00/ 08
HeroWhere next
Location intelligence · India · v1.0

The decision engine for retail expansion.

For brands — national and international — opening physical stores in India. Every micro-market in your footprint, scored against a model tuned to your brand. Updated as the city moves.

3
Cities live
72h
To a blind ranking
Live.
Re-scored weekly, not once
Specimen
Marsh · blind analysis Returned in 72 hrs
Three stores, ranked blind.
What you receive. You send three of your existing stores as coordinates — we rank them blind. No performance data shown; this is the format, not a result.
01
Store B · your existing store
Predicted strongest — footfall-to-rent ratio leads the set.
84
02
Store A · your existing store
Solid catchment, heavier competition nearby.
71
03
Store C · your existing store
Predicted weakest — catchment overlaps Store A.
58
You know which ranking is right — you've seen the numbers. If ours matches, the model just proved itself on your fleet.
01/ 08
The blindtest
For heads of expansion & land acquisition

Send us three of your stores. We’ll rank them blind.

One discovery call. Coordinates only. We return a blind ranking — best, worst, and what the score is sensitive to — without ever seeing your performance data.

The offer

Three stores. One ranking. Seventy-two hours.

Send three stores
01 · You send

Three coordinates.

Three live store locations. That’s the entire input — no profit-and-loss statements, no tenant mix, no secrets.

02 · We run

Your category’s model.

A blind ranking against the model built for your category — with no knowledge of how the stores actually perform.

03 · We show

Best, worst, why.

Which store should be strongest, which weakest, the reasoning behind each score — and an honest read of where we’re uncertain.

Why coordinates only You already know how these stores perform — we don’t. If our blind ranking matches the reality you can see in your own numbers, the model has proven itself on the one dataset you trust completely, before you share a single figure or sign anything. If it doesn’t, you’ve spent nothing but seventy-two hours.
02/ 08
Problem3 failure modes

Most stores open in the wrong place. Then they close quietly.

Site selection in Indian retail still runs on three tools, and each one breaks the same way: the decision outruns the information.

A · The familiar map

The city you know runs out.

Expansion naturally follows the neighbourhoods a team knows best — a sensible instinct, and a small map. The next hundred good sites are mostly in places nobody on the team passes on their commute.

B · Consultant studies

Paid once. Stale on arrival.

A site study takes months and lands as a PDF. By the time the lease is being negotiated, footfall, rents, and competitors have moved — and the study can’t move with them.

C · Spreadsheet sprawl

All the data. None of the model.

Footfall in one tab, rents in another, competitor screenshots in a folder. Nothing weighs anything against anything else — so the decision still gets made in the meeting, by voice.

03/ 08
Method3 stages
How it works

Your brand, modeled. The city, scored. The shortlist, delivered.

Input01

Your brand, modeled.

Category, target customer, and how your best stores behave. We learn what “a great site for you” means — we don’t guess it.

Process02

The city, scored.

Every micro-market in scope, ranked against your model and re-scored as the city changes. The how stays our trade — the why behind every score is always shown.

Output03

The shortlist, delivered.

A shortlist worth visiting. Field-ready briefs. Reasoning your real-estate team can defend in a board meeting.

04/ 08
DemoMap + score

Delhi, scored for a specialty coffee brand.

A preview of the console with illustrative scores — what your team sees once your model is tuned. Pick a category and the whole city re-ranks.

Category

Specialty coffee87
QSR · burgers72
Premium gym64
D2C beauty59
Bakery55
City
Delhi · NCR
Micro-markets, scored weekly
Score Footfall Competition
Illustrative scores · not client data
Selected · top of stack
Khan Market
Middle Lane
micro-market · NDMC
87
out of 100 · top of city
Footfall92
Demographics78
Competition65
Catchment81
Rent vs. revenue70
ConfidenceHigh
Re-scoredWeekly
Audit logview →
05/ 08
Outcomes3 commitments

What changes when the city is scored.

No borrowed case studies, no invented lift numbers — these are the commitments the engagement is built on.

Months weeks
Shortlist turnaround
From brief to a field-ready shortlist
72 hrs
Blind ranking
Coordinates in, verdict out — before any contract
100%
Auditable scores
Every number carries its why — quality of signal over quantity of signal
06/ 08
Engagements3 shapes

Three ways to work with us.

Scoped per engagement, priced in conversation — not on a rate card.

01 · Single market
Scout
One city · one category · 90 days
  • One metro, one retail category
  • Candidate sites scored & ranked
  • Signals refreshed as the city moves
  • Field-ready briefs & audit log
Start a market
02 · NetworkMost chains
Compass
All live cities · twelve months
  • Every Marsh-live city
  • Every candidate site in scope
  • Continuous re-scoring as cities move
  • Model weights tuned to your brand
  • Quarterly review with the founders
Talk to founders
03 · Enterprise
Custom
Embedded · API · analyst on call
  • Signals extended with your data
  • API & warehouse integration
  • Dedicated analyst
  • Data-room governance & audit
Scope an engagement
Offerings listed, prices deliberately not — every engagement is scoped to the brand. Ask and we’ll quote within the week.
07/ 08
FoundersDaedalus
The people behind the model

Built by Daedalus.

Daedalus is the parent company — the name on the card in your hand. Marsh is its first instrument: a working answer to where retail brands should open next in India.

We build quantitative infrastructure for decisions that Indian retail still makes on instinct and proximity. We’d rather be tested than trusted — which is why the blind analysis comes first, and the contract comes after.

Sharvi Jain
Sharvi Jain
Cofounder

Urban designer, Columbia University. Reads every blind analysis before it goes out.

Arnav Jain
Arnav Jain
Cofounder

Former product manager, Yahoo. Computer science, Purdue. Builds the model behind every score, and the why behind each.

08/ 08
FAQPlain answers

The honest questions.

Q.01We already have brokers and a real-estate team. Where does Marsh fit?+
Before them. Marsh decides where your team should be looking and negotiating; your brokers and acquisition team still do what they do best. The shortlist arrives ranked, scored, and explained — so the field visits start from the right ten places instead of the nearest ten.
Q.02Which cities are live today?+
Three at v1.0. New cities are commissioned as engagements need them — if yours isn’t live yet, ask. Bringing a city online is part of a Compass or Custom engagement, not a separate project.
Q.03How do you handle our internal data?+
Anything you share lives in a tenant isolated to you, and nothing identifying your brand ever appears in another client’s scores or leaves without your written consent. As we build toward a product, anonymised, brand-stripped signals may help improve the underlying models — never your raw data, never your performance, never anything that could be traced back to you. We’re happy to walk your security team through exactly where that line sits.
Q.04What does the blind analysis return, exactly?+
A ranking of your three submitted stores with scores, the signal-level reasoning behind each, a confidence band, and an honest note on where the model is uncertain. Seventy-two hours, no obligation.
Q.05Who else sees our results?+
Nobody outside your tenant. We don’t cross-reference brands, we don’t publish benchmarks that could re-identify you, and we don’t name clients without consent.

Where should you open next?

Talk to founders Request a blind analysis info@atdaedalus.com Reply within one business day · India
Marsh by Daedalus
Location intelligence platform for retailers. v1.0 · 2026.
Blind test Method Demo
Engagements Founders FAQ
© 2026 Daedalus
India
Privacy Data handling
Marsh is building toward a location-intelligence product. Data shared with us is used to deliver your engagement; where it improves our models it is used only in anonymised, brand-stripped form. We never name clients, reveal their internal data, or expose anything that could identify a brand without written consent.