Phenom just made its second acquisition in the behavioral science and assessments space in roughly ten weeks. BeApplied came first in February. Plum followed a few weeks later. Back-to-back moves in a category that has been a persistent gap in most enterprise talent platforms, and one that I think has been underdeployed for far too long.
I sat down with Phenom CEO Mahe Bayireddi to understand the thesis. It’s a conversation worth sharing.
What WorkTech Was Already Watching
Psychometric and behavioral science data have always had a credibility problem in enterprise talent technology. Not because the science is weak, but because the deployment has been narrow. As Bayireddi described it in our conversation, real-world usage has been concentrated at two extremes: high-volume hourly roles where you’re screening for reliability and show-up rate, and C-suite executive hiring where you’re trying to model team dynamics at the top. Everything in between has largely gone without it.
That is a real and costly gap. I’ve seen it from the inside. Earlier in my career, I was the general manager of a platform with embedded assessments and behavioral science, and I saw what it could do when implemented properly. The problem was never the science. It was getting it into the flow of work in a way that was consistent, contextual, and low-friction. Most platforms never solved that. They bolted psychometrics on as an optional module and left adoption to chance.
Most of the talent data flowing through ATS and HCM platforms today- resumes, job descriptions, performance reviews- is structured and already being consumed by LLMs. As general AI becomes commoditized infrastructure, the question of differentiation shifts: what data do you have access to that the base models don’t? Psychometric and behavioral data is a compelling answer to that question. The signal is richer, harder to replicate, and historically excluded from automated talent workflows. That is exactly the kind of proprietary data layer that matters in an AI-native architecture.
What Plum Brings
Plum’s role-modeling technology maps behavioral blueprints across 40,000 real-world jobs. Per Bayireddi, the methodology achieves four times the predictive accuracy for candidate success compared to traditional methods. That’s a meaningful asset, and importantly, it’s built on scientifically validated behavioral modeling rather than self-reported personality surveys. Those are meaningfully different things. The former has predictive validity. The latter has a long track record of producing data that feels useful but doesn’t hold up under scrutiny.
Bayireddi also named the people: Caitlin MacGregor (Plum CEO), Scott Allen (Plum CTO), and Neil MacGregor (Plum CPO), who came up directly as individuals aligned with Phenom’s direction. Phenom is explicit that their M&A model is built around product velocity; acquiring teams and technology to compress timelines on roadmap items they were already investing in. In his words, they want to know whether an acquisition can advance their direction by two or three years. I’ll be honest with you: this reads as an acqui-hire. That framing tells you something about how Phenom moves: they’re not buying scale, they’re buying acceleration. Whether you see that as a limitation of the deal or a feature of Phenom’s discipline depends on what you were expecting.
The Single CodeBase Mandate
One thing Phenom does consistently, and it is genuinely differentiated, is rebuild everything they acquire into their native codebase. No siloed business units. No patchwork integrations sitting on top of the platform. Every acquired product gets dismantled and reassembled inside Phenom’s architecture.
Bayireddi told me the expectation is full integration within 18 months, with meaningful customer benefits visible within six. The reason this matters isn’t just technical elegance. It’s that a psychometric marker captured during automated screening remains live data throughout the talent lifecycle — internal mobility, career pathing, retention modeling. A fragmented architecture makes that impossible. A single data integration flow makes it native.
This is where the promise of behavioral science has historically broken down. The data gets captured in one system and never travels. It sits in an assessment vendor’s database, disconnected from the ATS, invisible to the hiring manager, gone by the time someone is up for an internal role two years later. Phenom’s architecture, if it delivers what Bayireddi describes, solves that in a structurally different way.
See the full interview below…
The Five-Dimensional Context Problem
The more interesting strategic question isn’t whether Phenom can integrate Plum’s science. It’s whether they can solve the deployment problem that has always limited psychometrics in enterprise talent work.
Bayireddi frames this as a five-dimensional context matrix: industry, role, location, business-unit trajectory, and workflow-automation level. The point is that there is no single right answer for where in the talent workflow psychometric data belongs. In retail, you might apply it at sourcing. In healthcare, you likely can’t touch it until post-screening, and that changes again depending on whether you’re hiring nurses or home health workers. In executive hiring, the logic shifts entirely.
This is the right framing. The failure mode of prior psychometrics rollouts wasn’t bad science. It was one-size-fits-all implementation on workflows that required nuance the tools couldn’t provide. If Phenom’s agentic architecture can genuinely route assessments to the right point in the right workflow for the right context, that would solve a real, durable problem. The question is execution at scale across a 700-plus customer base spanning industries and geographies.
What It Means for the Market
For enterprise buyers evaluating Phenom, the near-term reality is that behavioral science integration is on the roadmap with a credible timeline and a team that has demonstrated they can execute post-acquisition. Whether the Plum IP meaningfully accelerates the roadmap or whether the primary value lies in the people and the validated job models will become clearer as product releases land over the next 12 to 18 months.
For competitors, two assessment-oriented acquisitions in ten weeks from a platform of Phenom’s scale is a signal. Not a category-defining one on its own, behavioral science has been promised before, but it is a signal that this cycle may be different because the infrastructure to deliver on the promise is materially more capable than in prior waves. Agentic architecture, single code base, AI-native orchestration: these are not the same conditions under which the last generation of psychometrics deployments struggled.
For Plum’s customers and former stakeholders: the science may actually receive better distribution within Phenom than it ever did as an independent company. A platform with 700-plus enterprise customers and a native data integration layer is a different delivery vehicle from a standalone assessment tool that asks buyers to bolt it onto whatever ATS they run.
The Bottom Line
The thesis Bayireddi is building toward is one I find compelling: as general AI becomes abundant infrastructure, differentiation shifts to context, judgment, and proprietary data. Psychometric and behavioral data are a legitimate candidate for one of those proprietary layers, arguably the most underutilized one sitting in the talent technology ecosystem right now.
Phenom has the architecture to deploy it differently than it’s been deployed before. They have a track record of actually integrating what they acquire rather than running acquisitions as disconnected business units. And they have a genuine strategic conviction here. Bayireddi told me they’ve been working on this direction for four years. That’s not a reactive move.
The story worth watching isn’t whether the thesis is right. It’s whether the execution at scale matches the ambition. On that, the next 18 months will tell you a lot.
Watch the full episode of WorkTech with Phenom CEO Mahe Bayireddi below.
George LaRocque is the founder of WorkTech, a market intelligence and strategic advisory firm covering the HR and work technology ecosystem. WorkTech tracks global investment, M&A, and tech strategy across 65+ specialized categories. The Opening and Closing Market Window Report Series, Q1 2026 Global Work Tech VC Update, and Q1 2026 Global Work Tech M&A Update are available at 1worktech.com. Phenom is not a WorkTech client.
Related Reading from WorkTech: