Making a property platform work for six very different people
Inceo Asset Intelligence Platform
Outcome
A structure that connects the information, the workflows and the decisions people actually have to make.
Inceo looks after UK property portfolios for international investors. All the information existed already. It was just spread across emails, spreadsheets and PDFs, so nobody could see the whole picture at once. I worked out who the platform was really for, then decided what it needed to know and what it should tell them.
- Company
- Inceo Group
- Product
- Asset intelligence platform, investor portal and staff workspace
- Role
- Business Development Associate
- Timeline
- October 2025 to September 2026
- Status
- Live
6
user groups mapped
17
data domains structured
5
portfolio KPIs defined
10
alert categories designed
Team and ownership
Said plainly, because it is what makes the rest believable.
I owned
- User group mapping and journey definition
- Data structure across 17 domains
- Portfolio KPI definition
- Alert categories and severity levels
I supported
- Scoping the first release with the delivery team
- Client onboarding and portal setup
- Reporting formats for investor reviews
I did not own
- Engineering and platform architecture
- Visual and interface design
- Valuation methodology and financial modelling
- Commercial terms with investor clients
The problem
Someone could own tens of millions of pounds of UK property and still not answer three basic questions. What is it all worth? What is about to go wrong? What needs me to decide something?
Valuations arrived by email. Lease dates arrived in a different email. Debt terms lived in a PDF somewhere. Compliance deadlines turned up at the point they became a problem.
That is not just annoying. Missed rent reviews, refinancing surprises and expired certificates cost real money, and the owner is usually the last to hear about it.
What made it hard
- Two audiences, one database. Investors want calm and simple. Staff need to edit everything. Those two wants fight each other constantly.
- Most owners live outside the UK. They own in pounds and think in dollars or euros, so every number needs converting in their head before it means anything.
- Trust was the real constraint. An investor had to be able to tell raw data from a calculated number from somebody's opinion. Otherwise the whole thing reads like marketing.
- Small team. Release one had to cover what genuinely changes a decision, not everything we could model.
My job
- Work out who the users are and what each one most needs to do, so the team had a clear scope for the first release.
- Structure the data. Decide what the platform needs to know about a property and how all of it connects.
- Define the headline numbers and the alerts that turn stored records into something worth acting on.
- I did not own engineering, architecture or visual design, and I did not decide how properties get valued.
What I found out
- Six user groups across two sides. On the client side: the owner, their investment manager, and their operations person. On the Inceo side: the asset manager, the analyst and the administrator.
- How often someone logs in mattered far more than their job title. The owner checks monthly and wants a headline. The analyst is in there daily and wants fast, tidy data entry.
- Five journeys carried nearly all the value. Onboarding, monitoring, reporting, alerts and decisions. Everything else could wait.
- The problem was never missing data. It was data that arrived too late, in the wrong place, for the wrong person.
What I decided
- 01
Two front doors, one source of truth
Instead of building two products, I scoped one database behind two experiences. Investors get a portal they can read. Staff get a workspace that publishes into it. Staff can preview exactly what a client will see before it goes live, which killed a whole category of trust problems.
- 02
Five numbers on the front page, everything else one level down
NAV, gross yield, LTV, WAULT and passing rent. The test was simple. Can an owner read the health of their portfolio in about a minute? Anything that failed stayed available, just not first.
- 03
Label facts, maths and opinions separately
Three layers. Data that staff entered. Numbers the platform worked out. Commentary written by a named person at Inceo. Keeping them visually different is a trust feature, not decoration. You always know which one you are reading.
- 04
Alerts that fire themselves
Ten categories, each triggered by the data, each with a severity and a linked property. The idea that mattered: tell the owner about the thing they did not know to ask about. Do not just hand them another list to check.
- 05
Put decisions inside the product
Anything needing sign off became a proper queue with approve and decline buttons and a record of what happened, instead of an email thread nobody can find six months later.
The thinking, not the screenshots.
These are the structures the work produced. That is the part that carries over to the next product.
Artifact · User group map
Six user groups, two audiences, one platform
The same records serve a principal who logs in monthly and an analyst who is in the system daily. Designing for the average of those two would have failed both.
Client side
International principal
Wealthy owner, usually living outside the UK, holding several properties
- Behaviour
- Logs in monthly, or when we tell them to. Wants the headline.
- Success
- I know what it is worth, what is risky and what needs me. In ten minutes.
Family office investment manager
The professional investor who supports the owner
- Behaviour
- Lives in the detail. Drills into properties, analytics, market notes.
- Success
- I can brief the owner with solid numbers without rebuilding a deck
Family office operations
Keeps the owner on top of paperwork and dates
- Behaviour
- Works the decision queue, sometimes acting on the owner's behalf
- Success
- Nothing slips on compliance or lease dates
Staff side
Asset manager
Looks after a set of investor clients and their properties
- Behaviour
- Starts at the triage screen, updates records, publishes notes
- Success
- I know what needs me today, across every client
Analyst
Keeps the data current and preps quarterly reviews
- Behaviour
- Fast, tidy data entry across every domain, every day
- Success
- Every property has a complete history behind it
Platform administrator
Handles access, setup and getting new clients started
- Behaviour
- Sets up new investor clients and switches on their portal
- Success
- A new client logs in and sees a full portfolio, not an empty one
Artifact · Journey map
The five journeys that carried the value
Scoping the first release meant naming which journeys mattered and accepting that everything else waited.
- 01
Onboarding
Create the client, enter the properties and history, switch on the portal
Administrator, analyst
- 02
Portfolio monitoring
Log in, scan the five numbers, check what is flagged, open one property
Principal, investment manager
- 03
Reporting
Raw data becomes calculated numbers, then published commentary
Investment manager, asset manager
- 04
Risk alerts
Something triggers, gets a severity, and shows up for both sides
Asset manager, operations
- 05
Investor decisions
A recommendation hits the queue, gets approved or declined, and is recorded
Principal, operations
Artifact · Information architecture
17 data domains, grouped by what they answer
Fragmented property and client information only becomes an operating model once every fact has one clear home.
Asset
1- Asset details
Leasing
3- Occupants
- Lease history
- Rent reviews
Financials
5- Valuations
- Debt history
- Financial transactions
- Capex planning
- Market benchmarks
Risk and compliance
4- Compliance certificates
- Insurance policies
- Asset reviews
- Exit scenarios
Portfolio
4- Commentary
- Decisions
- Health scores
- Watchlist
Artifact · KPI definition
Five numbers a principal can read in under a minute
Everything else stayed available, one level down. The constraint was attention, not data.
Portfolio NAV
What the whole portfolio is worth, and how that moved this quarter
Blended gross yield
What it earns as a percentage. Net yield sits one level down.
Portfolio LTV
How much is borrowed against what it is worth, and how close that is to the limit
Weighted WAULT
How long the rent is locked in for, on average
Annual passing rent
The rent coming in this year, under contract
Artifact · Alert logic
Ten triggers, each with a severity and a linked asset
The framing that mattered: surface what an owner did not know to ask about, rather than adding another list to check.
- Elevated
Lease expiry
A lease ending sooner than is comfortable
- Critical
Refinance window
Debt coming due while the room to manoeuvre shrinks
- Watch
Rent review
A rent review that is due now
- Elevated
Arrears and collection
Less rent arrived than should have
- Elevated
EPC risk
An energy rating or deadline that could bite
- Watch
Renters' rights
New rules that will affect this property
- Critical
Covenant deterioration
The tenant is looking financially shakier
- Critical
LTV breach
Borrowing is getting too close to the agreed limit
- Watch
Valuation movement
The valuation moved a lot this quarter
- Elevated
Compliance lapse
A certificate has expired, or is about to
Severity shows as intensity, not colour alone. Every band is labelled too.
What changed
- Seventeen data domains got clear edges and clear owners, so a property's full history became one record instead of a folder of loose files.
- Staff got a single screen answering one question: what needs me today? Built from expiring leases, rent reviews coming up, watchlist items and valuations going stale.
- Investor decisions got a paper trail, which turned reactive advice into an actual process.
Product outcome
A structure that connects the information, the workflows and the decisions people actually have to make.
What I learned
I assumed the hard part would be deciding what to build. It was deciding what to leave off the first screen. Every stakeholder had a number they swore was essential, and most of them genuinely were useful. But show an owner twenty numbers and they read none of them. Picking five, and being able to explain why the other fifteen sat one level down, did more for the product than anything I specified.
What I would do better
I built the alerts mostly from the platform's own logic and from watching how staff worked. Next time I would put a draft in front of two or three actual investors first, because severity is a judgement about someone else's appetite for risk and I was guessing at it. I also never defined how we would know the alerts were working. An alert nobody acts on looks exactly like noise, and I did not build any way to tell those two apart.