In this episode of SCRS Talks, Jimmy Bechtel sits down with Tom Farrand, Co-Founder and CEO of Delfa, to break down what a healthy recruitment funnel looks like from first inquiry to scheduled visit. Tom shares benchmark rates sites can measure themselves against, how to spot where a funnel is leaking, and why both speed to contact and consistent follow-up matter. They also walk through the three Cs of capacity, cost, and conversion, a simple framework to guide weekly staffing and spend decisions. Whether you’re managing a handful of leads or a large network, you’ll leave with practical numbers to track and ideas to try.
Audio and video play straight from the show's own hosting. PodSubs stores and serves no media.
Transcript
From the show. Tap a time to play from there.
Jimmy BechtelHello, and welcome to SCRS Talks provided by the Society for Clinical Research Sites. I'm Jimmy Bechtel, the chief site success officer with the Society, and today I'm joined by Tom Ferrand, who is the current co-founder and CEO of Delfa. Tom's gonna share with us a little bit about the recruitment funnel and how we maintain an effective, healthy recruitment pathway for our patients and participants. But before we get into the topic, let's start with you, Tom. I'd love to learn a little bit more about you and maybe a little bit of background on Delfa.
Tom FarrandPerfect. Thanks so much, Jimmy. It's a pleasure to be here. As you said, I'm the CEO, co-founder at Delfa. We're live with over 250 research sites globally, with our sites seeing productivity gains in their recruitment of over 70%. My background is working at the intersection of AI and the life sciences, I was initially introduced to this space by my co-founder, who lived and breathed this problem during his PhD, where he observed his colleagues having to work through big old Excel spreadsheets, calling participants one after the other, and really not achieving the enrollment goals that they had and delaying the studies that they were working on.
Tom FarrandSo yeah, it's a pleasure to be here, and happy to take any questions that you have.
Jimmy BechtelAwesome. I'm looking forward to talking with you about this, Tom. We know recruitment always sits at the top of the challenges that our site community faces I'm excited to peel back the curtain a little bit on some new ways to think about this problem. So let's talk about that. Can you walk us through what a healthy recruitment funnel really looks like from that first inquiry that you get, or, that first potential, solicitation instance, interaction instance with a patient, all the way through up to the successful enrollment?
Tom FarrandYeah, absolutely. So at Delfa, we think about patient recruitment as two sides of the same coin. One is you have to identify participants to join the trial, and this is, through marketing potentially. There are lots of solutions out there, fantastic solutions out there to go into health records and flag potential participants. But then there's the conversion, and there's the ability to turn those interested or potentially eligible participants into scheduled visits. And so really, Delfa focuses on the conversion side of things. How can we help you to take those potentially qualified participants, those participants who have put their hands up through a meta advert, for example, and actually get them into your site for a scheduled screening visit?
Tom FarrandAnd I think as we look across this, we see that there's a few different metrics which indicate health within the funnel. Obviously, there are some that are very study-specific, such as you have Jimmy on the phone, he is interested in joining your Alzheimer's study. The likelihood that he qualifies for your Alzheimer's study is going to be directly tied to the IE criteria of that study. So I'm gonna talk in more general terms without being too therapeutic area-specific and we can dive into TAs if they're interesting. But the first one is gonna be the reach rate.
Tom FarrandWe typically, see if sites have an advertising flow set up where they're generating new leads from a digital campaign, then good sites and best-in-class sites are gonna be getting anywhere between 60 to 75% of those participants engaging over the phone be that through SMS or phone call. And so that's the first kind of hallmark at the top of the funnel. Are you able to actually reach those participants who are expressing interest in your studies? The next piece then is the ability to, once they are on the phone, how many of them are actually going through the pre-screener.
Tom FarrandSo are they engaging all the way through, or are they dropping out early? Are they picking up the phone and then hanging up on you? Are you getting wrong numbers? Are people too far away from the site? All of these factors impact whether or not someone actually completes the pre-screener. So typically, we see that about half of the participants that you're gonna reach are gonna actually complete that full pre-screener. Whether or not they qualify for the study is obviously very dependent on the study that you're screening for. But overall, you can think of that fifty percent number as for whatever study you're qualifying for, if you have a well-designed pre-screening script and flow, then about half of those participants are gonna be coming through there.
Tom FarrandObviously, if you're pre-screening for something much simpler, that's gonna be higher. If it's much longer and much more complicated, then people will likely drop off just because of the length of those engagements. And then the final piece is based on the number of participants who potentially qualify, you're gonna want to see about eighty percent of those schedule and turn up for that visit, and that's gonna be your best-in-class sort of show rate for the visits that you're booking. And this is across if you're booking pre-screening visits, for example, if you're doing fiber scans for MASH studies or memory screeners for the CNS or the, the Alzheimer's dementia studies versus also screening visits.
Tom FarrandAnd so a lot of the work here goes into sending reminders sending follow-ups, and nurturing the participant from the point that they stop the pre-screener to the point that they turn up at the site. So those are the three agnostic numbers we think about. Reach rate being about seventy-five percent as best in class, fifty percent of those who are reached completing that pre-screener, and then eighty percent of those who qualify for the study turning up to the next step of that journey, whether that's a pre-screening visit, a pre-screening phone call, or an on-site screening visit.
Jimmy BechtelThanks for sharing that information, that data, Tom. I think it's really important to understand objectively what one might expect so that you can benchmark yourself. You can use that as sort of a KPI comparison to what your rates currently are, and explore where the issue is, right? Having that data and understanding that allows you as a site to pinpoint exactly what you should be looking at, what period along that spectrum is really where you might identify, again, any pitfalls or challenges.
Tom FarrandFor sure.
Jimmy BechtelWhen leads come in then, Tom, and, then enrollment sort of stalls, how do we, kinda getting at what I was saying, start to figure out if it's a lead quality problem- or something happened after the lead arrives? How do we diagnose where we are seeing the issues really along the spectrum of patient engagement and enrollment?
Tom FarrandYeah. It's a great question. And it's one that grows-- the challenge of this kind of root cause analysis of where is the issue in the funnel, where is the funnel leaking, grows as you scale the funnel. The more leads you have in your funnel, the harder this question becomes. Think for smaller sites and for campaigns that are generating, 20, 10, 20 leads per week it can be pretty straightforward to figure out where these are happening. You can go and listen back to the calls, you can review the transcripts of those calls, you can speak to the coordinator who's done those efforts.
Tom FarrandNow if you're a larger site or you're suddenly part of a network and those number of leads per week is an order of magnitude larger, then you can't rely as much on that anecdotal approach. You can't cover all of the bases there. And so in these types of scenarios, we look really at a few different factors. The first is that reach rate again. if you are generating poor lead quality, then ultimately you're not gonna be able to get many of those participants on the phone or engaging with you, whether that's over SMS or email, for example.
Tom FarrandAnd so the mere ability to actually engage a participant is a strong indication that lead is of reasonable quality. The second thing that we then look at is the length of the calls that you're doing. If those calls are very short and you see what we classify as an lots of early hang ups, so these are calls that are less than two minutes, generally that means that someone's picking up the phone, they're hearing what you have to say, and then they're deciding that's not for them. And this is a big indicator that your targeting isn't correct and that there's, something potentially wrong in how you've set up the ads.
Tom FarrandSo I think those are two quite straightforward general metrics that you can look at and that you can pull. I think then there are more detailed pieces of analysis. I think what we see, and certainly what we make use of, is we can use AI to analyze a lot of the call transcripts that happen and the text message transcripts, and we classify those. So in some cases, we found that sites are running simply poorly geographically targeted ads. they have a huge number of participants that are saying, I'm not in Michigan, I'm in Missouri," and you've typo'd somehow in your Facebook campaign, and that's caused that setup to be wrong.
Tom FarrandAnd so we certainly see that, and this is especially true with organizations that are across different geographies. It can be difficult to manage across those areas. And so once you analyze the transcripts and you look into the detail of what's going on these calls, I think that then really gives you a better sense to the, deeper issues and what the root causes of that might be. Some cases we found the pre-screening script just had a very confusing question. We've seen questions that have been phrased as double negatives, which confuse the participant and then ultimately cause a very big drop, even though the lead quality is good.
Tom FarrandSo I would say looking at the top of the funnel metrics around length of engagement and reach rate are the two key, the quick wins that you can do, and then it's about diving into the the data a little bit more and really understanding both your studies, but also the patient population you're dealing with.
Jimmy Bechtelgreat insight, Tom. One of the things I really took away from that was if the leads and the conversations that you're having with those leads, like you said, the top of that funnel, the starting point is not good, then all the rest of it is just gonna fall apart from there. So starting from the top and making sure that process, the execution of it, the questions, the quality, all of that is there first, and then you can move down the funnel to see, okay, where else? Are patients falling out for reasons that are aligned with our statistics and our data or is there another, influential factor here?
Jimmy BechtelAnd I think that's, another really important point that you brought forward is having, you continue to bring forward, is having the data to understand that as a site is, the first step in making sure that you can assess what's going on.
Tom FarrandAbsolutely.
Jimmy BechtelSo Tom, then what matters more, when we're looking at the, the different paths that an individual in patient recruitment or an organization broadly even could take, in how fast you move or how consistent you are? And how do you even compare speed across totally different studies, right? Because like we've been talking about, the data and the KPIs surrounding this are so important to understand. But as you've also alluded to, every trial and every TA and every region is different for a lot of regions. So how do you start to normalize that as an organization and, again, address the question moving fast and consistency?
Tom FarrandYeah I think it's really interesting and unfortunately for the sites listening, I think you have to be both fast and consistent. In today's world, we live in the attention economy, right? And participants aren't hanging around for long as they are generated from social media campaigns, which is really the recruitment funnels we're focused on here today. Participants aren't necessarily hanging around for long and, the attention span that they give to your study is very limited until you actually engage them and you are able to build that rapport, you're able to build that trust and become an authority figure in their mind for this type of engagement.
Tom FarrandAnd so To answer your question, I think it's both. You have to be fast and consistent, and the brutal reality that we see in our data is that for speed especially it is a very much an exponential drop-off. If you are able to contact a participant within fifteen minutes of them expressing interest to your study, you are twice as likely to pre-screen them as after fifteen minutes, and that drop-off continues to the hour mark, whereby fifteen minutes to an hour, again, you're half as likely to, to pre-screen, someone and continues on all the way up to the twenty-four-hour period.
Tom FarrandAnd obviously, that, places a huge amount of operational pressure on sites. And so what we've seen is that even if you're able to send an automated email, text message, you can trigger obviously AI agents nowadays to do that pre-screening, and that can be far quicker. That, that can really help to supplement all of these kind of the ability to move fast and engage participants. I think from a consistency point of view, we found that generally four calls is the gold standard, whereby after that point, if you're calling participants more than four times, this is both for humans and AI, it becomes somewhat of a diminishing return.
Tom Farrandfrom calls onward, you are increasing the likelihood that you get that participant to engage by between one and two percent, and so you're much more likely either pausing that and restarting that outreach a few months later, a few weeks later for that study if it's longer running, or focusing your attention on another participant. And so- You mentioned therapeutic areas. How do we manage across that? For most of our sites, what we always recommend is start with the gold standard. Start with the recommended outreach that we have within the product, which is effectively a text message.
Tom FarrandYou then wait 15 minutes. You trigger a phone call. You then call them. If they don't pick up, You send a follow-up text message saying,"Sorry we missed you." You then call them a second time on the same day, and then you have additional follow-up text messages and emails, and have two more call attempts on the day after, and then two days after that as well. That is reasonably all-encompassing for most studies and most participants. Where we see that diverging the greatest is when you're dealing with participants from your existing database, when you're reactivating older, colder leads and you want to be a bit more cognizant the fact that they haven't actually engaged actively with a campaign, and so you wanna be slightly lighter touch.
Tom FarrandBut for those who are actively sticking their hand up, you need to be as fast and as consistent as you possibly can.
Jimmy BechtelA lot to break down there, Tom. A lot of really great information to break down, just, some, again, how a site might approach this, how they might think of it, how they might benchmark themselves in terms of that. And I think it's a really important point what you said, right? Speed and consistency are both important in their own unique ways, right? Because we're not only trying to enroll the study as fast as we can, but I think what's even more important is the, what I would argue is this universal, maybe misunderstood truth that strike while the iron is hot, right? If the patient is engaged in that, and they're, they're reaching out, probably, as you mentioned, their interest can quickly fade.
Jimmy BechtelThey start to second-guess, right? We can come up with all the different reasons that rapid engagement is such a successful technique, and we've heard that before on different platforms at SCRS. I think that's a really great point particularly that you made there amongst several others.
Tom FarrandYeah, and ultimately, a lot of these studies these days are competitive enrollment, right? And so fundamentally, you need to strike while the iron is hot because you can be sure that others won't be waiting around, and they're gonna be working to ensure that their quota gets filled, and their site can succeed against their goals. And I think fundamentally, there is a wider competitive dynamic that happens as well for sites.
Jimmy BechtelYeah, no doubt, right? Especially in a world where we see the types of patients we're looking for continue to grow ever more specialized and niche, i- I guess, if you will, right? The, the requirements that we have of them to participate in clinical trials from an inclusion and exclusion, but also just a participation perspective narrows that funnel. Every single patient matters, and being the first in line for that patient in a potentially competitive or densely research populated community speaks to the point that you made very acutely. So then, Tom, if a site could look only maybe at a specific handful of numbers every week, we've been alluding to these KPIs and these things along the journey here in this conversation.
Jimmy BechtelLet's put a pin in a few here that guide their staffing and their spend in relation to the topic that we've been discussing here. What would be on that list in your mind?
Tom FarrandYeah, absolutely. I think when we think about this internally at Delfa, we think about the three Cs. We think about capacity, cost, and conversion. And so fundamentally, the site needs to be able to clearly understand what is the capacity they have to book screening visits, pre-screening calls, whatever it is. What is the availability either within people's calendars or physical space at the site? What is the capacity they have on a week-by-week basis? And from there you can work backwards to understand, the wider funnel. In doing so, you need to understand the conversion of, okay, a new lead comes in, what is the likelihood that they actually turn up for a screening visit or a pre-screening call?
Tom FarrandWhat is the conversion a- across my funnel from lead to visit? If it's a newer study, this can obviously be difficult to understand because you've just getting going. And I think it's what we've seen, exceptional sites do is to benchmark from previous funnels that they have from existing therapeutic, from previous protocols in the same therapeutic area. They can use these rules of thumb to actually have at least an informed estimate of what that conversion is gonna be. And then once you have that conversion, you can apply... and the capacity, you can then understand based on your cost per lead for your campaign, ultimately what that budget needs to look like to fill the capacity based on the conversion that you have and how much you have to spend in order to actually generate a new lead or generate interest.
Tom FarrandNow, there is some caveat here. There's plenty of caveats as there is, always is in clinical research. If this is assuming that you're paying for these new leads each and every time. We know lots of sites have fantastic databases that they can access, they can make use of. They have patients who within the community they have great relationships with that they can actually reengage, and we've seen that as a fantastic source for very well-established sites that do have these depths of databases. So this doesn't account for that, but certainly these three Cs of capacity, cost, and conversion are a great starting point to actually get your framing as a site around how should I be thinking about staffing cause we have that capacity.
Tom FarrandThat's really included there. How should I think about spend? And that's the cost angle. And the cost per lead, the marketing spend, and then finally what connects those two together is that conversion across the funnel. So that's really how we would frame it for sites. And happy to engage folks and go deeper into that discussion.
Jimmy BechtelI think those are really great metrics, Tom, to at least look at and to learn a little bit more about the various solutions and then how one might go about doing that. So thank you for sharing those. And also thank you for the conversation today. We're running short on time here, but it was great to learn. I always appreciate hearing about objective ways to approach a situation that does sometimes have a very subjective thought process behind it, right? We're dealing with patients here. We're dealing with humans. It's tough sometimes for sites to reconcile the balance between the human factor and the data behind that to make more effective decisions in what they do.
Jimmy BechtelSo again, thank you for sharing some of some insights into how a site might approach this going forward.
Tom FarrandAbsolutely, yeah. Thank you so much for having me. We're thrilled to be part of the SCRS community, and it's a pleasure to chat with you, Jim.
Jimmy BechtelIt was great to talk with you, as I said, too, Tom and again, thanks for being here. And for those that are listening, I also wanna make sure that you take some time to check out other site-focused resources on our website, myscrs.org, other podcasts, webinar opportunities, and articles centered around site relationships, recruitment, and engagement. But for now, thanks for listening, for tuning in, and until next time.
In this episode of SCRS Talks, Mike Wenger, Chief Innovation Officer at CRIO, discusses the opportunities and risks of AI in clinical research. The conversation explores how sites can benefit from AI-driven efficiency...
Clinical research software has rarely been built with sites in mind. Storm Stillman, CEO of SiteBase, by Curebase and one of SCRS’s newest Global Impact Partners, talks about how that’s changing. Storm shares what...
Technology is consistently one of the top pain points for research sites, and a big reason is that sites often aren't part of the conversation until a solution is already built. In this episode of SCRS Talks, Jimmy...
Site engagement often shows up only after a study is already in trouble. In this episode of SCRS Talks, Jimmy Bechtel sits down with Julie Wilhjelm, Senior Director of Site Strategy and Engagement at Novotech, to talk...
Lauren Stockwell sits down with Jennifer Larkin, VP of Business Development at M3 Wake Research, to break down what a modern site BD function actually looks like day to day. Jennifer walks through how data intelligence...