Two years ago, a regional commercial real estate brokerage engaged Innovation Vista with an ambition that went beyond running cleaner operations; they wanted technology to change what the firm could put in front of a prospect. This spring, the 24-month measurement window we agreed on at the outset closed, and the meter read: $5M in attributed new revenue and $3M in efficiency gains, on a $1.5M all-in investment. A 433% ROI.
Notably, there is not a single AI model in this story. In a year when every technology conversation begins and ends with generative AI, this engagement is a reminder of where durable returns actually come from: data worth trusting, and the discipline to keep it that way. The firms that build this foundation now are the ones whose AI investments will pay off later; the ones that skip it will buy demos.
The Situation
Our client is a regional CRE brokerage with roughly $10M in annual revenue: large enough to compete for institutional-grade assignments, small enough that every pitch matters. Like most firms of its size, it was rich in data and poor in information. Deal terms, counterparty details, and property records lived across systems that didn’t agree with each other, and the brokers and analysts closest to that data had never been shown a reason to treat it as anything other than exhaust, a byproduct of doing deals rather than an asset in its own right.
Determing How to Measure Impact Before Creating It
Before the first system was touched, our consultant sat down with the leadership team to answer a question most transformations skip: how will we know this worked? Our approach to this discipline is documented in Measuring the Monetization of IT, and this engagement put it to a genuine test, because the CRO was, constructively, a skeptic.
His instinct was deal-by-deal attribution: maintain a “tech scoreboard”, and credit the transformation only with specific wins the team felt it could not have landed otherwise. An early compromise floated was crediting the technology with half of any improvement. Our consultant pushed back on both, with a simpler principle: “when something changes, look at what changed it.” The firm’s historical win-rate was a stable, well-documented baseline; in effect, a control group the firm had been running for years without realizing it. If the win-rate moved after the new capabilities entered the pitch, and nothing else material had changed in the market or the team, the delta belonged to the change. Leadership agreed that the incremental win-rate above baseline, applied to pipeline value, would be the revenue yardstick.
That pre-agreement mattered. When the results came in this year, there was no relitigating the methodology; there was only reading the meter.
The Work
The transformation had four layers, each a prerequisite for the next.
- External Data: We subscribed the firm to two industry data sources, giving it market context its internal records alone could never provide.
- A Centralized Datalake: Internal deal, property, and counterparty data was combined with the new external feeds in a single platform, the first time the firm’s institutional knowledge had ever lived in one queryable place.
- A Data Quality Program: This was the unglamorous heart of the engagement. We designed tooling that let staff easily identify and correct data issues (mismatched counterparties, inconsistent property records, incomplete deal terms) and stood up the program disciplines to keep the datalake trustworthy. In several datasets, error rates were driven effectively to zero.
- Advanced BI: On that foundation, we built executive dashboard capabilities that could generate insights on a prospect’s own data, analysis no competitor was showing them. What began as an internal analytics project became the centerpiece of the firm’s pitch, heralded across the company and placed front and center in its marketing.
The Results
Over the 24 months following launch, the agreed measurement horizon:
- Win-rate rose from 31.1% to 42.4%. Applying that incremental 11.3% to the $44.25M in pipeline the firm pursued over the period yields ~$5M in attributed new revenue, under the methodology leadership had endorsed before the work began.
- ~$3M in efficiency gains, derived from measured time savings as data errors, and the rework and research hours they consumed, were dramatically reduced or eliminated. The estimate was calculated conservatively with the CFO’s review and endorsement; the true figure is likely higher.
- Total investment: $1.5M all-in, inclusive of Innovation Vista’s fees, the external data subscriptions, datalake infrastructure, and all development work.
Net return: 433% ROI over the measurement window. And the window likely understates the result; the measurement period has closed, but the win-rate advantage and the efficiency gains have not. Both continue to accrue with every quarter that passes.
Per our client confidentiality policy, the client is not named and certain figures are rounded; the win-rates, pipeline value, and investment total are as measured.
From Back Office to Front of the Pitch Deck
The number worth dwelling on is not the ROI; it is the win-rate. Most IT investments at firms this size aspire to make operations cheaper or less fragile. This one changed the sales conversation. When a brokerage can open a pitch with insights about the prospect’s own portfolio that no competitor can produce, technology has stopped being a cost center and started being a reason clients sign. That is the Monetize tier of our Innovate Beyond Efficiency® framework in its purest form: IT and data creating revenue, not merely protecting it.
And while this engagement required no AI to deliver its returns, it has quietly delivered something else: readiness. Every capability described here (clean, centralized, trustworthy data, with the tooling and habits to keep it that way) is the exact prerequisite for the AI use cases now arriving in commercial real estate. When this client decides the moment is right, it will be building on rock while competitors discover their foundations are sand.
What We’d Do Differently
Candidly, the hardest part of this engagement was not the architecture; it was persuading brokers and analysts to take data quality seriously. They had never seen data treated as an asset, and early participation in the quality program reflected that. Adoption came only once the first BI outputs made the payoff visible to the people doing the correcting. Were we starting again, we would stage a small, visible BI win earlier in the timeline, before the full datalake was complete, to give the data quality effort a “why” from week one. The lesson generalizes: the technology is rarely the constraint; adoption is, and adoption is bought with demonstrated value, not mandates.
