Capital allocation strategy
Where should this year's capital budget go?
Six capital deployment scenarios, generated and ranked against your own portfolio data — with the quality of that data scored and shown to you before a single recommendation appears.
Every remodel programme, every closure sequence and every renewal strategy runs on the data your team can actually reach — not the data sitting in disconnected systems or buried in unabstracted lease PDFs. Bad inputs produce confident-sounding wrong answers.
The quality report comes before the recommendation. Always.
If your data cannot support 95% confidence on a scenario, Surfaice says so — and names the sources to connect or the leases to re-abstract to fix it.
Assembling it by hand
~4 wks
To assemble portfolio data manually
- Data pulled from four to six disconnected systems
- Lease clause gaps missed until it is too late
- Confidence levels never stated
- Recommendations with no interval around them
With connected sources
Same day
From connected sources to ranked scenarios
- Data quality scored before any modelling
- All six scenarios generated and ranked
- A 95% interval on every headline metric
- Proxy values require explicit approval
How a run works
The same sequence every time.
No result is surfaced before data quality is disclosed, and no capital decision is made without human confirmation.
- 01
Context and connected sources
The capital budget, the planning horizon, the primary and secondary objectives, and the portfolio in scope. Data then resolves through your connected systems rather than a fresh export.
Lease admin · ERP · facilities · development · site selection
- 02
The data quality report, before any result
Each domain is scored, a weighted overall score computed, critical gaps flagged — missing co-tenancy data, absent P&L, stale comparables — and the confidence gate evaluated. No scenario output appears before this.
Mandatory, no exception
- 03
Six scenarios generated and modelled
All supported scenarios run from your own portfolio data, with intervals on every primary metric. If fewer than five are runnable, you are told exactly which data unlocks the rest.
Minimum five runnable, or it says why not
- 04
Ranked against your objectives
Each scenario is scored against your primary objective at 60% weight and your secondary at 40%, producing one recommendation, two or three alternatives, and contingency guidance if budget or market conditions shift.
Earnings · risk · growth · efficiency · relationships
- 05
A dashboard that refreshes
Portfolio overview, scenario comparison, capital timeline, co-tenancy exposure, renewal runway and data lineage. A weekly refresh runs by default, and a material change to lease data, store financials or the development pipeline triggers a full re-run.
Weekly, and on material change
The confidence gate
Any scenario whose 95% interval is wider than 30% of the mean needs human confirmation before it surfaces as a recommendation.
| Score | Gate | What happens |
|---|---|---|
| < 60% | Halt | Quality report and a remediation plan. No modelling runs. |
| 60–74% | Warn | Quality report, then your confirmation before modelling. |
| 75–84% | Proceed with disclosure | Quality report, and any scenario below 95% confidence is named. |
| ≥ 85% | Proceed | Quality report, and 95%+ confidence is achievable throughout. |
Six scenarios, every run
Each one answers a question you already have.
A minimum of five runnable scenarios per engagement. Readiness and confidence are disclosed before modelling, not assumed after it.
Strategic closure optimisation
Ready with disclosure · sample confidence 91%
Which stores should close, in what sequence, to maximise earnings improvement while minimising exit cost and co-tenancy cascade?
Remodel ROI optimisation
Ready · sample confidence 94%
Which stores should be remodelled, in what order, to maximise sales lift per dollar deployed?
Relocate, renew or exit
Ready · sample confidence 90%
For leases expiring in the horizon, should each store renew in place, relocate within market, or exit entirely?
Expansion optimisation
Ready with disclosure · sample confidence 88%
Where should new stores open to maximise net portfolio earnings growth after cannibalisation?
Hybrid capital optimiser
Ready · sample confidence 92%
What mix of closures, remodels, relocations, renewals and openings is optimal under budget, execution and cash-flow constraints?
Renewal negotiation leverage
Ready · sample confidence 89%
For upcoming renewals, where is there leverage, and does an aggressive, moderate or defensive strategy maximise value?
What it needs, and how it weights it
Four required domains. Two that add confidence.
Every score feeds the confidence gate and determines which scenarios are runnable at all.
Lease administration
Complete abstracts for every active store, with every clause: renewal, termination, kickout, co-tenancy, assignment, improvement allowances, dark-store provisions and percentage rent.
Store financials
Revenue, gross margin, earnings, occupancy cost ratio, sales per square foot and year-on-year trend — three or more years of history to support the remodel regression.
Facilities
Deferred maintenance, equipment age and replacement schedules, operating cost per square foot, energy cost and preventive maintenance compliance by store.
Store development
Capital models, project pipeline, committed capital, permitting timelines, vendor capacity, and historical actuals against budget.
Site selection
Trade areas, candidate sites, market rent comparables, competitor locations and cannibalisation risk. Enhancing rather than required.
Landlord relationships
Negotiation history, concessions granted, landlord portfolio intelligence and known pressures. Enhancing rather than required.
Sample output
fictional portfolioWhat a finished analysis reports.
A 48-store portfolio with a $10M budget, run end to end: the quality report, all six scenarios, the hybrid comparison, a twelve-month roadmap and the executive risk controls.
$5.2–6.6M
12-month earnings improvement
At 95% confidence
18 mo
Expected payback
On $10M deployed
91%
Data quality score
Weighted across four domains
6/6
Scenarios runnable
In the sample engagement
Recommended action mix — hybrid allocation
| Action | Count | Capital | 12-mo impact | Risk |
|---|---|---|---|---|
| Close or exit underperforming stores | 5 | $2.1M | +$2.4M | Medium |
| Remodel high-ROI stores | 8 | $4.6M | +$1.7M | Low |
| Relocate constrained store | 1 | $1.8M | +$0.6M | Medium |
| Open high-confidence new site | 1 | $1.2M | +$0.4M | Medium |
| Renewal negotiations | 9 leases | $0.3M | +$0.8M | Low |
| Total | — | $10.0M | +$5.9M | Medium |
Fictional data, for demonstration only — not a basis for business decisions. A real run uses your connected data and shows the quality report before any result.
Five rules the run cannot talk itself out of.
These are not aspirations. They are constraints enforced on every run, and each one can stop it.
Real values only
Every modelled input is a retrieved value, a mapped canonical value, a value you supplied, or an explicitly approved proxy with a cited source. Nothing is invented.
No proxy without approval
Where an input is missing, the source, the value, why it applies and its effect on confidence are shown — and the scenario halts if you do not approve it.
No headline number without an interval
Every primary financial metric carries a 95% interval or is labelled as not modelable with acceptable confidence. A wide interval requires confirmation before it becomes a recommendation.
Humans decide anything consequential
Lease terminations, capital commitments, landlord communications and CRM updates require approval. This is decision support; it does not act.
Critical gaps halt the run
Co-tenancy data missing for more than 40% of the portfolio, store P&L missing for more than 25% of stores, or any missing lease expiration date: the run stops and remediation steps are provided.
Turn portfolio data into a capital allocation strategy.
Bring your budget and your connected systems. The first thing you will see is what your data can and cannot support.