In this episode of Mastering CS: Candid Leader Insights, Irina Cismas sits down with Anne LiCata, VP of Customer Success at Qoniq, an AI-powered analytics platform for pharma and life science teams. Anne brings over fifteen years of experience serving the life science industry from different seats, and now finds herself at the intersection of highly regulated enterprise clients and rapidly evolving AI technology.
She shares what it actually takes to get pharma teams to adopt an AI platform, why change management is the hardest part of the job, how she thinks about measuring value in a world where traditional SaaS metrics fall short, and why she believes AI will take over customer service but will never replace customer success.
What You’ll Learn
- What customer success looks like inside an AI-powered pharma analytics startup
- Why change management is the real challenge when selling AI to life science teams
- How Anne builds trust with customers who are asked to relinquish manual processes they have relied on for decades
- Why traditional SaaS metrics fall short when measuring the value of an AI product
- How Anne uses AI in her own day-to-day work and where she draws the line
- What skills she looks for when hiring for a CS team in an AI and life science context
- Why she believes AI will replace customer service but not customer success
- What getting back to basics actually means and why it still works
Key Insights & Takeaways
Change management is the core challenge in AI adoption. Asking people to trust AI with tasks they have done manually for years is not just a product question. It is a human question, and it requires transparency, patience, and constant reinforcement.
You have to show AI being right, not just tell people it works. Pointing out specific moments where the AI performed correctly and walking customers through those moments is more effective than any pitch.
Traditional CS metrics don’t fully capture AI value. Adoption and usage numbers are a starting point, but measuring the quality of AI outputs requires new frameworks that the industry is still developing.
Think of AI as your newest intern. It needs clear, precise direction. Without it, both the intern and the manager fail. The quality of your prompt determines the quality of your output.
You can teach the science, you can’t teach relationship-building. The most important thing Anne looks for when hiring is whether someone is inherently a customer success person, not a customer service person. That distinction matters more than any technical skill.
AI will take over customer service, not customer success. Transactional interactions can be automated. Trust, advocacy, and genuine relationship management cannot.
Get back to basics. When things feel stagnant, the answer is almost always to start having more real conversations with people. That has been true at every stage of Anne’s career.
Podcast Transcript
What Qoniq Does and What CS Looks Like There
Irina (0:06 – 0:42)
Welcome to Mastering CS, Candid Leader Insights, the podcast where we dive into the world of customer success with industry leaders. I’m your host, Irina Cismas, and today I’m joined by Anne LiCata, VP of Customer Success at Qoniq an AI-powered analytics for pharma and life science teams. Anne, I’m really happy to have you here.
Thanks for joining. Help me understand your world, the Qoniq, today.
What does the platform do? Who are your customers? And what does CS actually look like in that context?
Anne (0:43 – 1:20)
Terrific. At Qoniq, we help pharma and medtech companies to quickly and accurately and securely surface evidence and insights from across multiple data sources using generative AI. This is to better inform decision-making.
We’re a startup, so we’re very small. So we’re just beginning to mature. And as the VP of Customer Success, I wear a lot of hats, not only acting in a traditional customer success way, but I’m also doing product management, project management, and really whatever is asked of me.
Irina (1:21 – 1:37)
I feel you. I know how you work in smaller and lean teams where you are wearing multiple hats, as you said, and in this hour, you are running a report, then you are dealing with an escalation. Then you are, I don’t know, the startup world is a totally different thing.
Anne (1:38 – 1:40)
Oh, completely. It’s crazy.
What Anne Spends Most of Her Time On
Irina (1:41 – 1:48)
Oh, for sure. And in that setup, what do you spend most of your time on?
Anne (1:48 – 2:22)
I would say the bulk of my time right now is being spent in project and product management. As a startup, there’s a lot of work that has to be done, writing the business requirements. And as the person who is the voice of the customer and who is working and interacting with the customer regularly, I have the best understanding of what it is that they need and what it is that they require.
So now I take off that customer success hat and I put on that very limited technical knowledge hat. And then I start to create the documents that I give to others to dig in deeper too.
Getting Life Science Teams to Actually Adopt AI
Irina (2:24 – 2:39)
Pharma is one of the most conservative or compliance-driven industries out there. What does it actually take to get the life science team to genuinely adopt a new platform? Not just signing the contract, but really using it.
Anne (2:40 – 3:48)
Yeah, this is the tough one. Change management.
Change management is inherently challenging when introducing any kind of new solution. And then you want that new solution to drive meaningful adoption at the same time. This becomes extremely difficult when you’re asking your team or any team to move away from an established manual process and to place all of their trust in AI, in an AI-powered platform.
It’s not simply a matter of learning a new tool. You’re asking people to relinquish the methods that they know and they’ve relied on for a bazillion years and to believe in tech, that technology can replicate and improve upon what they’ve already always done themselves. So you’re from a change management aspect.
You’re not just asking them, Hey, look at this platform. It’s going to use, it’s going to support you and help you do X, Y, and Z. You’re asking them to place all of their trust in what you’re delivering.
Because let’s face it, as humans, we know best and it’s hard to give up that control.
How Anne Handles Change Management in Practice
Irina (3:51 – 4:20)
And this change management, how do you deal with it? Because it’s one of the toughest, I think it’s easy because it involves humans and everything that involves changing the behavior that it was like this for thousands or for dozens of years. It’s hard.
It’s basically going out of your comfort zone. So how do you do that? What are the tricks that you go to?
Anne (4:20 – 6:48)
Honestly, it’s not so much a matter of tricks as it is a matter of really learning to develop trust. When you’re working with anyone who’s moving a regular manual task into an AI environment, you have to build that trust from the ground up. You have to be very transparent in what you’re doing, how you’re doing it, and why you’re doing it this particular way.
A lot of people think of AI as the big bad, but really what it can do is take the manual out of certain tasks. Let’s take literature analysis as an example. In pharma and medtech, teams are hammered daily with hundreds of articles, publications, and scientific information. As humans, they have to ingest it, summarize it, determine what’s most important, and then act on it.
I have a customer that every week has to send a literature report all the way up to the top on their products and their competing products. They really have to keep their finger on the pulse of a lot that’s going on.
What AI can now do is go out there and find the publications, taking away the manual need to search across multiple sources like PubMed Central, Embase, and Scibite, and allowing the AI to identify and surface all of those publications.
Then AI can summarize each one in depth, giving five or six bullets about a particular publication. The user can quickly scan those points and identify which publications are worth their attention. The AI can also provide reasoning as to why a specific publication would be essential for them to read. So it takes the need to read about 500 publications a week, depending on the indication and the drugs, down to a handful, and then that handful down to another handful that can actually be presented to the organization. It takes away probably 8 to 12 hours of manual work and gives people the time to make much more valuable decisions and create their next best actions.
Irina (6:49 – 7:07)
You know what I’m curious, in your current role, because you are wearing multiple hats, what’s the hardest thing you are trying to move right now? Is it on the CS, on the product, on the project management, on the relationship with the accounts?
Anne (7:07 – 9:01)
Yeah, that’s a great question. With regards to relationships, that’s just inherent to who I am. It comes naturally. But I think the real challenge is that level of trust, specifically the trust that AI is going to perform well for you. That’s the most challenging thing: the change management, getting them to trust that the AI has identified the right articles for them rather than doing it manually themselves.
How I’m doing it right now, and this actually came up yesterday, is first pointing out to the customer where AI was correct. That’s number one, and you do it nicely and gently. Then offer to set up a quick training session to show them how AI did it correctly. That can be a quick ad hoc session, just 15 minutes, or I can record a session and go through it step by step so they always have something to refer back to and show their colleagues: look, they put this together for me and the system is doing the job for us. Now let’s move on to our next task.
You’ve got to come up with a number of different approaches and you have to stay with them. You can’t just assume they’re going to use it. You have to keep checking in, making sure they got what they needed, and doing their job side by side so you’re aware of where they are. So when the
What a Healthy Account Looks Like and How Value Gets Measured
Irina (9:02 – 9:19)
What does it mean in your world that an account is healthy? What are the things that you are monitoring or what are the things that you need to constantly be aware so that a renewal happens or a client remains with you?
Anne (9:19 – 10:47)
Yeah, that’s another really great question. In a traditional SaaS platform, you’re doing QBRs throughout the course of the year and you’re able to provide very quantitative metrics. You can say you have this many users, and we’ve watched usage increase in this way. We’ve got all of that.
But now you also have to bring KPIs around AI itself, which is an emerging discipline. How do you bring KPIs to an AI product? Conventional metrics like adoption, usage, and revenue expansion are familiar baselines, but they fall short when it comes to evaluating the quality of AI performance, particularly the quality of LLM responses. How do you score relevance? How do you trust an output? These are observational, non-numerical judgments that demand new frameworks. Getting them right is one of those nuanced challenges in your overall AI strategy.
That’s actually something I’m working on right now: developing an additional way to show that the customer is deriving value from the platform, not just through the standard QBR metrics that we’re all so familiar with.
Measuring Value When Traditional SaaS Metrics Fall Short
Irina (10:48 – 10:56)
And how does this process work out that you are working on as we speak?
Anne (10:56 – 11:34)
Early this morning someone asked me how that’s coming along, and I said it’s still half-baked. So it’s still half-baked as of this morning.
One of the things I’m also doing, because I genuinely believe in AI, is asking the models themselves: how do you measure value? How do we find value with your system? And starting to develop that mindset of thinking like AI does, so that you can further educate your customers.
How Anne Actually Uses AI Day to Day
Irina (11:35 – 12:39)
I’ve talked with a few leaders who told me that knowing how to use tools like Claude, GPT, and Gemini is becoming a requirement in this world. And you also work at an AI company. How do you actually use it on a day-to-day basis to simplify what you do?
I think there are two extremes. On one side, there are CSMs who use it only to summarize meeting notes. On the other side, there are those who develop sophisticated models, build workflows, and create dedicated agents. Where do you fit?
Anne (12:39 – 15:24)
I am not the casual user who will just ask for a recipe to bake a chocolate cake. And I’m definitely not the person developing agentic AI using different agents and things like that. I leave that to my experts. When they start talking about it, I sometimes glaze over because it’s technical above my head. I sit right in the middle.
I’ve learned how to write a good prompt in order to give AI the right direction about what I need. For example, I’m working on something right now where I needed to know what data fields come from a particular API. I just went to Claude and asked what data fields I can expect from that API, and it gave me an entire list. Now I don’t have to go ask my data scientists or engineers. I can build my requirements around what I know I have access to.
I also ask AI to rewrite things for me. Maybe my tone was a little too harsh in something I wrote. We’ve all been there, fired up, typed an email, and rewritten it 14 times before sending it. I’ll tell AI, I was too harsh, I need to tone this down, this is the lens I want it through. A friend sent me their resume and asked for help. I thought, this is a bad one and I don’t have time to rewrite an entire resume. I used AI, it rewrote the resume, and she was quite pleased with it.
Now, we all know that AI can hallucinate, and some models are better than others at certain things. You always have to have a human in the loop, go through it, read it, make adjustments, and adjust your prompt if you’re not getting what you need. We look for 95 to 100 percent correctness, but AI is never going to be 100 percent correct. It doesn’t have the reasoning that we have yet. It can’t apply objective judgment to something. So you always have to check it. Always.
Irina (15:25 – 16:09)
I like to say that you should be the one leading the AI, not the other way around. If you’re just taking what AI gives you and passing it along, you aren’t leveraging anything. You’re just copy-pasting, which is a lazy job. But when you are guiding the AI, telling it exactly what it needs to do and why, and challenging it on specific things, that’s where the outcome improves. You need to know your line of business and have that clear expertise. That’s what allows AI to actually help you.
Anne (16:11 – 17:11)
It’s not going to help you if you don’t know what you need. When I have a new user coming on board and they’re asking about AI, one of the analogies I use is: think of AI as your newest intern. Someone you’re going to give a lot of manual work to, but that you have to give very clear direction. Otherwise your intern is going to fail and you’re going to fail.
You have to give it very clear, very precise direction on what you’re trying to accomplish. You can ask it to quickly summarize an article and it will summarize that article. But if you want more specific information, maybe you’re in pharma and you’re looking for specific guideline information or dosing information, you need to be very specific in the question you ask. Otherwise you’re just going to get generalizations back.
What the CS Team of the Future Looks Like
Irina (17:12 – 17:33)
With everything that changes in the world now, how do you think the setup of a CS team will change? How would you build your team? For instance, what roles and what skills were you looking for in order to have a successful CS team?
Anne (17:33 – 19:07)
I’m scaling at this current time, and one of the key things, especially in AI and life sciences where everything has to be so precise and goes up against regulations, is that I still go back to basics from my recruiting days. As a recruiter, I know what I’m looking for and I know how to find it.
One of the key things you can’t teach somebody is how to be an inherent customer success person. Customer service people are entirely different from customer success people. Customer service people are transactional. And this is one of the key questions I always ask candidates: are you looking to do customer service or customer success? Because they are two separate roles.
I see AI pretty much taking over customer service. That transactional interaction, your chatbot, asking where your next TV is going to be in stock, that’s customer service. Customer success is going to go out and find that TV for you and make sure you’re completely happy with it.
People and leaders have laughed at me when I say I look for burned-out recruiters first, because I know they can build a relationship. And that’s key. You can teach the science. You can teach the technology. You can’t teach how to build a relationship. That has to already be in somebody.
How Anne Is Scaling Her CS Team Right Now
Irina (19:09 – 19:25)
You mentioned you are in the process of scaling. We don’t have time to go deep on it today, but we’ll definitely do a dedicated conversation on scaling at some point. For now, give me the summary. How are you scaling as we speak? What does scaling mean in your case?
Anne (19:26 – 20:33)
Scaling means first knowing what an individual customer success person can carry. How many customers can they support? You think about that and you think about the logos you’re currently managing.
As I scale, there has to be that interplay with sales. I don’t want to bring somebody on board only to have to let them go because we don’t have enough customers. I have to work closely with sales to find out what’s coming in so that I can do some predictive modeling on when we might need that person, what type of person I’m looking for, and how much they can handle. It’s a process I collaborate with sales strictly on. And honestly, if you’re not a good collaborator with sales, you’re not good at customer success.
That collaboration helps me build those predictive models for when I’m going to need additional people. It’s all based on the work sales does and how many logos they’re going to bring in.
What Most CS Teams Get Wrong
Irina (20:35 – 20:41)
So that I can scale up from there. From everything you’ve seen, what do most CS teams get wrong?
Anne (20:42 – 21:59)
They don’t get back to basics. Back in my recruiting days, I started a program called service delivery, which would have been considered a precursor to customer success. The goal was to maintain contracts longer, same philosophy, same mentality. And it was successful because you were building relationships.
Current CS teams, and I have gone from one CS team to another, are really reliant on just sending out mass emails, texting, relying on technology to build the relationship. You’re not going to do that. Before the pandemic, you traveled a lot. You got on site, you sat with your customers. What I see now from customer success people is that they don’t want to turn on the camera. They just want to send out a bunch of emails, automate everything, and send an automated QBR. No. You have to get back to basics. That’s where I say AI will definitely take over customer service. It cannot take over customer success.
The Lesson That Has Stayed with Anne Through Fifteen Years
Irina (22:00 – 22:12)
Last question. You’ve had a pretty varied path across healthcare and life science. Looking back, is there something you learned early on that still shows up in how we work with clients today?
Anne (22:12 – 22:56)
It goes back to when I first started out as a recruiter, and it’s getting back to basics. When you find yourself in a position where you feel stagnant, get back to basics. What made you successful to begin with? It was having conversations with people. Sometimes just getting back to basics and having a good conversation like the one you and I are having right now can light that spark and bring back the enthusiasm.
That’s what I do going forward. When I’m mentoring or managing a team and someone is struggling, I’ll say, okay, it’s time to get back to basics. It’s what my first boss told me to do many times.
Irina (22:57 – 23:28)
Anne, this was a really interesting conversation and 15 years serving the same industry from different seats. And now at the point where AI is actually landing in that world, there was a lot here that doesn’t come up in a typical CS episode. So thanks so much for joining me.
And to everyone listening, thanks for tuning in. Until next time, stay curious, keep learning and mastering customer success.
