Case study 04 — Clinical research network · Centricity Research

Multi-site clinical research network, North America

A digital marketing transformation: 84% lower recruitment cost per lead, without adding risk.

Centricity Research was already running paid campaigns to recruit trial participants. The challenge was not getting attention — it was making the economics predictable. With cost per lead ranging from $79 to $900, every new campaign raised the same question: what is actually working, and where should the next dollar go? We set out to answer it.

84%

Reduction in cost per lead, on the same channels and the same audience pool.

$15 – $69

The new cost-per-lead range, down from a $79 – $900 swing.

~3x

Monthly ad spend scaled, $50K to $147K, without CPL drifting back.

Weekly

Cadence of performance audits and budget reallocation, not quarterly.

Context

The ads were not the problem. The uncertainty was.

Centricity had traffic, leads and several active channels. What was missing was a reliable way to separate the campaigns, audiences and creative generating quality results from those quietly consuming budget.

So instead of increasing spend and hoping performance would follow, we went in the opposite direction.

Organization

Centricity Research, multi-site clinical research network

Scope

Paid patient recruitment across Meta, Google and TikTok

Problem

Cost per lead ranging from $79 to $900 with no reliable read on why

Engagement

Funnel diagnosis, structured testing, attribution build, ongoing optimization

Our approach

Test first. Scale second. Not the other way around.

The common first move is to increase budget and hope volume smooths out the variance. We did the opposite. Before touching spend, we needed to know exactly which parts of the funnel were working — so scaling would not just mean scaling the problem. Four things ran continuously, not as a one-time audit.

Refined targeting

Audiences on Meta and TikTok narrowed against the segments that were actually converting, not the ones generating the most impressions.

Creative testing

Conversion-focused testing on a continuous cycle, because a scroll-stopping ad and a lead-generating ad are not automatically the same ad.

Weekly audits

Cost per lead reviewed every week rather than every quarter. A campaign bleeding money for eleven weeks has already cost eleven weeks.

Dynamic reallocation

Budget moved toward what was working in near real time, instead of waiting on a monthly report nobody could act on for another month.

The breakthrough

Cost per lead: $79 – $900 → $15 – $69.

Before this engagement, cost per lead swung between $79 and $900. After, it settled into a tight $15 to $69 range — an 84% reduction. Same channels, same general audience pool. The difference was not a bigger budget; it was finally knowing which parts of the existing funnel were earning their spend, and cutting the ones that were not.

Going from a range that could hit $900 to a ceiling of $69 does not just save money. It makes the whole recruitment budget plannable — which is worth more than the savings themselves.

The same spend, read two ways. Above: cost per lead week to week, before and after. Below: the loop that keeps it inside the band.

Attention was never the constraint. Centricity already had traffic, leads and multiple live channels — what was missing was a way to tell earning spend from quietly wasted spend.
Platform dashboards were grading their own homework. Each channel claimed credit on its own terms, so no two numbers could be compared and no total could be trusted.
The volatility itself was the cost. A range that could reach $900 makes a recruitment budget unforecastable, which is more expensive than the wasted spend inside it.
Scaling first would have scaled the problem. Raising budget before fixing the signal would have amplified whichever campaigns were losing money fastest.

Scaling with confidence

$50K → $147K in monthly ad spend.

Once cost per lead came down and, more importantly, stayed down, Centricity had something they did not have before: a real reason to trust the next dollar spent. Monthly ad spend grew from $50K to $147K — close to 3x — without cost per lead creeping back toward its old volatility.

$50K

Monthly spend before

$147K

Monthly spend after

Held

CPL band through the scale-up

What we built

One connected system, not five disconnected efforts.

A lot of "we manage your ads" engagements are really someone logging into Ads Manager and adjusting bids. That is not what held this together. This is the actual stack.

Paid ads management as one coordinated strategy

Meta, Google and TikTok run as a single recruitment programme rather than three separate budgets competing for the same candidates. One plan, one set of thresholds, one place where trade-offs get made.

Multi-channel attribution and reporting

Spend tied to what a lead actually did afterward, not to what each platform's own dashboard claimed credit for. Platforms are, unsurprisingly, not neutral referees of their own performance.

Funnel optimization and landing page strategy

The ad and the page treated as one unit. A strong ad pointing at a confusing landing page is just an expensive way to generate a bounce, and it shows up as a targeting problem when it is not one.

Retargeting and lookalike campaigns

Audiences rebuilt from the segments the data showed were converting, rather than a generic “people who might be interested” definition inherited from platform defaults.

Ongoing analytics and strategy consulting

The loop keeps running after launch, so the system sharpens over time instead of quietly decaying the way most set-and-forget campaigns do.

Why it worked

Every decision — which audience to keep, which creative to kill, where the next dollar goes — was made against a real weekly number, not a hunch or a platform performance score.
Thresholds were set before spend moved, so results were never graded after the fact.
Channels were compared on a single attribution model rather than on each platform's self-reported conversions.
The framework was handed over as something Centricity's own team can keep running, not a dependency on an agency login.

Before and after

The same recruitment budget, on a different footing.

Measure

Before

After

Cost per lead

Before$79 – $900

After$15 – $69

CPL ceiling

Before$900

After$69

Monthly ad spend

Before$50K

After$147K

Performance review cadence

BeforeAd hoc / quarterly

AfterWeekly

Operating model

BeforeFive disconnected campaigns

AfterOne framework across every platform

Figures reflect the engagement's measured cost-per-lead range and spend levels across Meta, Google and TikTok.

Impact

Nothing here was a guess. That is the actual product.

This is not a campaign that spiked once and faded — the kind of result that looks great in a single screenshot and terrible six months later. It is a system Centricity's own team can keep running and keep trusting, because it is built on a feedback loop rather than a one-time optimization.

84%

Lower cost per lead

~3x

Ad spend scaled with confidence

$69

New CPL ceiling, down from $900

1 framework

Running across every platform

Before

Spend decisions made against platform-reported scores, with cost per lead anywhere between $79 and $900.

After

A $15 – $69 band, one attribution model across three channels, and a budget that can be planned rather than guessed.

Ongoing

Weekly audits and dynamic reallocation continue, so the system keeps sharpening instead of decaying.

Take it with you

Get the full recruitment marketing case study as a PDF.

The complete write-up: the funnel diagnosis, the testing cadence, the attribution build, and the before-and-after cost-per-lead and spend figures. Useful if you need to take a recruitment budget case into a leadership or sponsor conversation.

Questions first? contact@matechnologies.net

Full case study

Recruitment marketing and cost per lead

Enter your work email and we will send the complete write-up — funnel diagnosis, testing cadence, attribution design and the measured cost-per-lead impact.

If your recruitment spend feels more like a bet than a decision, let's fix that.

Book a strategy call All case studies