How to Run 50+ Self-Storage Locations Without a Data Science Team
Running 50 self-storage locations the way you ran 5 doesn’t work. Running 50 the way a REIT runs 500 is overkill. The sweet spot for mid-market PE operators is a lean central team powered by the right software. No data science team required.
The Traditional (Wrong) Approach
As portfolios grow, the instinct is to hire:
- A data analyst to pull reports from Yardi, SiteLink, and Excel
- A BI engineer to build dashboards
- Maybe a “data scientist” to “optimize” something
Total cost: $300K-$500K annually. Output: dashboards that are outdated by the time they’re built and reports that take two weeks to compile.
The problem isn’t talent. It’s that self-storage data is structured. Occupancy, revenue, delinquency, move-ins, move-outs — these are standard metrics with standard definitions. The work is aggregation and presentation, not discovery. That’s a software problem, not a headcount problem.
The Lean Central Model
Instead of hiring a data team, invest in a platform that:
- Integrates with your PMS — Yardi, SiteLink, StorEDGE, QuikStor. One connection, all locations.
- Aggregates automatically — No manual exports. No copy-paste. Data flows in daily or real-time.
- Surfaces the right metrics — The 15 KPIs you actually need, not 150 you don’t.
- Generates reports on schedule — Weekly, monthly, quarterly. Emailed to LPs. No one opening Excel.
That’s it. No PhD required. No SQL. No “data pipeline” meetings.
What You Actually Need From “Data”
PE operators need five things from their data:
| Need | Old way | New way |
|---|---|---|
| Portfolio health at a glance | Call each manager, compile spreadsheets | Single dashboard, live |
| Investor reporting | CFO + analyst, 2 weeks | Automated, same-day |
| Underperformer identification | Monthly review, manual comparison | Flagged daily, ranked |
| Pricing decisions | Guess, or hire revenue manager | AI recommendations |
| Acquisition underwriting | Manual rent roll analysis | Upload, get NOI projection in minutes |
None of these require a data science team. They require connected systems and intelligent defaults.
The 10-Person Portfolio Team
For 50-100 locations, a lean central team looks like:
- 2-3 Portfolio/Asset Managers — Your eyes on the ground. They use the dashboard; they don’t build it.
- 1 Operations Lead — Vendor relationships, CapEx, manager oversight.
- 1 Finance/Reporting — Fractional or part-time. Handles audit, tax, anything the platform doesn’t.
- 0 Data Analysts — The platform is the analyst.
Your “data team” is a single platform that costs a fraction of one FTE and delivers more than three ever could.
Migration Path
If you’re currently reliant on manual reporting or a patchwork of tools:
- Pick a single source of truth — Your PMS. Everything else feeds from it.
- Connect the platform — API or export. Most modern PMSs support this.
- Kill the old process — Stop the weekly spreadsheet drill. Cold turkey.
- Train the team — 30 minutes. Show them the dashboard. Answer questions. Done.
Operators who’ve made this switch report going from “reporting day” (a full day of panic every month) to “reporting happens” (it’s just there). The time savings alone justify the platform cost. The quality improvement is a bonus.
The Bottom Line
You don’t need a data science team to run 50+ self-storage locations. You need a platform that does the work a data science team would do — aggregation, visualization, and basic inference — without the headcount. Build the team around the platform, not the other way around.