A customer you verify at signup does not stay the same for long. People move, switch jobs, change devices and phone numbers, open and close accounts. The identity check that recognized someone on day one is slowly describing a person who no longer quite exists, and most verification systems never go back to ask the question again. Closing that gap is the work I care about most.
I’m the CMO of Jumio, where we verify identities for fintechs, banks, and marketplaces, then keep reading risk long after the first check clears. When Filip Verley had me on Liminal’s Friday Five, the rapid-fire questions kept pulling me back to one conviction: identity was never meant to be a one-time check, and treating it that way gets more expensive every quarter.
Identity is not a moment. It’s a moving target
Filip asked what problem we exist to solve, and my answer was about the lifecycle, not the login. For years, the industry has treated verification as a gate: confirm someone at onboarding, wave them through, consider it closed. But people are constantly in motion. They move between organizations and accounts, pick up new devices and numbers, and yes, some of them change the color of their hair. Each shift quietly erodes the confidence of that first check.
So you have to track it, hold that trust continually, and see it across the whole lifecycle. A clean signup is a snapshot. Whether someone is still who they say they are, and still trustworthy, is a question that has to stay open.
Even a continuous check has to survive deepfakes
Filip also asked what buyers in my category still underestimate, and for me, it is the challenge of deepfakes and AI-powered fraud. It is only going to get worse. The fraudsters are getting better and better at what they do, and the tools to fake a face or inject a video stream are cheap and continuously improving.
This is the part that makes the lifecycle argument harder, not easier. It is not enough to verify someone continuously if any check can be fooled by a synthetic face. Liveness detection, confirming there is a real and present human on the other side of the screen, stops being a feature and becomes the floor the whole system stands on. As providers, we do not get to be a step behind the people we are defending against.
How we are operationalizing this at Jumio
The fair objection is that no single company sees enough of someone’s behavior to track them well over time. My answer is that the signal does not have to live inside one company’s walls. That is the thinking behind cross-transaction risk and identity intelligence: capturing information while someone is in one organization and, with the right controls, making it useful when they move to another.
Framed that way, identity is something the whole network reads together, not a verdict any one company reaches on its own. Someone flagged at one institution does not get a clean slate at the next. Continuous verification works best when the network behind it sees more than any single participant could.
The small hill I’ll die on: be on time
Filip closes the Friday Five by asking for a small hill you would die on. Mine is punctuality. I am a stickler for being on time. At Jumio, we are good at it, even on video calls. It sounds trivial next to deepfakes and lifecycle risk, but it comes from the same place. Being on time is respect for other people’s time, and that is the instinct that should shape how we treat customers too. Every unnecessary check spends their time as if it were free. Done right, continuous verification gives us more certainty and asks less of good customers, not more.
Read full transcript
Filip: Welcome to the Friday Five. I am joined by Julie Roy, CMO of Jumio. Julie, thank you so much for joining us.
Julie Roy: Great to be here.
Filip: In one sentence, what problem does your company exist to make less painful, and for whom?
Julie Roy: Companies have been operating in terms of identification as a one-time check. The problem we solve runs across the customer lifecycle: being able to track individuals over time, see whether they are still trustworthy and whether any risk is involved, so that an organization can control that identity and work with people it can trust throughout the relationship.
Filip: What assumption about your market turned out to be wrong, and what forced you to confront it?
Julie Roy: What we have seen is that companies can partner across the risk of all their users, something we call cross-transaction risk. You capture information while a user is in one organization, and if they move to another, you can share that information. We launched cross-transaction risk to create a kind of community for managing risk and tracking users throughout their lifecycle across different companies.
Filip: What trade-off do buyers in your category still underestimate?
Julie Roy: Deepfakes and liveness. It is only going to get worse, and companies need to be on their toes, because the fraudsters keep getting better at what they do. As organizations and solution providers, we have to stay on top of all the deepfake creativity out there.
Filip: What will matter more in your market over the next 12 to 18 months than most people are preparing for today?
Julie Roy: It comes back to managing the lifecycle. People are moving around, moving between organizations, taking part in different transactions. They change devices, get new phone numbers and new addresses. Some even change the color of their hair. You have to track that and continually maintain trust across the lifecycle. That is where we underestimate what both trustworthy users and fraudsters will be doing.
Filip: What’s a small hill you’re willing to die on that most people would think is trivial?
Julie Roy: I am a punctual person, and I think it is important to respect others by being on time. At Jumio we are very good at it, even virtually. We show up to our calls right on time, and the work gets done quicker for it.
Key Takeaways
- Identity verification cannot be a one-time check. People change devices, phone numbers, addresses, and circumstances, so trust and risk have to be continuously analyzed across the customer lifecycle.
- Deepfakes and liveness are the threats that buyers most often underestimate. Continuous verification only holds if each individual check can withstand synthetic attacks.
- Cross-transaction risk extends the thesis across organizations: a signal seen at one institution can inform another, so risk does not reset every time a user moves.
- Done right, continuous verification adds certainty for the business while asking less of good customers, not more.
