ROTAS: Return on Total Ad Spend

A Non-Attribution Marketing Performance Model

Author: David Chadderton, CMO, Homes for Students Version: 1.0 — June 2026 Status: Draft for Implementation

1. The Problem ROTAS Solves

Attribution modelling is broken. Not slightly broken, fundamentally broken.

Multi-touch attribution assumes a clean, trackable customer journey. In PBSA, that journey includes a parent’s influence, a friend’s recommendation, a TikTok they saw in October, a Google search in January, a campus visit in February, and an AI assistant’s recommendation in March. No attribution model captures that. Not last-click, not data-driven, not any of them.

Meanwhile, marketing teams spend increasing time and budget trying to attribute credit to individual channels, while the actual question the board asks is simpler and more important: “What did we get for what we spent?”

ROTAS answers that question. Not by assigning credit, but by measuring the whole system.

2. The ROTAS Formula

2.1 The Core Equation

ROTAS = Marketing Composite Score (MCS) ÷ Total Marketing Spend (TMS)

Where:

MCS = a weighted composite of five marketing outputs
TMS = total marketing spend, excluding agent commissions

ROTAS is expressed as a points-per-pound (or points-per-£1,000 for readability) metric. Higher is better.

2.2 Two Versions

Version	Formula	Use Case
ROTAS Score	MCS (absolute)	Board reporting, trend analysis, city/brand comparison
ROTAS Ratio	MCS ÷ TMS	Efficiency comparison, budget allocation, YoY performance

Both are reported together. The Score tells you what marketing produced. The Ratio tells you how efficiently it produced it.

3. The Marketing Composite Score (MCS)

3.1 Components

The MCS is built from five weighted components, each measuring a distinct dimension of marketing output:

#	Component	Weight	What It Measures	Why It Matters
1	Leases Signed	35%	Total new leases executed in the booking cycle	The ultimate conversion metric. Directly attributable to marketing + sales pipeline
2	Revenue Generated	25%	Total revenue from leases signed (net of discounts/concessions)	Quality of conversion. A lease at £150/week is worth more than one at £120/week
3	Enquiry Volume	20%	Total qualified enquiries (calls, forms, live chats, walk-ins, emails)	Pipeline health. Leading indicator of future lease performance
4	Brand Awareness	10%	Aided and unaided awareness scores (survey-based)	Top-of-funnel strength. Measures whether marketing is building future demand
5	Student NPS	10%	Net Promoter Score from current residents	Retention and reputation. High NPS reduces future acquisition cost

3.2 Why These Five

Each component answers a different question:

Leases Signed: Did marketing fill the beds?
Revenue Generated: Did it fill them at the right value?
Enquiry Volume: Is the pipeline healthy for next cycle?
Brand Awareness: Are we building long-term demand, not just harvesting short-term intent?
Student NPS: Are current residents advocates or detractors?

Together, they give a complete picture of marketing’s contribution: short-term performance, medium-term pipeline, and long-term brand equity.

3.3 What’s Deliberately Excluded

Agent commission leases: These are a sales channel, not a marketing output. Agent-generated leases are tracked separately and do not contribute to MCS.
Channel-level metrics (CPC, CTR, impressions): These are operational diagnostics, not strategic outputs. They inform channel optimisation but do not belong in the board-level model.
Website traffic: Volume without quality. A million visits that don’t convert are noise.
Social media followers/engagement: Vanity metrics unless they correlate with the five components above.

4. Normalisation & Scoring

4.1 The Problem of Different Units

Leases are counted in units. Revenue is counted in pounds. Enquiries are counted in units. NPS is on a -100 to +100 scale. Brand awareness is a percentage. You cannot simply add them together.

Each component must be normalised to a common scale before weighting.

4.2 Normalisation Method: Index to Target

Each component is scored against a target set at the start of the booking cycle (October). The target is based on:

Prior year actuals (adjusted for portfolio changes)
Growth ambition (board-approved)
Market conditions

Formula:

Component Score = (Actual ÷ Target) × 100

Example:

Component	Target	Actual	Score
Leases Signed	8,000	7,600	95.0
Revenue	£65M	£68M	104.6
Enquiries	45,000	47,200	104.9
Brand Awareness	65%	62%	95.4
Student NPS	+55	+52	94.5

4.3 Calculating MCS

Apply the weights:

MCS = (95.0 × 0.35) + (104.6 × 0.25) + (104.9 × 0.20) + (95.4 × 0.10) + (94.5 × 0.10)
MCS = 33.25 + 26.15 + 20.98 + 9.54 + 9.45
MCS = 99.37

An MCS of 100.00 means marketing hit exactly its target across all five components. Above 100 is overperformance. Below 100 is underperformance.

4.4 Floor and Cap

To prevent a single extreme component from distorting the composite:

Each component score is floored at 0 and capped at 150.
This means a single component blowing its target out of the water (e.g. revenue at 200% of target) contributes proportionally, but doesn’t mask poor performance elsewhere.

5. Total Marketing Spend (TMS)

5.1 What’s Included

Everything the marketing function spends, except agent commissions:

Category	Examples
Paid Media	PPC, paid social, display, OOH, print, radio, programmatic
Organic & Content	SEO, content production, photography, video, design
Technology	CRM, marketing automation, analytics tools, website hosting
Agencies	Media agencies, creative agencies, PR agencies
People	Marketing team salaries, benefits, contractor costs
Events	Open days, freshers’ fairs, campus activations
Research	Brand tracking surveys, NPS surveys, market research

5.2 What’s Excluded

Agent commissions: These are a sales/distribution cost, not a marketing cost. Tracked and reported separately.
Capital expenditure: One-off website rebuilds, office fit-outs, etc. If it’s capex, it’s not in TMS.

5.3 TMS Reporting

TMS is reported in two ways:

Absolute: £X.XXM total
Per bed: £X per available bed (for portfolio comparison)
Per lease: £X per lease signed (for efficiency trending)

6. ROTAS in Practice

6.1 The Reporting Hierarchy

ROTAS operates at five levels:

┌─────────────────────────────────────┐
│         GROUP ROTAS                 │
│    (HFS + VervLife + Orla)          │
├──────────┬──────────┬───────────────┤
│ HFS ROTAS│VervLife  │  Orla ROTAS   │
│          │ ROTAS    │               │
├──────────┴──────────┴───────────────┤
│         CITY ROTAS                  │
│   (Per city, per brand)             │
├─────────────────────────────────────┤
│       BUILDING ROTAS (PRIMARY)      │
│   (Per property, per brand)         │
├─────────────────────────────────────┤
│     CHANNEL DIAGNOSTICS             │
│  (Attribution-level, operational)   │
└─────────────────────────────────────┘

Group ROTAS: Board-level, overall health of the marketing function
Brand ROTAS: Brand-level, comparing HFS, VervLife, Orla performance
City ROTAS: City-level roll-up, identifying market-level trends
Building ROTAS: The primary unit of analysis. Each building has its own MCS, TMS, and ROTAS score. This is where strategy meets reality.
Channel Diagnostics: Operational, for the marketing team to optimise specific channels

Building-first, not city-first. ROTAS is fundamentally a building-by-building measurement. A city-level ROTAS is a weighted roll-up of its buildings. A brand-level ROTAS is a weighted roll-up of its cities. The building is where the beds are, where the spend goes, and where the return comes from. Everything above building level is aggregation.

What building-level means for TMS:

Direct building spend (local PPC, local events, building-specific content) is allocated to that building
Shared spend (brand campaigns, central tech, headcount) is allocated proportionally, typically by available beds or enquiry volume
Agent commissions are excluded at building level, same as everywhere else

What building-level means for MCS:

Leases Signed: Per building. Clean.
Revenue Generated: Per building. Clean.
Enquiry Volume: Per building where possible. Enquiries routed to a building count to that building. General brand enquiries allocated proportionally.
Brand Awareness: Surveyed at city or brand level, not building. Allocated equally to buildings within the survey area.
Student NPS: Per building. Clean (surveyed per property).

6.2 Reporting Cadence

Frequency	What	Audience
Annual	Full ROTAS with all five components, targets set, full MCS calculated	Board
Quarterly	ROTAS update with available components (NPS and brand awareness lag)	Board, leadership
Monthly	ROTAS proxy using leases, revenue, enquiries only	Marketing team
Weekly	Channel diagnostics (attribution-level, operational)	Marketing team, agencies

6.3 The ROTAS Dashboard (Conceptual)

A single-page view for the board:

╔═══════════════════════════════════════════════════════════╗
║  ROTAS: ORLA EUROPE — BOOKING CYCLE 2025/26              ║
╠═══════════════════════════════════════════════════════════╣
║                                                           ║
║  MCS: 99.4    TMS: €2.1M    ROTAS: 47.3 pts/€1k        ║
║  Target: 100  Target: €2.0M  YoY: +8.2%                  ║
║                                                           ║
║  ┌─────────────┬────────┬────────┬──────┐                 ║
║  │ Component   │ Weight │ Target │ Act. │                 ║
║  ├─────────────┼────────┼────────┼──────┤                 ║
║  │ Leases      │  35%   │ 3,200  │3,040 │                 ║
║  │ Revenue     │  25%   │ €28M   │ €29M │                 ║
║  │ Enquiries   │  20%   │ 18,000 │18,900│                 ║
║  │ Awareness   │  10%   │ 55%    │ 52%  │                 ║
║  │ NPS         │  10%   │ +50    │ +48  │                 ║
║  └─────────────┴────────┴────────┴──────┘                 ║
║                                                           ║
║  TOP 5 BUILDINGS            BOTTOM 5 BUILDINGS            ║
║  1. Lisbon Central: 118.2   1. Warsaw West: 72.1          ║
║  2. Barcelona Gràcia: 112.7 2. Prague North: 76.8         ║
║  3. Milan Bovisa: 109.4     3. Berlin Mitte: 81.3         ║
║  4. Paris 11e: 106.1        4. Madrid Sur: 84.6           ║
║  5. Dublin Docklands: 104.3 5. Lyon Part-Dieu: 87.2       ║
║                                                           ║
╚═══════════════════════════════════════════════════════════╝

7. ROTAS vs Existing Metrics

7.1 What ROTAS Replaces

Old Metric	Why It’s Replaced
Channel-level ROAS as the primary metric	Attribution is unreliable in PBSA. ROTAS measures the whole system.
CPA (Cost Per Acquisition) as a board metric	CPA only measures cost efficiency, not output quality. A low CPA with low revenue is worse than a moderate CPA with high revenue.
“How did we do?” conversations based on gut feel	ROTAS provides a structured, comparable, targetable framework.

7.2 What ROTAS Doesn’t Replace

Metric	Why It Stays
Channel ROAS/CPA	Still useful for operational optimisation within the marketing team. If PPC is underperforming, you need to know.
Conversion rate by stage	Funnel diagnostics are operational, not strategic.
Individual campaign performance	Campaign-level measurement is valid for tactical decisions.

The principle: ROTAS is the board-level strategic lens. Attribution metrics are the team-level operational lens. Both exist. They just serve different audiences.

8. Implementation Plan

Pilot: Orla (Europe) First

ROTAS launches at Orla Europe as a controlled pilot before rolling to HFS and VervLife. This allows model validation in a smaller, cross-border context before scaling to the 60,000+ bed UK portfolio.

Why Orla Europe first:

Smaller portfolio, faster feedback loops
Tests the model across multiple European markets and currencies
Different booking cycles and market dynamics stress-test the framework
Lessons learned reduce risk when HFS (the largest brand) adopts
European PBSA is less mature, so a new model faces less institutional resistance

Orla Europe-specific considerations:

Currency: TMS and revenue will be in EUR (and potentially other local currencies). ROTAS Ratio is currency-agnostic (points-per-unit), but MCS targets must be set in local currency or normalised to EUR
Booking cycles: European university calendars vary by country. The Oct-Sep cycle may need adjustment per market
Data infrastructure: Confirm NPS and brand awareness tracking exist in each Orla market. If not, Phase 1 includes building that capability
Agent model: Confirm agent commission structures across European markets to ensure clean TMS exclusion

Phase 1: Define & Baseline — Orla Europe (Month 1-2)

Action	Owner	Deliverable
Agree final component weights with leadership	Dave	Signed-off weight matrix
Audit Orla Europe data availability (NPS, brand awareness, enquiry tracking)	Orla Marketing	Data gap analysis
Set per-building targets for current booking cycle	Orla Marketing + Finance	Target document (per building, per market)
Pull prior year data for all five components	Orla Marketing	Baseline dataset
Calculate prior year ROTAS as historical benchmark	Orla Marketing	Baseline ROTAS score
Confirm agent commission exclusion per market	Orla Finance	Clean TMS definition

Phase 2: Build & Integrate — Orla Europe (Month 2-4)

Action	Owner	Deliverable
Build ROTAS calculation in reporting stack	Marketing Ops	Working dashboard
Connect data sources (CRM, ad platforms, NPS, brand tracker)	Marketing Ops	Automated data feeds
Define building-level, city-level, and market-level reporting views	Marketing Ops	Multi-level dashboard
Train Orla marketing team on ROTAS methodology	Dave	Team briefing document

Phase 3: Operationalise — Orla Europe (Month 4-6)

Action	Owner	Deliverable
First quarterly ROTAS report for Orla Europe	Dave	Board/leadership presentation
Integrate ROTAS into Orla budget planning process	Dave + Finance	Budget model uses ROTAS
Use building-level ROTAS to identify intervention targets	Orla Marketing	Building action plans
Document lessons learned from pilot	Dave	Pilot retrospective

Phase 4: Scale to HFS & VervLife (Month 6+)

Action	Owner	Deliverable
Adapt ROTAS model for HFS UK portfolio	HFS Marketing	HFS-specific targets and weights
Adapt ROTAS model for VervLife (BTR)	VervLife Marketing	VervLife-specific targets and weights
Group ROTAS dashboard (HFS + VervLife + Orla)	Marketing Ops	Group-level view
ROTAS becomes standard board reporting metric	Dave	Standing agenda item
Annual weight and target review	Dave + Leadership	Updated model
ROTAS integrated into team KPIs and performance reviews	HR + Dave	KPI framework
Explore predictive ROTAS (forecasting MCS from pipeline data)	Marketing Ops	Predictive model

Note on VervLife: The MCS weights may need adjustment for BTR. Leases and revenue will still dominate, but the brand awareness and NPS components may carry different significance in a young-professional renter demographic vs student PBSA. The Orla pilot will reveal whether the weights need per-brand customisation.

9. Strategic Implications

9.1 Budget Allocation

ROTAS enables a fundamentally different approach to budget allocation. Instead of asking “which channel gets credit for this lease?”, you ask “which building has the highest ROTAS opportunity?”

High ROTAS, high spend: Maintain. These are your efficient engines.
High ROTAS, low spend: Invest. These are your underfunded buildings.
Low ROTAS, high spend: Investigate. Something is wrong with the building’s marketing system, not just one channel.
Low ROTAS, low spend: Evaluate. Is this building worth the investment, or should spend be reallocated?

Aggregated to city and brand level, the same matrix applies for portfolio-level decisions.

9.2 AI-Mediated Discovery

As AI assistants increasingly form the student’s consideration set (the “Invisible First Click” framework), attribution becomes even less reliable. A student may never click an ad, never visit a tracked link, never fill in a form. They ask ChatGPT “best student accommodation in Bristol” and it recommends HFS. That’s marketing output. No attribution model captures it.

ROTAS doesn’t need to capture it. It measures the system’s total output regardless of how the student arrived. This makes it future-proof in a way attribution models are not.

9.3 The Agent Question

By excluding agent commissions, ROTAS isolates marketing’s direct contribution. This creates a clean comparison: what does marketing deliver on its own vs what the agent channel delivers? If ROTAS Ratio is healthy, the case for reducing agent dependency strengthens. If it’s weak, the case for agents strengthens. Either way, the data supports the decision.

9.4 The PBSA Industry

No PBSA operator has a standardised, non-attribution marketing performance model. If HFS builds ROTAS and it works, it becomes a publishable framework. Dave’s positioning as a thought leader in marketing performance, combined with the STAR framework’s behavioural science foundation, makes this a credible industry contribution, not just an internal KPI.

Metric	What It Measures	Attribution Required?	ROTAS Difference
ROAS	Revenue per £1 of ad spend per channel	Yes	ROTAS measures total system, no channel attribution
MER	Total revenue ÷ total marketing spend	No	MER is revenue-only. ROTAS includes pipeline, brand, and satisfaction
CAC	Cost to acquire one customer	Yes (per customer)	ROTAS doesn’t track individual acquisition cost
LTV:CAC	Lifetime value vs acquisition cost	Yes	ROTAS is cycle-based, not lifetime-based
iROAS	Incremental ROAS from experiments	Yes (experimental)	ROTAS is the whole-system view, not incremental

ROTAS is closest to MER but enriched with pipeline, brand, and satisfaction data. It’s MER with a brain.

11. Version History

Version	Date	Changes
1.0	June 2026	Initial model design

David Chadderton is the creator of the STAR Framework and the CMO at Homes for Students, VervLife, and Orla, covering 60,000+ beds across 56 UK cities.