A situation that comes up regularly in search conversations: a hiring manager brings a brief that looks, on paper, identical to one they filled eighteen months earlier. Same seniority, same salary band, same job title. What had changed was everything underneath the title, and it took two interviews for everyone in the room to realise the person hired last time wouldn't have been right for the role this time.
That's not a one-off. Across the platforms we recruit for, the skill set sitting behind a senior Salesforce, Adobe or Oracle CX hire has shifted meaningfully in the last year, and most hiring panels haven't updated their interview questions to match.
What's actually changed
The shift has a name in the Salesforce world: Agentforce, and the broader move toward AI agents that take action inside a CRM rather than just surfacing information. But the underlying change isn't really about one product. It's that the job has quietly split into two different skills that used to be one.
The first is the skill that's always mattered: understanding the business problem well enough to configure a platform around it. The second is newer, and it's the ability to design, govern and trust an automated or AI-driven workflow well enough to let it run with limited supervision. People who are strong at the first and weak at the second can still do good work, but they can no longer do all of the work a senior platform role now requires.
The market isn't short of people who know Salesforce. It's short of people who can be trusted to let an AI agent touch a customer record and walk away comfortable with that decision.
This is showing up clearly in where the market has gone tight and where it's gone soft. Entry-level admin work and junior development are, if anything, oversupplied right now, a few years of bootcamps and certifications pushed a lot of people into the bottom of the market at the same time the easiest admin tasks started getting automated away. Meanwhile demand for genuine data architects, governance-minded technical leads and anyone with real Data Cloud experience has kept climbing. Specialisation now matters more than it did three years ago, and generalists are having a noticeably harder time finding senior roles.
What this means for how you interview
Most interview processes we see for senior platform roles still ask the old questions: tell me about a complex implementation, walk me through a data model you built, how do you handle a difficult stakeholder. All still fair, none of them tell you whether the person in front of you can be trusted with the newer half of the job.
A better question, and one we now push clients to ask directly, is to describe a time they had to decide whether to trust an automated process with something that mattered, and what they did to make sure it wouldn't go wrong quietly. The answer tells you more about how someone will handle the next eighteen months of their role than almost anything else you can ask in an hour.
The hiring mistake we're seeing most often
The single most common error right now isn't hiring the wrong person, it's writing a job spec for the role as it existed two years ago and then being disappointed when the person who matches it perfectly can't do half of what's actually needed in month three. Job specs for platform roles have a shelf life now, and that shelf life has got shorter.
If you're about to open a senior CRM or MarTech platform role, it's worth spending twenty minutes before the spec goes out asking what's actually changed about the job since the last time someone did it. Often nobody has asked that question out loud, and the answer changes who you should be looking for.