Solving operational challenges and accelerating insights through data integration

Understanding a data problem versus knowing how to fix it

Data integration challenges have long plagued the CRE industry. With property management data stored in one system, ARGUS data stored in another, and forecast assumptions living in a spreadsheet that three analysts have saved under three different names, the promise of data-driven decision-making remains just out of reach. "There's been this never-ending problem in the industry, the data is just everywhere," said Vlad Starobinets, Senior Director, Client Enablement at Altus Group, who has spent 25 years working on precisely this challenge.

While the challenge is well understood and frequently discussed, few CRE firms seem to have a clear view of what fixing it actually looks like, and what becomes possible on the other side.

At a breakout session during Altus Connect, Starobinets was joined by Alyssa Mencel, Senior Associate in FP&A at BKM Capital Partners, an early adopter that made the move from a legacy ARGUS Enterprise environment to ARGUS Intelligence, for an in-depth conversation about the journey from disconnected, manually assembled data to a connected, decision-ready environment. The picture they painted was neither a quick fix nor an overwhelming overhaul; it was a defined starting point that led to a compounding payoff.

The problem is universal, and the cost of leaving it alone keeps growing

BKM Capital Partners operates as an investor and operator of light-to-mid-bay industrial assets. Before integrating their systems, the firm's analysts were doing what analysts at most CRE firms do: pulling ARGUS data manually, reconciling it against Yardi, and assembling forecasts in Excel. "A lot of different versions of our forecast were combining Yardi data and ARGUS data in Excel," Mencel said. “It was taking the analysts a lot of time to process, grab the ARGUS data, put it in the models, and also to visualize it in our various reporting for JV partners."

But what has become the industry norm can no longer afford to be the industry standard. At some point, “the way things have always worked” inevitably stops working. As portfolios grow, as reporting requirements multiply, as JV partners and fund investors ask for more transparency, the cost of manual reconciliation stops being a background annoyance. Instead, it starts eroding analyst capacity, and slowing reporting turnaround times.

The data fragmentation problem also runs deeper than most senior leaders realize. A significant portion of the data most important to CRE decision-making never makes it into any transactional system. Instead, it lives in spreadsheets, email threads, and local drives where it cannot be analyzed or acted on at scale. And inside the systems that do exist, the accumulation of redundant and outdated models compounds the problem. On average, between 18 and 20% of models in most legacy ARGUS Enterprise environments are identical duplicates — copies made for a scenario run, a sensitivity analysis, a deal that didn't close — that no one ever cleaned up.

Stop asking how to move data and start asking what it can do instead

The integration conversation in CRE tends to focus on the mechanics: how do you get data from one system to another, how do you handle format mismatches, how do you manage the volume? These are real questions, but Starobinets argues they are the wrong place to start.

"We should really start asking: what can our data do for us?" he said. "It's not about moving it. Moving it is easy."

The harder and more valuable question is what the data needs to do once it gets where it's going. Who needs to see it, in what form, at what frequency, and for what decision? The business use case has to precede the architecture. Firms that mobilize their data without answering those questions tend to end up with a well-connected system they can't act on or the data equivalent of building a highway that goes nowhere.

The foundational work – consistent property addresses, correct model tagging, normalized and deduplicated files – has to happen before any downstream value is realized. For example, if a property's address isn't consistent across systems, there is no reliable way to associate data across models over time, leaving the intelligence layer with nothing clean to work with. "If you start with garbage, you end up with garbage," Starobinets said. That groundwork is unglamorous and it doesn't produce an immediate output, but skipping it means every analytical capability built on top of it is built on a shaky foundation.

What the journey actually looks like

For BKM Capital Partners, the move from ARGUS Enterprise to ARGUS Intelligence was never just a platform migration. From the outset, the goal was to use the transition as an opportunity to build the integration infrastructure that would connect their ARGUS data with Yardi data, making it queryable and enabling the kind of reporting and visualization that manual workflows had never allowed. That goal shaped every decision that followed.

With the support of ARGUS Intelligence Platform Data Integration Services team, the data integration process began with an honest assessment of where their data actually lived, including years of acquisition models for all deals, both won and lost. As it turned out, that historical data mattered. "We actually wanted everything," Mencel said, "because we wanted to be able to look back and see what we were forecasting previously, even with assets that we didn't get. Sometimes those assets come back on the market. It gives us a better starting point."

Because BKM came in as an early adopter of the ARGUS Intelligence Platform with their full historical model library loaded (not just current assets), the process took approximately six months. That timeline is the exception, not the rule.

The custom build required to accommodate that request has since been made reusable. A more recent client engagement, starting from a standard configuration, went from first conversation to live data flow in three weeks. A typical client with 45,000 models in their legacy ARGUS Enterprise system would be migrated to ARGUS Intelligence in approximately two weeks. The perceived complexity of data mobilization almost always exceeds the actual complexity. Firms do not have to mobilize everything at once, and whether a firm starts from today or wants to bring its full model history along, the entry point is more manageable than it looks.

How BKM built its data environment

BKM's technical setup is instructive. With their models now moved across into the ARGUS Intelligence Platform, BKM worked with a third-party data partner, who had already built a cloud-backed SQL environment that houses Yardi data, to now flow in their ARGUS data in a unified, queryable structure. Analysts who need a cashflow don't have to navigate raw data files or interpret unfamiliar IDs because the data is structured for use. Updates from ARGUS Intelligence flow incrementally, not as full-file replacements, keeping the system current without creating a maintenance burden. To reconcile property codes between Yardi and ARGUS, which don't always match neatly, BKM used the external ID field in ARGUS Intelligence to create a mapping table that bridges the two systems. Where the property codes align, the match is automatic. Where they don't, the client table resolves it. It’s the kind of practical, thoughtful solution that comes from a team that understood their data before they started moving it.

Instrumental to the third-party integration work is ARGUS Intelligence’s JSON files, the same files used internally to power its own fund management and portfolio reporting capabilities. For firms evaluating whether the integration is production-ready, that's an important point: the architecture being offered to clients for data integration is the same one running functionality in ARGUS Intelligence.

What becomes possible

The operational upside BKM is now realizing didn't happen by accident. It’s the direct result of approaching the integration journey with ARGUS Intelligence as an infrastructure investment rather than a system swap, one designed from the start to connect data, not just move it.

The return on their investment first manifested as time saved by analysts. Work that previously required manual assembly across multiple systems gets automated. Reporting that took days gets reduced to hours. Analysts reclaim time from data gathering to invest in actual analysis.

For BKM, the connected infrastructure is already enabling a roadmap that would not have been practical before. The team is rebuilding their forecast model to draw cleanly from both Yardi and ARGUS without manual intervention, and developing new visualizations and reporting. "We’ve been growing pretty rapidly, and unless we want to hire a team of 20 additional analysts on the portfolio management side, we're working on automating some of the processes there," Mencel said. "We're also developing new dashboards for executives and also for our capital partners to use." The transparency that comes with cleaner, more consistent reporting is not just important internally but externally, as JV partners and fund investors are increasingly asking harder questions, and the ability to answer them quickly and confidently is itself a competitive differentiator.

The longer-term picture

The longer-term ambition reflects where BKM is heading as a business. As the firm has grown into portfolio acquisitions, the analytical requirements have grown with them — and the connected infrastructure is what makes those requirements answerable. "We've started to grow more into portfolio acquisitions and wanting to report at that level," Mencel said. "There are other data points that we like to connect at that level that's not a standard report." The integration wasn't built for where BKM was. It was built for where they are going.

Where to start

The most important step in the integration journey happens before any data moves at all. "Get the correct base of data," Starobinets said. "Normalize your files. Clean up the duplicates. This work starts before data movement begins."

In practice, that means auditing what is actually in your legacy valuation solution, understand how many models exist, how many are current, how many are identical copies or outdated scenarios from analysts who are no longer with the firm. It means making sure property addresses are consistent, and deciding what you actually need to carry forward versus what can be archived.

That work is a data discipline decision, not a technology decision. But how well it's done determines whether everything that follows — from the integration to the automation, the dashboards, and the analytics — produces something a firm can act on or simply adds a new layer of complexity to an existing mess.

"The challenge isn't new," Starobinets said. "The way we solve it is." The firms that treat integration as a capability to build rather than a project to complete are the ones that end up with data infrastructure that compounds in value over time instead of requiring a rebuild every few years.

Want to be notified of our new and relevant CRE content, articles and events?