Most multifamily operators treat centralization as the destination. For Michael DiMella, it’s table stakes. As Managing Partner of Charlesgate, the Boston-based integrated real estate firm he has helped build since 2003, DiMella has spent the better part of a decade replacing the generalist property management model with what he calls functional specialization.
Rather than moving the same jack-of-all-trades roles to a back office and calling it centralization, Charlesgate gives narrow, outcome-owning roles to the people closest to residents and runs leasing as a true commission-driven sales function. The firm operates primarily as a third-party manager across New England and Colorado, with additional arms in development, condo sellouts, marketing, and investment management.
DiMella will bring that operating thesis to the stage at Blueprint this year. His argument is that structure has to come before software: operators who bolt AI onto a broken org chart keep hitting walls, while those who redesign roles around outcomes first can use technology to extend a model that already works. The conversation below previews that argument, lightly edited for length and clarity.
BH: Centralization has become the default narrative in multifamily, but Charlesgate has pushed past it into what you call functional specialization. Walk through the distinction operationally. What does a specialized role do day-to-day that a centralized role at a peer operator does not?
MD: Honestly, centralization as most operators practice it doesn’t go far enough. The typical move is to take the assistant property manager off site, relocate them to a back office, and hope that global teams or AI fill the gap. But it’s the same workflows, the same processes, the same jack-of-all-trades job.
What we’ve built is different. We call it functional specialization: roles tasked with driving specific outcomes inside a team structure that, together, drives property performance. For example, we’ll have a resident hospitality manager. It’s an on-site role whose entire job is the resident experience. They’re not also handling rent collection, maintenance dispatch, and everything else. They own that one outcome.
The traditional property manager role, which owns the financial outcome and overall strategy, becomes more centralized. They may visit the asset to check on it, but their job is to direct the team driving performance. Back-office functions coordinate maintenance, collections, and accounting. The principle is simple: the roles that should be closest to the customer are closest to the customer, and they’re accountable for that and nothing else.
Leasing works the same way. Ours is a professionalized leasing function staffed by salespeople, not entry-level property managers three months into the job working off a weak script. We run an on-demand leasing squad that covers multiple properties, and those people are on a sales career track, not a property-management one.
That team is tightly integrated with our marketing agency, so we close the full loop: brand awareness, demand generation, and conversion to lease. We don’t have marketing sitting in an ivory tower or a property manager spinning up templates on the side. Everyone owns a specific outcome rather than one or two people trying to do everything at once. It also sets us up well for AI, because we’ve already done the workflow redesign that AI will eventually fill in.
BH: Describe the org chart before and after. What roles changed, and how did the compensation structure change alongside it?
MD: Before, you typically have a property manager doing leasing, or a junior person who’s nominally a leasing consultant but is really an entry-level employee trying to become a property manager. Their whole career track points toward property management.
What we’ve built instead is a team of people who come in to do leasing and stay in leasing. Their career path looks like any other sales organization, from individual contributor to sales manager to director. It’s the full sales trajectory you’d find outside multifamily.
Compensation follows from that. It’s highly leveraged salesperson compensation. We run a few different models, but it ranges from a small base plus commission all the way to full straight commission. It’s structured like a real sales role, not a salaried generalist position.
BH: What are the indicators you watch to know the model is working, the leading signals before the NOI shows up?
MD: We start from the top-level outcomes. We’re trying to improve the revenue side of the P&L, which means fewer days to lease, higher occupancy, lower cost per lead and per conversion, and better resident reviews. Then we cascade those into the metrics each team owns.
Every team operates on weekly scorecards, and daily in some cases, so it’s not a monthly review cycle. It’s closer to an EOS-style system where each level of the organization dials in on the specific numbers it owns. An individual contributor watches granular metrics like tour-to-conversion rate. The marketing team owns leads created. But ultimately everything rolls up to outcomes: shorter days to lease, lower vacancy, and better-than-market rents.
The way I frame it for the team is that our job is to drive alpha. I don’t know why you’d hire an operator for any other reason than to outperform the market. I’ve always pushed back on the traditional mindset that the market is what the market is. There’s an aggregate level of demand in any market, but there’s no reason a given property or set of properties can’t outperform it through better leasing and better marketing. I want us out of the spreadsheet-only mentality where the comps are the comps and that’s the end of the story.
BH: Can you quantify the impact, and how much do you attribute to the structural redesign versus the technology layer underneath it?
MD: I was in Denver this week for a pitch, and I’ll give you a live example. We took over a 400-unit portfolio there about six months ago, and the before-and-after is dramatic. We drove occupancy up roughly 35% on some of the properties over the course of the winter. We typically see about 4x on marketing results, both leads and tours, before versus after, along with a meaningful jump in occupancy and fewer concessions. These are tangible results, and I’m happy to share the case studies.
At this point, I attribute the impact primarily to the structural redesign. The technology layer will let us extend it, but the fundamentals are about clear ownership of outcomes by specialized roles inside a team. A lot of the AI pieces are still relatively new for us, so they’ll grow in importance over time.
But I’d caution against going the other way around. We’ve watched the industry chase things like self-guided tours purely to cut costs, and outside of a few rare exceptions it has saved cost while costing real dollars on the revenue side. Our sequence is to structure it right, define clear outcomes, put the right people in the right seats, and then layer automation and AI on top to handle the mechanical work. If you bolt technology onto a broken structure and expect it to instantly fix things, it doesn’t work.
BH: Describe the actual tech stack. What’s the system of record, what sits on top, and where are the integration seams?
MD: We use a lot of the standard industry pieces. AppFolio is our system of record, including some of their AI tooling, and we use EliseAI on the leasing side, among others. That’s the industry layer.
Underneath it, we’ve built our own knowledge and context layer, covering our processes, institutional knowledge, and workflows. We’re wiring that in with agentic systems, including Claude, so the team can rapidly review, approve, and deploy automations that accelerate a lot of our back-office work. The goal is to use technology to free up humans to do what is uniquely human, which is building relationships and delivering a better resident experience. We rely on best-in-class industry software where it makes sense, but we’re increasingly building our own layers behind the scenes.
BH: Most operators who attempt a redesign like this hit resistance from regional and property managers who lose direct reports or budget control. How did you manage the internal transition?
MD: We were fortunate because we started this early, six or seven years ago, so we’ve built more from the ground up than most. I don’t want to minimize it. Change management is real and it’s hard. But because we weren’t reinventing a large legacy organization, we avoided a lot of that friction.
What I’d point any operator to, regardless of size, is the employee side of the equation. Our people are happier and more empowered in this system. Turnover is meaningfully lower and our employee NPS scores run above the industry. One of the original drivers for building the model wasn’t just property outcomes. It was employee outcomes. People aren’t stuck on an island doing dozens of different things and pulling their hair out. They get to focus, work within a team, and have a better experience overall.
BH: The model presumably behaves differently across asset types. A 300-unit Class A lease-up is not a stabilized 60-unit walk-up. Where does it break, and what do you do differently across asset classes?
MD: Smaller and mid-market assets are actually where we built this model, and as you know from your own experience, that stuff is often much harder to manage than the large institutional properties. That’s where we cut our teeth, broke some things, and made improvements, and it’s still a big part of our bread and butter.
But we built the model to flex by property type, asset, portfolio, and the goals of ownership. We keep the same concepts but curate the specific roles and outcomes to what a given property needs. A Class A-plus lease-up needs a different emphasis on resident experience than a stabilized building, so we tailor that while retaining the centralized layer underneath.
The flexibility extends to staffing. We’re not stuck assigning an integer number of employees to a site. If a building needs a 1.5-person maintenance allocation, we can structure and price it that way rather than forcing it to be one or two.
That said, it raises fair questions from owners. How do I know I’m getting what I’m paying for, and how do I know the allocation is fair when I don’t see an employee on my own payroll? We get those questions constantly, and it’s genuinely an education process. The honest answer is that we’re delivering a service designed to drive outcomes. If an owner just wants to pay for headcount with no clear path to outcomes, we’re not the right fit. If they want a service program engineered to hit the outcomes they’re underwriting, we can design that for them.
BH: For an owner-operator running tens of thousands of units who wants to move toward functional specialization, what are the first moves, and what’s the biggest mistake operators make when they try to copy a model like this?
MD: The first thing you can bite off fairly easily is leasing. That’s where I’d attack first, and then you back into the rest. I’d also set up a more on-demand maintenance program early. Those are the two I’d tackle first, and from there you can redesign the more specific management roles around outcome-based work.
The biggest pitfalls are about people, not tools. It’s a real education for your team. When we hire managers out of a traditional model, it takes time to get them to understand the ownership required: owning specific outcomes, curating roles, and working within a team instead of running their own island. There’s a genuine learning curve and a reset in how people think about their jobs.
The other big pitfall is starting with technology instead of structure. That’s where I see most operators run into walls, and we’ve hit it ourselves when we expected software to deliver a result that actually required a structural change first. It’s people, then process, then systems. That idea goes back to Henry Ford. It’s nothing new, but it’s true. And for an owner-operator, this is much easier to do than for a third-party manager. If they’re not moving in this direction, I’d argue they’re leaving a lot on the table.
BH: What are you excited about sharing at Blueprint this year?
MD: What I’m most excited to get across is that this isn’t a technology story first. It’s an operating-model story. The industry is racing to bolt AI onto the same generalist structure and call it centralization, and I want to make the case for the opposite sequence: redesign the roles around specific outcomes, put the right people in the right seats, and then let technology extend a model that already works.
I’d love to walk through what functional specialization actually looks like on the ground, from the professionalized leasing function to the on-demand squads to the flexible staffing, and share some of the hard numbers behind it, like the portfolio where we drove occupancy up roughly 35% in six months. And I’m looking forward to the debate, because plenty of operators still believe the market is what the market is, and I think Blueprint is the right room to make the case that an operator’s whole job is to drive alpha.
