Logan Marsh

“Talent acquisition is part of the business, and I'm a strategic partner.”

Logan Marsh does not think of talent acquisition as overhead. As Head of Talent Acquisition at Calendly, Logan works as a strategic partner to the business, meeting with the C-suite regularly and treating hiring as a lever for company goals and roadmaps rather than a queue of open requisitions.

That conviction has been 18 years in the making. Logan has built talent organizations at Amazon (twice), Microsoft, Okta and DocuSign, and has led TA at Calendly for about two and a half years. Along the way came global teams, new offices in India, Europe, South America and North America, and hiring for products found in people's homes, including the Amazon Echo.

The same instinct now shapes how Logan approaches AI. It is less a threat to the profession than the next test of whether recruiting operates as a true business function.

The philosophy traces back to two leaders Logan reported to at Amazon. Both had worked on the business side before moving into recruiting leadership. From them, Logan learned to see recruiting as a business in its own right, a function and even a product of the company, not the overhead many assume TA to be.

"Talent acquisition is part of the business, and I'm a strategic partner," Logan says. "We are not here to just fill roles." Day to day, that means helping leaders drive their goals and roadmaps. Filling roles is the output. Enabling the plan is the job.

Logan fell into recruiting, as so many in the field do. Fresh out of college, with a competitive streak from sports, Logan agreed to a quick chat with an agency recruiter in late 2007 or early 2008, just as the market was turning down. The recruiter asked whether Logan knew what C++ was. Logan had no idea. Nearly two decades later, Logan laughs about the question, but at the time it sparked a curiosity about how the whole business worked.

A couple of years in agency work showed that competitive drive translated naturally into hitting targets and the sales motion of the job. Having led people in earlier roles, Logan wanted to get back to leadership. In 2010 came a move to Amazon as a corporate recruiter and, that same year, a first TA people manager role. Logan backfilled the individual contributor seat and built a team of 10 to support businesses previously handled solo.

Since then the roles have grown more complex, more global and larger in scope. The path has zigzagged, with some turns by design and some dictated by the industry, and Logan counts the challenges alongside the rewards.

The lessons from the leaders who shaped that growth are specific. A former chief technology officer at a previous company, a brilliant engineering mind with whom Logan partnered closely to build out global sites, became a pseudo mentor. Watching that executive navigate acquisitions, combined teams, global expansion and budgets of $100 million to $200 million taught Logan humility and how to stay steady and calm amid chaos. The executive was also generous with time, which Logan calls precious, and the two still keep in occasional touch.

Another leader at a previous company, at one point Logan's direct and skip-level manager, modeled the human side of leadership. The higher you rise, Logan notes, the lonelier it gets and the harder the decisions become. Through difficult periods, including layoffs, that leader showed how to stay grounded and keep the business connected to the people it affects.

Looking across these four leaders, the two at Amazon included, Logan sees a pattern that is no coincidence: each was grounded and calm, and each was excellent at the work. Strong org and people leadership, in Logan's view, rests on that foundation of knowing who you are and how you operate.

AI is not new territory for Logan, who was hiring machine learning scientists and researchers for Amazon businesses in 2012 and 2013, many of them Stanford and MIT educated pioneers of the field. Today's landscape is no surprise, though the pace of advancement is. LLMs now touch nearly every product in the TA tech stack, often without users realizing it.

Logan's response is curiosity. A personal Claude license serves for building schedules, asking questions and tinkering on projects purely for fun. At work, the team uses AI operationally every day, including developing an AI sourcing layer on top of the applicant tracking system to surface the best talent as efficiently as possible. Logan has also trained Calendly's people team on Claude at basic and advanced levels.

To those who are afraid, Logan offers respect and realism. The fear is real, and a healthy fear should be balanced with healthy curiosity. But ignoring AI "will only hurt your own career," Logan says. "That's not a threat. That's just the reality of where we are." Offloading administrative work frees recruiters to act as strategic advisors, and Logan tells the team the skill-building is an investment in their own careers, not just the company's.

There is a caution, too. Never assume the AI knows best, and never put out work that has not been strongly vetted and made your own. "AI slop," Logan argues, comes from overstretched people submitting raw output as a finished product. "You have to use it as a strategic tool," Logan says. "It's not you."

Logan's first piece of advice comes half in jest: lock in your headcount plans now. Finalizing next year's plan in the front half of Q4 rarely happens, but talent leaders should be working with the business on the roadmap and strategic plan, then tying it to their own capacity model, team structure and delivery timelines.

Second, rethink team design. Many longtime leaders equate success with scale. Logan, who has managed large teams, expects talent organizations to look different, potentially half agents and half people, with automations augmenting the work. Org design is not an afterthought or a capacity exercise. It is part of the strategic plan for the business.

Third, budget for AI training on Claude, ChatGPT, Gemini or whatever the tools may be. Capable people without enablement face something like writer's block, and they need to be up to speed over the next three, six, nine and 12 months. Finally, plan the full cost picture now, AI credits, team size and structure included. The alternative, Logan warns, is discovering in May or June that you overspent and now have to cut back.