The phone
Imagine you're trying to track down the source of like a massive roach infestation in a large apartment complex nightmare scenario, right? Totally. So where do you look you check the trash rooms? Sure, you check under the sinks you look for the tenants who maybe leave food out on the counters But you probably wouldn't look at a perfectly clean virtually unused dishwasher
I know it's completely counterintuitive an unused appliance doesn't exactly scream You know ground zero for a pest outbreak exactly But today we're gonna look at why that perfectly clean dishwasher is actually acting as a secret subterranean highway for pests It is wild and we're looking at the person who figured that out because today we have a very specific
Mission and a very specific audience. We are doing a one-of-a-kind of bespoke deep dive Custom tailored for one single listener just one Eileen O'Malley. So Eileen welcome to your very own deep dive Yeah, it is a genuine privilege to be digging into your world today Eileen and what a complex world it is, right? Let's let's set the stage here Eileen
You are the director of operations at Situs group in Denver, right which is an owner operator setup Yeah, we are talking about roughly 550 apartment units and about 500 investors. Oh man That is an enormous surface area. You've got physical buildings. You've got Financial capital and you know, just a ton of human emotion
It's a massive ecosystem and the raw material we are using to explore this ecosystem today is incredibly unique We have the raw transcript of a five-hour whiteboard session five hours five hours. This took place on September 1st 2026 it was a deep completely unscripted sit-down between you Eileen and Federico chop a feed from PropFlow. Yeah
43,000 words of you and feed at a whiteboard It's like it's like getting the flight recorder data from a master operators brain It really is and early in this transcript Eileen you described your specific role in a way that just like locked everything into place for me You called yourself the center of the bowtie. I absolutely love this visual. Let's kind of unpack the mechanics of that bowtie
Yeah, okay Yeah So on the left side you have the wide end the ownership the capital the high-level strategy, right? You have Hugo Noel and Matt right and they are looking at yields interest rates overall portfolio health
They funnel every piece of good news bad news and financial pressure right into the center And then on the right side of the bowtie the other wide end you have the execution the ground game Exactly. You have the people dealing with the physical reality of the buildings. That's Courtney lovely Davey Miriam Lee Chauncey and your intern drew right and right in the middle
The literal knot of that bowtie is Eileen all the pressure from the top and all the chaos from the ground Just converges in this one tiny high friction space Yeah you are the translation layer like you have to take a mandate from the owners something like reduced turn costs and
Translate that into actionable non demoralizing steps for Courtney and Davey on the ground, which is not easy No, and conversely you have to take a tenants emotional complaint about a leaky pipe and Translate that into defensible objective data for Hugo Noel and Matt which requires a very very specific Toolkit and reading through this five-hour session Eileen your DNA is just all over these problem-solving methods
Okay, you come from an ad agency background Yeah, which you know completely explains your obsession with visual data and tracking systems totally plus you have this incredible Dry sharp sense of humor which you clearly need to survive property management. It's 100% But most importantly, there's this dual nature to how you operate
You are completely allergic to excuses absolutely allergic, but you have this deep protective warmth for your team You make sure they get the credit by name Yeah, you see a problem you engineer the solution and you throw absolutely everything at it to make it work But instead of just talking about your philosophy, let's look at how it actually manifests in the physical world. Let's do it
Pest maps and dishwasher pipes
Let's start where your longest most energized riff in the entire whiteboard session started the absolute bleeding edge of operation The phone line the phone. Oh my gosh. This part of the transcript is amazing You were passionately advocating to Fadi that there needs to be a live effective human voice answering the phone 24-7 right especially for new leasing leads your central argument was simple quote
Things are slipping through the cracks it seems like a totally simple premise right answer the phone, right, but the logistics of this in a 550 unit portfolio are just Staggering if someone calls at 7 p.m
On a Wednesday or like a Sunday afternoon and it goes to an automated tree or a voicemail Yeah Well, then the follow-up becomes entirely dependent on the administrative discipline of whoever checks the messages the next morning Let me push back on that a bit though because you know, we're in 2026
Why not just use an automated AI routing bot for after hours calls? Wouldn't that be cheaper than paying a live call center? I mean in theory sure But the reality of property management is that the initial data input from a tenant is almost always messy, right? It's emotional very emotional and imprecise
So a rigid automated system or even a poorly trained third-party call center Forces that messy human reality into the wrong boxes and if the data is corrupted right at the entry point Exactly the physical dispatch of a vendor the next day is just doomed. Okay, which brings us to the live demonstration To prove exactly what you meant about the entry point being broken Eileen
You literally put the after hours maintenance call center on speakerphone right there in the room with feed I love this. It is a master class and watching a system fail in real time. It really is So you dial the number an agent named Alex answers you state the problem incredibly clearly He said I have these yellow jackets outside in the front at 1965 South High Street
Now let's just pause and look at the cascading data failure that happens next please first the agent miss yours yellow jackets, right? She searches the ticketing system for the word bees which I mean sounds minor bees wasps yellow jackets. They all sting but Operationally, it's a massive distinction. Oh because of how you treat them Exactly the vendor you dispatch to safely relocate a protected hive of honeybees is
Completely different from the exterminator you send to just spray a hostile yellow jacket nest, right? Totally different vendor and a good system or the kind of AI architecture Eileen actually wants to build would understand the semantic Relationship between those words and ask a clarifying question, right? Like what kind of bees but the human call center agent is rigidly searching the exact string
BES finds no previous ticket and just gets confused with the breakdown gets so much worse you give her the address 1965 South High Street you say it twice twice But because she is relying on the caller ID that pulled your personal cell phone number Her screen pops up a completely different property a single house you man it on the side over at 446 Galapagos
Yes, and despite you explicitly telling her the address is South High Street She associates the new work order for bees to the Galapagos address unbelievable You literally looked at feed and said I can't even make this up and what blew my mind here is that you caught the error Instantly because she didn't even have to look at a chart
No You have the entire nested phone tree like the routing logic for every single property in the status group portfolio Memorized which is insane You could see the invisible architecture of where the call is failing in real time
And this just illustrates the immense friction at the center of the bowtie You have an exterminator getting a work order for honeybees at an address where there are no honeybees You pay a trip charge for a vendor to drive to the wrong zip code The tenant at South High Street is furious because the yellow jackets are still there right the owners are annoyed by the wasted expense
And all of it stems from a five-second miscommunication at the very edge of the system It perfectly sets up why you are so obsessed with tracking the physical reality of a problem Which brings us back to the dishwasher well, this is the best
Receipts over anecdotes
Let's talk about what I think is the most impressive piece of operational detective work in this entire transcript the pest outbreak mind map This is where Eileen's ad agency background really collides with the gritty reality of property management So you had a problem Certain buildings were experiencing these clusters of pest infestations now Cetus group uses an AppFolio, right?
Which is a massive enterprise property management software Yeah, but you were frustrated with how AppFolio handles this data, right? Because at folio like most CRMs Handles data in a tabular format. It essentially a spreadsheet Okay, you can pull a report that says unit 101 unit 205 and unit 312 had pest work orders in the last 60 days
But tabular data cannot show you spatial relationships because it's just a list exactly It doesn't tell you if those units share a wall or a ceiling or a plumbing stack and pests don't care about spreadsheet rows No, they do not they care about physical proximity So you didn't just accept the software's limitations you teamed up with your intern Drew and I love that you credit him so heavily
For this you had Drew physically walk the properties Drew did the light work He mapped the physical layout of the buildings where the doors are how the vent stacks run which units sit directly above or adjacent to others and Using that physical map you took the raw AppFolio data and built an adjacency model You build a system that flags neighboring units the absolute second one unit reports bug
It's so smart in the transcript you showed Fede how it works in real time You looked at the dashboard and literally said I'd recommend cottonwood 208 and 210 keep an eye on 207 That is predictive operations Yeah
Instead of waiting for the tenant in 208 to call in a panic two weeks later You were tracking the physical migration of the pests But the most incredible part of this modeling is what it finally proved the dishwasher pipe epiphany Yes, you had been carrying around this hunch for years about why pests were moving in these highly specific vertical patterns
So let's look at the mechanics of apartment plumbing in a multi-story building the kitchen sinks and dishwashers for stacked units Often share a primary vertical drainage pipe like a wet wall Right a wet wall or a plumbing sack and if everyone is using their appliances normally water flushes through the system regularly But through your demographic knowledge of the tenant base you realize something
certain tenant populations particularly within some immigrant communities in these buildings Culturally prefer to hand wash their dishes, right? They don't use the mechanical dishwasher to clean They use it as a drying rack or for dry storage which creates an unintended mechanical failure See a dishwasher is plumbed to retain a small amount of water in its lower pump and drain hose
To keep the seals from drying out. Okay, if the machine is never run that water just sits there It becomes stagnant standing water and what do roaches need more than anything else water exactly You realize that these unused dishwashers were creating stagnant water reservoirs Directly connected to the shared plumbing stack the pipes had literally become a dark moist
Super highway for pests to commute between units you called it wild and insane in the session and it really is But notice how this changes the entire operational response Right because a standard operator sees a tabular spike in pest control invoices assumes the exterminators just doing a bad job And yells at the vendor or they assume the tenants are just dirty exactly they make an emotional or anecdotal assumption
But by mapping the data physically you found a structural Mechanical root cause yeah, you can now solve the problem at the infrastructure level by doing what well perhaps by capping Unused dishwasher drains or running flush cycles during preventative maintenance rather than just blindly spraying more poison It is the ultimate example of catching a problem with receipts
You didn't argue with people's opinions about the pests you just built a map that proved reality Yeah, and this transitions perfectly into how you handle the ownership side of the bowtie
The vendor whisperer
Let's look at how you handle the financial data because this is where that scrappy founder energy really comes out Oh scrappy founder energy inside an established corporate structure is a rare and highly effective trait very rare It usually manifests when someone is fundamentally allergic to making decisions based on anecdotes Yes
for example The leads are trash myth your team on the ground was telling you that certain marketing leads just weren't converting The prevailing narrative in the office became that the leads themselves were low quality and this is so dangerous Because when a narrative like that takes hold
Ownership starts making budget decisions based on office chatter right like well. I heard they're bad leads, so let's cut it exactly They might cut a marketing channel that is actually performing really well Simply because the leasing agents find those specific prospects slightly more difficult to close So you stopped the chatter you built a lead attribution
Dashboard you force the conversation out of the realm of feelings like I feel like these leads are bad and into the realm of facts Right, but you didn't just track digital clicks you brought this tracking into the physical world the physical heat map This is again where the ad agency background just shines You set up physical leasing banners out on the street with unique QR codes
And you tied those specific QR code scans to a geographic heat map of vacancies across Lakewood, Arvada and Denver So cool You can literally watch the data as a prospective tenant drives past a physical building in Lakewood Scans the code and enters your digital leasing funnel
You tied a physical piece of vinyl on a fence to a digital conversion rate by doing this You can definitively prove to Hugo Noel and mad Exactly, which marketing dollars are yielding signed leases and what you're just generating noise you completely neutralize the anecdotes But the anecdote that cracked me up the most was the story about the Carr Street website the single-scroll solution
So you knew from your marketing days that multi-page? Clunky websites lose people you wanted a clean responsive single scroll website for the Carr Street property Right. The problem is at folio is default templates can be really rigid. So I have to play devil's advocate here Eileen you are the director of operations you manage 550 units
Why on earth are you spending a Tuesday afternoon? Personally using a site builder to design a property website, right? Shouldn't you delegate that? Yeah, you think so but delegation only works if the person you're delegating to understands the required architecture Okay, what do you mean? Eileen didn't just want a pretty website
She needed the architecture of that site to eliminate administrative friction The guest cards exactly when a prospective tenant fills out an interest form that creates a guest card a digital profile Containing their name contact info and unit preferences, right?
if the website doesn't talk directly to the CRM a leasing agent has to manually read an email and Retype that guest card into AppFolio which introduces the exact human error We saw what the phone call precisely it creates a massive data lag by building the site herself Eileen hardwire the intake forms so that a guest card completely bypasses the human data entry bottleneck and
Feeds straight into the leasing workflow. That's amazing. She invested a few hours on a Tuesday to permanently eliminate Thousands of future manual keystrokes for her team. It's engineering the workaround and you did this again with the language tagging in Colorado There are compliance requirements regarding communicating with tenants in their preferred language But shockingly AppFolio didn't have a simple built-in field to tag whether a tenant preferred Spanish or English
Which is just a massive blind spot for a software company to have especially in a diverse market like Denver, right? You cannot just guess someone's language preference based on their surname. That's unprofessional and honestly Operationally risky. So what did you do? You didn't wait for a software update You sat down with your maintenance tech and you manually combed through the entire rent roll unit by unit English
Spanish Spanish English you manually created the tags yourself So you could confidently send bulk communications that people could actually read and understand It just shows that when the system fails, you don't shrug your shoulders You brute force a solution until it becomes a standard procedure and this command of the details this absolute mastery of the numbers gives you
Incredible leverage when you are dealing with the outside world
The final boss
Which brings us to the segment of the whiteboard session that I am calling the vendor whisperer. Yes Negotiating with contractors and supply chains in property management is usually a zero-sum game Vendors make their margins in the gray areas of information asymmetry, right if a turn vendor You know the contractor who comes in to paint and repair a unit after someone moves out
Assumes you don't know the going rate for a standard repaint They will quote you a thousand dollars for a job that should cost five hundred because they assume you are way too busy managing 550 units to price check them, but they don't know Eileen you showed Fed your running rate card You actively track the exact down to the penny turn costs across five or six different vendors simultaneously
Yeah in the transcript you rattled off the numbers from memory. I have 650 550 550 620 and 290 You know exactly what a touch-up costs versus a full two coat smoker unit repaint and the mechanics of this are so powerful When a new vendor tries to inflate a quote you don't argue with them emotionally you just open the rate card You show them exactly what their competitors are charging you in real time, right?
You instantly collapse that information asymmetry You talk them right back down to market reality because you have the receipts but talking down a painter is one thing What you did with the kitchen cabinets is supply chain hacking on a whole different level the Amish made shaker cabinet arbitrage
This is a phenomenal example of isolating the friction points in a budget walk us through it Let's break down the standard operating procedure for most property managers a kitchen cabinet is rotting Usually because it's made of cheap retail press board that basically dissolves the second it gets wet, right? So the manager tells the turn vendor go replace the cabinets the vendor goes to a big-box store
Buys new slightly less cheap press board cabinets at retail markup Marked up the price again to the property manager and charges for labor So the owner pays a massive premium for a product that is just gonna rot again in five years. It's retail rot exactly So you completely bypass that entire system you set up a direct institutional discount with a wholesale distributor called Rue Shalhoux
You separated the materials from the labor you use the institutional account to purchase high quality hardwood shaker cabinets Literally Amish made wooden boxes that can withstand a flood and the logistics of this are amazing You have the premium cabinets delivered and then you hire the turn vendor strictly for their labor You just pay them to screw the boxes to the wall by disaggregating the supply chain. You are securing top-tier
Hundred year quality hardwood for the exact same total price that a lazy property manager pays for cheap rotting particle board It is brilliant and the fact that you find humor in this whole ecosystem makes it so much better You were joking with fate about pest control companies. Oh, this is so funny You said that doing admin for bug companies could easily be a lucrative side hustle because the exterminators are great at killing bugs
But notoriously terrible at filling out the state mandated compliance paperwork, right? You literally choose your pest vendors almost entirely based on their ability to file a piece of paper Because you understand the real risk a Slightly cheaper spray job isn't worth a massive compliance fine from the state of Colorado
You are managing the risk not just the task. This brings us to a really fascinating pivot point in the transcript
Five data sources, zero truth
We've talked a lot about systems data and hard numbers But property management is ultimately about human beings living in their homes Yeah Let's look at how you handle the emotional volatility of the job the empathy and the spine in the whiteboard session
You used a phrase that perfectly captures the psychology of this industry the cortisol spike metaphor Let's walk through the anatomy of a cortisol spike Housing is inherently emotional as you pointed out a tenant can be a perfectly rational even keeled professional in their daily life Yeah
But if the ceiling starts leaking onto their bed in the apartment they pay a huge chunk of their salary for Their brain enters a fight-or-flight state their cortisol spikes, which is completely valid But the problem is how that spike travels the panic tenant calls the leasing office and Unloads that cortisol onto a leasing agent, right?
The leasing agent who is likely younger and less experienced in crisis management Absorbs that panic their cortisol spikes and then the leasing agent transmits that panic down the chain to the maintenance team Turning a routine work order into a screaming emergency the entire operational chain becomes infected by the initial anxiety and
Eileen you view your job as the circuit breaker for that panic. You are the final boss Yes, you told a story about a situation at 1965 South High Street There is a delay in a unit turn The incoming tenant was getting anxious their cortisol was spiking and your leasing agent absorbed that panic and proactively offered the tenant a
Massive rent credit a huge financial concession for what was really a minor scheduling delay just to make the yelling stop They freelanced an incentive out of sheer emotional fatigue Exactly, and you stepped in immediately you stopped the concession you took over the communication Calm the tenant down with a clear professional set of options and resolve the issue without giving away the owner's margin
Now let's play devil's advocate again a critic might look at that and say well Eileen is just being a corporate enforcer She's prioritizing the budget over tenant empathy Why is stepping in to stop a rent credit actually a form of empathy? It's such an important distinction
True empathy in an operational environment isn't about panicking alongside the customer and it's definitely not about throwing money at a problem Inconsistently exactly if you give a massive credit to the tenant who yells the loudest You are inherently discriminating against the tenant who is patient and polite true empathy is providing a calm Predictable and guaranteed standard of care for everyone and you prove this to feed with the stove burner cure period in Colorado
There is a legal requirement if a stove is broken the landlord has a 24-hour cure period to fix it It's a strict habitability law, right? You were reviewing open work orders and saw a ticket where one single burner on a four burner electric stove Was taking a little too long to heat up. Okay, it wasn't completely broken
Just slow and you noted to your team that this specific tenant was older incredibly polite and would probably never complain or escalate the issue to a legal threat a Reactive operator would look at that and say great push that work order to next week They won't mind and we have louder fires to put out
But you insisted that the burner get fixed within the 24-hour window with the exact same urgency as if a lawyer had drafted a demand Letter you said that's not a reason to not fix this person's burner They should be treated no differently the standard is the standard that is what we mean by having a spine your empathy actually requires you to be Uncompromising about the operational standards you protect the silent tenant just as fiercely as the screaming one
It's a beautiful philosophy, but here is the brutal reality We have to confront to maintain that objective standard you need perfect data Yeah You do you need to know exactly when the stove broke when the vendor was called and when it was fixed and that?
What she would build
Leads us to the most maddening section of the entire five-hour whiteboard session the messy reality of software Eileen you delivered a line to feed that should be etched in stone above every property management office in the country Your data is only as good as the people updating it We tend to think of enterprise software like AppFolio as a flawless machine
The owners look at the dashboard and assume it represents absolute truth But you showed feed that there are five different data sources in your office and zero absolute truth Let's roleplay the lifecycle of a lease to see how human entropy destroys perfect data Okay, I'll be the tenant. I find the apartment I apply and I get approved great. I'm the leasing agent
I verbally tell you that you're approved you move your boxes in but because the office phone is ringing and I'm dealing with a cortisol Spike from someone else. I forget to go into AppFolio and click the final countersign button on your digital lease file So what happens to the data the rent roll drifts the system still categorizes your unit as vacant even though you've been sleeping there for three weeks
The financial projections for the owners are now artificially deflated because the last click never happened Okay, but surely the agent will catch up eventually they try to which leads to the night shift to lease processing ritual you described You have an employee who gets so over stimulated by the chaotic energy of the daytime office that she simply stops Processing paperwork during business hours Wow
She waits until 6 p.m. When everyone leaves and the phone stopped ringing and then processes massive batches of leases all at once in the quiet it a completely Understandable human coping mechanism for stress Oh totally but operationally it creates a massive daytime latency If Hugo or Matt check the system at 2 p.m. They are looking at a ghost town
But the data destruction gets so much wilder you discovered that for years team members were taking digital leases out of the system entirely This blew my mind because at folios editing tools were slightly clunky employees were ripping the leases out Manually retyping the terms into external PDF editors on their desktops and then uploading the finalized flat PDFs back into the system Let's explain why that is catastrophic, please when data is native inside a CRM
It's trackable you can run a query that says show me all leases expiring in June But a flat PDF is just a digital photograph of a piece of paper The system can't read it by retyping the leases the employees had rendered the entire database unsearchable It's like building a beautiful digital library and then burning the card catalog
Yeah, and the communication data is just as fractured You told feed that roughly 50% of all tenant text communications vital records about repairs complaints and legal notices are living entirely On your employees personal cell phones half of the institutional memory of a multi-million dollar business is sitting in Courtney or Davies pocket completely inaccessible to you if they call in sick and
The final boss of data black holes the move out bill back process when a tenant moves out There are supposed to be a process to assess damages calculate security deposit deductions and build them You assume this system was functioning for five years But when Miriam went on maternity leave you had to step in and handle her cue and you discovered the void you realized that the entire
Process simply didn't exist. There was no systemic tracking for billbacks It was just a loose collection of habits and memory until you physically mapped it out Thousands of dollars in legitimate damage recovery were just vanishing into the ether This is exactly why you build your own custom dashboard
So I lean you aren't building them because you love spreadsheets you build them because a dashboard is a translation layer for human behavior Right you build systems designed to catch the drift flag the missing clicks and force the messy reality of human employees To align with the rigid requirements of the digital record It is relentless exhausting work to constantly tighten the bowtie
What comes next
Which makes so exciting to see you and Fede look toward the horizon in the latter half of the whiteboard session You start articulating a vision for what property tech should be and to be clear for the listener Fede was at the whiteboard as a Facilitator he was putting language and structure to Eileen's ideas. This wasn't a PropFlow sales pitch This was Eileen outlining her dream architecture. You named your absolute highest ROI dream. You called it renewal automation
currently the renewal process is highly manual and deeply fraught with human hesitancy an Agent has to decide how much to raise the rent worry if the tender will get mad Draft the letter and remember to send it you envisioned a predictive offer engine that removes the human entirely from the initiation phase You want an algorithm that automatically pulls in three massive data sets
First the property's financial pro forma the required yield the owners need Second real-time market comps across Denver and third crucially the physical condition of the specific unit Does unit 204 still have the old press board cabinets or did he get the new Amish hardwood upgrade? The AI synthesizes all of that and exactly 90 days before the lease expires it autonomously
generates and emails a perfectly tailored Dynamically priced renewal offered to the tenant no human hesitation. No forgotten emails It's a brilliant concept, but your second idea is the one that really signals a paradigm shift
You articulated a desperate need for a digital co-worker Yes Because you are suffering from dashboard fatigue the problem of dashboards Even the beautiful ones you build is that you still have to remember to log into them refresh the data and
Interrogate the system to find the problem. You are pulling the information You told feed you want a system that pushes information to you you want an AI that acts like a proactive employee Imagine sitting at your desk and instead of digging through spreadsheets the system taps you on the shoulder It pings you and says hey Eileen
You are 75% of the way through the turn budget for the lamb later building this month Should we hold off on non-essential painting or hey the snow removal contracts for 2027? Need to be bid out in the next 30 days You want me to email the vendors on the rate card think about the cognitive load that lifts off your shoulders if
An AI can proactively manage the informational flow and handle the routine alerts You are freed from the constant low-level anxiety of wondering what you forgot to check if the AI manages the data The center of the bowtie can spend 100% of her time Making the high-level human judgments that a machine can never do like negotiating a labor rate or intervening in a cortisol spike or walking the
Property with Drew to find the standing water you get to focus entirely on the physical and the strategic It's a vision of operations that moves from reactive defense to proactive architecture