Occupancy, yield and risk at a glance in Power BI for a commercial property group

The portfolio picture that lives in twenty spreadsheets

A commercial property group runs on a handful of numbers, and almost none of them are easy to see across the whole portfolio. How full is each building. What is it really earning once incentives come out. What is it yielding against today’s value. And when do the leases roll. The data exists for every asset, but it lives in that asset’s own management system, rent roll or spreadsheet, in that asset manager’s own format.

So the portfolio view is hand-built. Once a month someone copies figures out of each asset’s pack, pastes them into a master spreadsheet, and reconciles the differences until the totals look right. By the time the board pack is ready, it is weeks old, and the asset-level detail behind it has already moved. Anyone who asks a follow-up question sends the team back to the source files.

The harder problem is that the numbers do not even mean the same thing from one asset to the next. Is a holdover tenant counted as occupied or vacant. Does “rent” mean the face rent in the lease or the net effective rent after a fit-out incentive. One asset manager answers one way, another answers differently, and the portfolio roll-up quietly mixes them. The board reads a single occupancy figure that was never calculated consistently underneath. For numbers that feed investor reporting and valuations, that is a real exposure, not a cosmetic one.

Why Power BI, and the modelled layer beneath it

The aim is one portfolio view that the board, the investors and the asset managers all read from, with each measure calculated the same way every time. We headline these builds on Power BI for three practical reasons. A semantic model defines occupancy, net effective rent, yield and WALE once, so every report agrees. Row-level security lets each asset manager see their own buildings while the group sees the whole portfolio. And a nightly refresh puts current numbers in front of people without anyone rebuilding a spreadsheet.

Power BI is only as good as the data beneath it, so the architecture shapes that data first. A governed data layer, built on Snowflake or Microsoft Fabric depending on the group’s existing stack, ingests each asset’s lease, rent-roll and valuation data and models it into a clean structure. Power BI reports from that single source rather than from per-asset extracts. The semantic model defines each measure once, so an asset manager and the board open a report and see the same occupancy and the same WALE, calculated the same way.

The model sits over the spreadsheets, it does not replace the systems each asset already uses. Each asset keeps its own management system. The modelled layer maps every source into one agreed shape, which means a new asset is onboarded by mapping its data in, not by rebuilding the reports. That separation also lets the same governed layer feed valuation and investor reporting later, not just today’s dashboard.

Building it, and where it got hard

The friction in portfolio analytics is rarely the dashboard. It is the definitions, and on this build one issue stood in for the rest. The same word meant different things in different assets.

Early in the work the occupancy figures looked plausible per asset but would not reconcile across the portfolio. The cause was definitional, not technical. One asset counted a holdover tenant, one still in the building past lease expiry, as occupied. Another counted that space as vacant because the lease had ended. “Rent” was worse. Some assets reported face rent straight from the lease, others netted off the incentive, so a building with a large fit-out contribution looked far stronger than it was earning. Rolled together, the portfolio occupancy and the portfolio rent were an average of inconsistent things, which is to say they were wrong in a way no one could see.

The fix was a single semantic model that defined each measure once and a mapping layer that reconciled every asset’s lease data into it. We agreed one definition of occupancy, one of net effective rent, one of yield and one of WALE, then handled the awkward cases explicitly. Holdover tenancies were classified one way across the whole portfolio. Incentives were amortised so net effective rent meant the same thing in every building. Lease expiries and incentive end dates were modelled together, so a leasing event and the incentive attached to it lined up. None of this is glamorous. It is the difference between a portfolio number the board can sign off and one the asset managers quietly distrust.

Two constraints shaped the rest. Data governance mattered, because these figures feed valuations and investor reporting, so the modelled layer is the agreed source of truth and changes to a definition are made in one place. And because asset managers should see their own buildings without seeing the whole group’s commercials, row-level security was built in from the start rather than bolted on.

What changed

In a representative build the monthly per-asset spreadsheet pack became a single Power BI view of the portfolio, refreshed nightly, covering occupancy and vacancy, net effective rent and yield, and WALE across every asset. The group stopped rebuilding the picture by hand and started reading current numbers instead of month-old ones. Because each measure was defined once, the board pack reconciled with the asset-level detail rather than drifting apart between teams.

Modelling lease expiries and incentive end dates together surfaced exposure the old packs had hidden. A cluster of leases expiring in the same window, spread across several assets and never seen side by side before, became visible early enough to plan around. Arrears and outgoings recovery sat on the same view, so income at risk was no longer buried in an individual asset’s file.

These figures are illustrative. They describe the pattern we see rather than a published result for a named group. The shape is the point. The numbers that were always there, scattered across each asset’s records, start reaching the people who set strategy, report to investors and negotiate the next lease, calculated consistently and current enough to act on.

Where this fits

Portfolio analytics is one application of our Data Insights and Analysis service, built on Power BI, for a commercial property group. It is a contained, high-return starting point, because the data already exists in each asset’s records and the value comes from defining the measures properly and getting them onto one trusted view. If your portfolio picture takes a week to build and the definitions argue with each other, the place to start is to map your lease, rent-roll and valuation data and agree what occupancy, rent, yield and WALE actually mean across the group.