Strategy guide
Build vs buy for CRE technology: how to decide.
Every institutional CRE firm hits the same fork: license a SaaS platform, build in-house, or engage a specialist to deploy infrastructure into your cloud and operate it alongside your team. This is how to pick the right path and avoid the middle.
TL;DR
Buy a platform when your workflows match the platform's assumptions and you are willing to live inside their data model. Build in-house only if you have permanent engineering capacity and 12+ months of runway. For most institutional CRE firms, the right answer is a specialist partner that brings a pre-built CRE shell, customizes it to your firm, and operates it alongside your team, multiples faster than a pure consultancy build.
Option 1
Buy a SaaS platform.
Platforms like Yardi Elevate, Juniper Square, Dealpath, and VTS give you working software on day one. You pay per seat or per property, you log in through their portal, and you live inside their data model. For simple use cases on common workflows, this is the fastest path to a functional system.
When buying wins:
- Your workflow is standard and the platform's opinionated model fits.
- You need functional software fast, not a multi-month build.
- You do not want to own any engineering or ongoing operation.
- You are under $250M AUM and the total platform cost is manageable.
When buying fails:
- Your fund structure, asset class, or LP mix has custom economics the platform does not model.
- You need your data out of the platform to feed Excel, BI, or AI agents and the export is partial or slow.
- You want to layer AI agents on top of your data and the platform keeps the data locked in.
- You scale past the platform's sweet spot and pricing compounds faster than value.
Option 2
Build in-house from scratch.
Some firms decide to hire a head of data and a small engineering team and build everything from the ground up. In theory, you get full control, full customization, and no platform fees. In practice, most in-house CRE tech projects take 12 to 18 months to reach parity with what a platform delivers on day one, and a meaningful percentage fail to ship at all.
When building wins:
- You are a large enough firm to support a permanent engineering team (typically $2B+ AUM).
- Technology is core to your strategic thesis, not operational overhead.
- You have 12 to 18 months of runway before the system needs to produce value.
- You are prepared to lose and rehire engineers at industry churn rates.
When building fails:
- Your engineering team has never worked in CRE and learns the domain while building.
- Scope expands faster than the team can ship, and the first release never lands.
- The head of data leaves and the project stalls while you backfill.
- You fund the build but never allocate operations budget, so it degrades after year one.
Option 3
Engage a specialist to deploy custom infrastructure into your Microsoft Azure tenant, operated alongside your team.
The third option is a specialist firm that brings a templated CRE stack, tailors it to your firm, and deploys into your Microsoft Azure tenant. You get platform speed with custom control. You operate the system yourself or have the specialist operate it alongside your team.
When this wins:
- You have custom economics (fund structures, asset classes, LP classes) a platform does not model.
- You want your data, models, and workflows in your own Microsoft Azure tenant, not on a vendor portal.
- You want to deploy AI agents and reporting that sit on top of your data.
- You do not want to build and maintain an in-house engineering team.
When this fails:
- You pick a generalist consultancy that has never done CRE, and you pay for them to learn your domain.
- You pick a specialist with no templated stack, so every deliverable is genuinely custom and the build takes as long as in-house.
- You skip the managed-services contract and the system degrades after handoff with no one to evolve it.
Decision framework
Which option fits your firm?
| Factor | Buy a platform | Build in-house | Specialist partner |
|---|---|---|---|
| AUM fit | Under $250M | $2B+ | $250M to $5B+ |
| Time to value | Fast | Slow (multi-quarter) | Multiples faster than in-house |
| Customization | Low | High | High |
| Data control | In vendor platform | In your cloud | In your cloud |
| Engineering capacity needed | None | Permanent team | None to light |
| Ongoing cost pattern | Per seat / per property | Salaries + cloud | Cloud + committed engineering hours |
| AI agent layer | Platform-dependent | Build it yourself | Included via Operate |
Bottom line
The middle is gone. Pick one of the three cleanly.
The pattern that fails is the half-commit: buying a platform and then paying consultants to customize it heavily, or hiring one engineer and asking them to build an enterprise data stack from scratch in a quarter. Those efforts usually produce the worst of both worlds: platform lock-in with custom-build cost, or custom-build complexity with no team to maintain it.
If you pick a platform, fully buy the opinion. If you build in-house, resource it like the long-running program it is. If neither of those fits, pick a specialist partner whose templated stack and forward-deployed model collapses the build dramatically.
That is the case for the third option. For most institutional CRE firms with custom economics and AI ambitions, it is the only one that delivers platform speed with the control and AI extensibility of custom infrastructure.
Tell us what's broken. We'll show you what fixing it looks like.
Walk us through your stack and your pain points. We map the engagement and outline the path from there. Paid Discovery scopes the SOW, blueprint, and timeline.