographic coverage and technology all get read as table stakes, because a sponsor assumes any site they would consider already has them. Lead with any of those and the message reads as a capability brochure, which is filed rather than answered. We ran that version in the first fortnight and the replies were courteous and empty.

So the enrolment claim goes in the first line with a number attached, and everything else in the network's story waits until someone has asked. That ordering is not a stylistic preference. It is the difference between a message that gets a reply and one that gets an acknowledgement.

16

## Before and after

Before After eight months

Study revenue$1,375,467 $3,713,760

Studies signedWhatever the medical director's contacts produced 12, at an average of $309,480

Route to a new sponsorAn introduction, or nothing 81,287 contacts reached directly

Qualified opportunities in viewNot tracked as a stage 50, worth $11,762,500

ForecastingGuesswork against last year A pipeline with counts and values at every stage

17

## If you want to know whether your market has this in it

A quarter of an hour usually settles it. Tell us which sponsors and CROs you want placing studies with you, and what a study is worth to the network when it lands. We come back with how many of them are genuinely reachable, the conversation volume that is realistic against that number, and a direct answer on whether this channel suits you.

If the answer is no, you will hear it on that call. Neither of us gains from discovering it in month four.

## Want to know whether your market has this in it?

The first conversation is short. You tell us who your buyers are and what one is worth to you. We tell you how many we can actually reach, what the meeting volume looks like, and whether outbound is the right lever for you at all.

If we think it is not, we will say so.

Book a consultation call

→

## Read another one

[Commercial Solar & Storage EPC](https://stonehaven.capital/showcase/commercial-solar-storage-epc/)[How we closed $7,650,000 of solar and storage projects for a commercial EPC in nine months.](https://stonehaven.capital/showcase/commercial-solar-storage-epc/)[Compliance Automation Platform](https://stonehaven.capital/showcase/compliance-automation-platform/)[How we closed 105 compliance automation accounts worth $4,401,390 in seven months.](https://stonehaven.capital/showcase/compliance-automation-platform/)[All case studies](https://stonehaven.capital/case-studies/)