Why These 10 Sydney Restaurants Get Location Right
Locatalyze Research Team
Location intelligence, Locatalyze
Sydney counts the people on its own footpaths, twice a year, block by block, from six in the morning until midnight — and has done since 2013. Almost nobody in hospitality reads it. We pulled the full thirteen-year record and ran it against ten restaurants. What it shows is not how busy Sydney is; it is who each street belongs to. Two footpaths 946 metres apart, inside the same council area, differ sevenfold on that measure — and a restaurant that would thrive on one would struggle on the other. This is a location-intelligence read of ten Sydney sites, not a food ranking, and not a claim about anyone's private accounts.
How to read this article
**FACT** — supported by a named, linked source, or computed by us from a named open dataset with the method stated. **LOCATALYZE ANALYSIS** — our interpretation of that evidence. **INFERENCE** — a reasonable reading that is not proven. Every pedestrian figure below is our own calculation from the City of Sydney's published walking counts; the method, and its limits, are set out under [Method, evidence and limits](#method).
What this is not
This is not a ranking, and it is not a claim that any restaurant named here is profitable. Trading figures, rent deals and margins are private, and we do not estimate them. What is observable is the site: its catchment, access, anchors, competitors, published trading hours, and — inside the City of Sydney — the counted number of people who walk past it. That is what we analyse. See our [methodology](/methodology) for how Locatalyze scores locations.
Key takeaways
Sydney counts its own footpaths, and almost nobody reads it
Every hospitality lease in this country is signed on an assumption about how many people walk past the door, and when. In most of Australia that assumption stays an assumption. In the City of Sydney it does not have to. Twice a year — in March and October — the council sends counters to roughly 100 locations across its local government area and counts pedestrians from 6am to midnight, on one weekday and one weekend day, in fair weather. It has done this since 2013, and it publishes every count as open data.
We pulled the full history — 5,453 records covering 130 sites — and calculated four things for each block: the average weekday count, the average weekend-day count, the ratio between them, and the ratio of the evening window (6pm to 11pm) to the lunch window (11am to 3pm). Current figures are the mean of the three most recent surveys, autumn 2025 to autumn 2026. Where we quote change, the baseline is 2019, the last year before the pandemic, the light rail and the Metro all landed on this city at once.
The weekend-to-weekday ratio is the number this article is built on, and as far as we can establish nobody publishes it. It is not a measure of how busy a street is. It is a measure of who the street belongs to.
FACT: Martin Place counted an average of 71,146 pedestrians on a weekday and 22,266 on a weekend day. FACT: Argyle Street in The Rocks counted 13,948 and 32,086. LOCATALYZE ANALYSIS: these two footpaths are 946 metres apart, inside the same local government area, and they are not in the same business. One is a five-day office corridor that empties on Saturday. The other is a two-day visitor precinct that is busiest when the first one is dead. A restaurant that would thrive on one would struggle on the other, and no suburb-level number can tell them apart, because for most purposes they are the same suburb.
The hourly curves make the point harder than the totals do. FACT: 43% of King Street Newtown's weekday pedestrians pass between 6 and 11pm, and 37% of Argyle Street's. On Martin Place the figure is 16%, and in Pitt Street Mall 17%. LOCATALYZE ANALYSIS: for a business whose revenue is concentrated in dinner, that single percentage is closer to a forecast than the daily total will ever be. Martin Place peaks at 8am — 10,142 people in a single hour, the largest hourly count anywhere in the dataset — then again at 1pm, then drains. By 8pm it is carrying fewer people than King Street in Newtown. INFERENCE: an 8am peak measures arrival at work, not appetite. It is the one hour of the day in which almost nobody is deciding where to have dinner.
Where the data stops
These are the City of Sydney's counts, so they cover the CBD, The Rocks, Haymarket, Pyrmont, Glebe, Redfern, Surry Hills, Darlinghurst, Potts Point, Paddington's western end, Alexandria and the Newtown end of King Street — and nothing else. There is no count for Bondi, Manly, Parramatta, Cabramatta, Chatswood or Balmain. For sites outside the boundary we say so, and reason from other evidence, rather than quietly extending a number that does not exist.
What the counts change about choosing a Sydney site
Read who, then when, then how many
Volume is the last question, not the first. Pitt Street Mall counts 50,139 people on a weekday — roughly two and a half times the busiest counted block of King Street in Newtown — and returns an evening-to-lunch ratio of 0.42, the lowest of any major block we measured. Newtown returns 1.93. FACT: both figures are ours, from the same survey. LOCATALYZE ANALYSIS: for a dinner-led restaurant, the Newtown block delivers roughly four and a half times more evening trade per person walking past, on frontage that costs a fraction of the Mall's. The Mall is not a worse street. It is a different street, and it is priced for the business it actually does.
The unit is the block
Five blocks of King Street, Newtown are counted separately. They range from 4,646 people a weekday between Angel and Union streets to 20,480 between Australia and Eliza — a 4.4-fold spread. FACT: all five are on King Street; all five are in Newtown. LOCATALYZE ANALYSIS: "a site on King Street" describes an address, not a catchment. If one street name contains a 4.4-fold range, a suburb name contains more, and a suburb score cannot resolve a lease.
Sydney's geography makes access part of the concept
This is a city cut into pieces by water and cliffs, and the pieces are joined by scheduled services rather than by roads you would want to drive. That has a specific consequence for hospitality: on much of the peninsula, your catchment does not live within walking distance — it arrives, and then it leaves, on a timetable. FACT: the Metro's city section opened on 19 August 2024 with new stations at Barangaroo, Martin Place, Gadigal and Waterloo, and about 120,000 of the network's 210,000 daily metro trips now run under the harbour (Sydney Metro). INFERENCE: on a site like that, the question is not how many parking bays are out the front. It is what time the last service runs, and whether your kitchen is still open when it does.
Rent is a claim on covers you have not sold yet
Counted footfall does not pay rent; converted covers do. The arithmetic that decides a Sydney site is how many covers per week it must produce, at your real average spend, before rent and wages clear — and then whether the street's shape can deliver them in the hours you actually trade. Our guide to why a cheaper lease is often the more expensive location works through how that goes wrong, and high foot traffic, low sales covers the failure mode this whole article is about.
Ten Sydney sites, and what each one is actually renting
These are not the ten best restaurants in Sydney and they are not ranked. They were chosen to cover ten distinct location strategies, to be currently trading and publicly documented, and — wherever possible — to sit on or beside a counted block, so the reasoning can be checked rather than asserted. Six of the ten are counterintuitive: they are on the wrong street, in the wrong suburb, or in a precinct our own model marks down.
1. Rockpool Bar & Grill — 66 Hunter Street, Sydney CBD
The location. A basement dining room in a heritage office building, on a block of Hunter Street that counts 27,676 people on an average weekday and 8,268 on an average weekend day. FACT: that is a weekend-to-weekday ratio of 0.30, the lowest in this article alongside Martin Place. The block sits between Bligh and O'Connell streets, in the middle of the country's largest concentration of premium office towers.
The demand engine. Office workers, and the expense accounts attached to them. FACT: Sydney CBD office vacancy fell to 13.3% by mid-2026, below the 14.9% national CBD rate and far below Melbourne's 19.0%, with Sydney premium-grade vacancy at just 8.9% (Property Council of Australia). LOCATALYZE ANALYSIS: the premium end of the market, which is the part of it that eats here, is close to full.
The trade-off, and why it works. FACT: Rockpool trades lunch and dinner Monday to Friday, dinner only on Saturday, and closes on Sunday. LOCATALYZE ANALYSIS: that published roster is an almost exact fit to a 0.30 street. The business does not attempt to sell a Sunday that the footpath cannot supply, and it puts its highest-value service — corporate lunch — into the hours when the block is at its fullest. What it gives up is the entire weekend leisure market, and with it any hedge against a downturn in white-collar employment.
The lesson. On a five-day street, a seven-day business is not more ambitious, it is more expensive. Match the roster to the ratio.
2. Saké — 12 Argyle Street, The Rocks
The location. A heritage warehouse on the counted Argyle Street block between Harrington and George. FACT: 13,948 people on an average weekday, 32,086 on an average weekend day — a ratio of 2.30, the highest of any major block in the dataset. It is the mirror image of Hunter Street.
The demand engine. Visitors. FACT: Sydney received 3.9 million international visitors and 63.6 million domestic visitors in the year to March 2026, and international visitors alone spent $13.9 billion in the city (Destination NSW). The Rocks is the oldest and most heavily signposted part of that itinerary.
The trade-off, and why it works. FACT: Saké trades lunch and dinner seven days, with its longest hours on Friday and Saturday. LOCATALYZE ANALYSIS: on a 2.30 street the seven-day roster is not optional — Saturday and Sunday are the week's two best days, and a venue closed on them would be shut for its own peak. The trade-off is that this catchment is borrowed rather than owned: it responds to airfares, cruise schedules and the exchange rate, none of which the operator controls. INFERENCE: that is the risk premium built into a tourism site, and it is the reason a visitor precinct rewards a format with high covers and fast turns rather than a slow, small, booking-only room.
The lesson. Tourism does not just add demand, it inverts the week. Check which two days are your peak before you write the roster or the wage budget.
3. Chat Thai — 20 Campbell Street, Haymarket
The location. The original Chat Thai, on the Thai Town block behind George Street's southern end. FACT: the counted George Street block between Goulburn and Campbell streets carries 55,146 people a weekday and 76,856 on a weekend day, up 43% on 2019. FACT: the intersection block at Liverpool Street carries 94,426 and 109,646 — the busiest footpath in this dataset, up 28% on weekdays and 206% on weekends since 2019.
The demand engine. A shifted city. FACT: the L2 light rail opened along a pedestrianised George Street on 14 December 2019 and the L3 branch on 3 April 2020 (Transport for NSW). Our 2019 baseline predates both. INFERENCE: the growth at the southern end of George Street is consistent with the light rail and the removal of through traffic, though the counts alone cannot prove causation.
The trade-off, and why it works. FACT: this location trades daily until roughly 2am. LOCATALYZE ANALYSIS: the Liverpool Street block returns an evening-to-lunch ratio of 1.72 and peaks at 6pm on weekdays and 7pm on weekends — it is a genuine night street, unusual for a CBD. A venue open to 2am harvests hours that most of the city centre has already given up on. The trade-off is a late roster, a security cost and a kitchen that must hold quality at midnight.
The lesson. Infrastructure moves footpaths, and it moves them within a decade. Check what has been built near your site since the last lease was signed, and check the counts on both sides of it.
4. Firedoor — 23–33 Mary Street, Surry Hills
The location. A back street off Devonshire Street, not on Crown Street and not on Oxford Street. FACT: Firedoor cooks entirely over fire (Firedoor).
The demand engine. Reservations. FACT: Firedoor serves dinner Tuesday to Saturday and lunch on Friday only — six services a week, and none at all on Sunday or Monday. LOCATALYZE ANALYSIS: a roster that narrow only works if demand arrives booked. Passing trade cannot fill a room that is closed two days a week and open for one lunch.
The trade-off, and why it works. LOCATALYZE ANALYSIS: the site's defining feature is what it does not buy. There is no frontage premium on Mary Street, no corner position and no window onto a counted block. What the business buys instead is a Surry Hills postcode, a short walk from Central and the light rail, and a building that can take a wood-fired grill and its flue — a genuinely scarce requirement that would be far harder to satisfy, and far more expensive, on a main strip. The trade-off is total dependence on being worth a special trip.
The lesson. A destination pays for the building, not the footpath. If your concept generates its own demand, frontage is a cost you can decline — but only if it really does.
5. Mary's — 6 Mary Street, Newtown
The location. A former industrial building on a small street one block off King Street. FACT: the counted King Street block between Australia and Eliza streets — the nearest counted block to Mary's — carries 20,480 people a weekday and 27,956 on a weekend day, with an evening-to-lunch ratio of 1.93 and a 6pm weekday peak. FACT: that is the strongest evening ratio of any high-volume block in this dataset.
The demand engine. The King Street catchment, without the King Street frontage. LOCATALYZE ANALYSIS: Mary's sits close enough to be found by the people who came to Newtown to eat and drink, and far enough off the strip that it is not paying for a shopfront on the busiest block. That is a deliberate arbitrage, and the counts show what is being arbitraged.
The trade-off, and why it works. FACT: Mary's opens at 4pm Monday to Thursday and from midday Friday to Sunday. LOCATALYZE ANALYSIS: skipping weekday lunch on a 1.93 street costs almost nothing — the trade is not there — while opening from midday on the three days the ratio is best captures it. The trade-off is that a laneway address must be earned every single week: there is no accidental customer, and the venue lives or dies on being a place people already intend to go.
The lesson. Proximity is not frontage, and the difference between them is often the whole rent gap. Measure the block you are near, not the block you are on.
Comparing two Sydney addresses on the same strip? Locatalyze reads each frontage separately for demand, competition, catchment and rent pressure — because, as the counts show, one street name can hold four very different markets.
Free location score, map, data confidence and PROCEED / VERIFY / AVOID recommendation. Modelled financials stay optional.
Analyse a Sydney address6. Saint Peter — 161 Underwood Street, Paddington
The location. The Grand National Hotel, an 1890s corner pub on Underwood Street. FACT: Josh and Julie Niland moved Saint Peter here from its original Oxford Street shopfront, reopening in the pub in August 2024 (Broadsheet).
The demand engine, and the trap it avoids. This is the most instructive move in the article. FACT: the counted Oxford Street block at Ormond and Underwood carries 5,816 people a weekday and 11,790 on a weekend day — a ratio of 2.03. FACT: the neighbouring block at Newcombe and Elizabeth runs at 2.32, and on weekends its evening-to-lunch ratio is 0.41. LOCATALYZE ANALYSIS: Paddington's weekend is more than twice its weekday, and that weekend crowd is a daytime crowd. It is shopping, not dining. A restaurant reading only the weekend total would conclude that Oxford Street is a strong weekend dining street; the shape says it is a strong weekend retail street that empties before dinner.
The trade-off, and why it works. LOCATALYZE ANALYSIS: moving off Oxford Street onto a residential corner behind it trades passing visibility for a larger, cheaper building with accommodation above it and a local catchment that is present in the evening. FACT: the Oxford Street block Saint Peter left is down 34% on weekday footfall since 2019. INFERENCE: for a destination restaurant, the corner pub was the better asset — but this only holds because the business does not need the strip to find it.
The lesson. A busy weekend is not a busy dinner. Split the weekend count into daytime and evening before you price the site.
7. Fratelli Paradiso — 12–16 Challis Avenue, Potts Point
The location. A small Italian room on Challis Avenue, open since 2001 (Fratelli Paradiso). It sits in a precinct that has lived through the most violent demand shock in this dataset.
What happened around it. FACT: the counted Darlinghurst Road block between Baroda Street and Barncleuth Square recorded 9,530 pedestrians on a weekday in 2019 and averages 1,122 now — down 88%. FACT: NSW introduced lockout laws in February 2014 and lifted the last of them, in Kings Cross, on 8 March 2021 (ABC News). FACT: the ten surveys taken from late 2021 onward all sit between 798 and 1,446. LOCATALYZE ANALYSIS: the restriction was removed and the footfall did not come back.
The trade-off, and why it works. FACT: roughly 250 metres away on the same street, the Darlinghurst Road block at Bayswater and Roslyn is down only 10%, at 22,880 a weekday, with an evening-to-lunch ratio of 1.38. LOCATALYZE ANALYSIS: what collapsed was the late-night entertainment engine, not the neighbourhood. Fratelli Paradiso trades seven days from midday into the evening on a residential street a short walk from the worst of it, serving people who live there rather than people who travelled in to drink. INFERENCE: the residential catchment is what survived, and a venue built on it was insulated from a shock that removed nine tenths of the trade one block away.
The lesson. Ask what a precinct's demand actually rests on, and what would happen if that one thing were removed. Sometimes it is removed, and sometimes it does not return when the policy reverses.
8. Temasek — Shop 2, 100 George Street, Parramatta
The location. An arcade shopfront in Parramatta's centre, trading since 1992 (Temasek) — thirty-four years on one site. Parramatta is outside the counting boundary, so there is no footfall figure here and we do not invent one.
The demand engine. Sydney's second CBD, and a very large surrounding catchment. FACT: our own model scores Parramatta 67 out of 100 for restaurants with a CAUTION verdict, and marks its competition density 9 out of 10 — the highest of any Sydney suburb we score. FACT: Parramatta's office vacancy rate reached 23.5%, well above Sydney CBD's 13.3% (Australian Broker).
The trade-off, and why it works. LOCATALYZE ANALYSIS: on paper this is a hard site — maximum competition, a soft office market, and a suburb our model will not give a GO. Thirty-four years of continuous trading is the counter-evidence, and the plausible reading is that the competition score is measuring the wrong thing. FACT: Temasek trades Wednesday to Sunday only, lunch and dinner. INFERENCE: a five-day roster with two full weekend days suggests a business drawing on a residential and shopping catchment rather than an office one — which is precisely the catchment that a 23.5% office vacancy does not touch.
The lesson. A high competition score can mean a saturated market or a proven one. The distinguishing question is whether the cluster brings its own customers, and how long the incumbents have lasted.
9. Tan Viet Noodle House — 100 John Street, Cabramatta
The location. A shopfront on John Street, the spine of Cabramatta's Vietnamese precinct and a short walk from Cabramatta station, trading for more than thirty years.
The demand engine. A destination food precinct with a rail line running into it. FACT: our model scores Cabramatta 70 out of 100 for restaurants with a GO verdict — higher than Bondi (68, CAUTION), Paddington (68, CAUTION) and Sydney CBD (64, CAUTION). LOCATALYZE ANALYSIS: that is not a mistake in the model and it is not a claim that Cabramatta is more glamorous than Bondi. It is what happens when you score rent pressure (3 out of 10 in Cabramatta against 6 in Bondi and 9 in the CBD) and seasonality honestly alongside demand.
The trade-off, and why it works. FACT: Tan Viet trades seven days from 9am to 8pm. LOCATALYZE ANALYSIS: that is a daytime roster — it closes before most restaurants have finished their first sitting. INFERENCE: a precinct that draws visitors from across Sydney for lunch does not need a dinner service to fill a room, and skipping the evening removes a shift's wages, a late roster and the security costs that go with it. The trade-off is a low average spend per head, which only works at volume and at speed.
The lesson. Prestige and profitability are different variables. The cheapest defensible rent in a precinct people deliberately travel to is one of the strongest positions in this city.
10. Icebergs Dining Room and Bar — 1 Notts Avenue, Bondi
The location. Above the Bondi Icebergs pool at the southern end of the beach — a site with no through traffic, difficult parking, and one of the most photographed outlooks in the country.
The demand engine. The view, and the fact that the beach is a destination in its own right. Bondi is outside the counting boundary. FACT: our model scores Bondi 68 for restaurants with a CAUTION verdict, on a seasonality risk of 4 and a tourism dependency of 7.
The trade-off, and why it works. LOCATALYZE ANALYSIS: this is the clearest case in the article of a business that overrides its access problems rather than solving them. Nobody walks past 1 Notts Avenue on the way to somewhere else; every customer decided to come. FACT: Icebergs trades Tuesday to Sunday and closes Mondays. INFERENCE: the closure is what a seasonal, weather-exposed, destination-only trade looks like on a roster — the quietest day is not worth opening for. The trade-off is a revenue line that moves with the weather and the season in a way that an office-catchment restaurant's does not.
The lesson. Destination demand can beat bad access, but it does not beat seasonality. If the site's demand is weather-dependent, the cost base has to be able to shrink.
What the ten have in common
Laid side by side, the ten stop looking like ten stories and start looking like one. Every venue here has matched its trading pattern to the shape of its catchment, and where we can measure the footpath, the match is close enough to be checked.
Nobody here is trading against their street
Two venues at opposite ends of the same city centre make the point on their own. Rockpool sits on a 0.30 block and closes on Sunday. Saké sits on a 2.30 block and opens seven days. LOCATALYZE ANALYSIS: neither roster is more ambitious than the other — each is the correct answer to a different footpath, and swapping them would damage both businesses.
Six of the ten declined the best frontage
Firedoor is on a back street, Mary's is in a lane, Saint Peter moved off the strip onto a residential corner, Temasek is in an arcade, Tan Viet is a shopfront in a suburb no property page calls prestigious, and Rockpool is in a basement. LOCATALYZE ANALYSIS: frontage is a service you buy for customers who have not already decided to come. Six of these ten do not need that service, and have arranged not to pay for it.
The strongest evening ratios are outside the city centre
FACT: the best evening-to-lunch ratios among high-volume blocks are King Street, Newtown at 1.93 and George Street at Liverpool at 1.72. FACT: Pitt Street Mall returns 0.42 and Martin Place 0.46. LOCATALYZE ANALYSIS: with one exception — the southern, light-rail end of George Street — Sydney's city centre is a daytime restaurant market wearing a prestigious address, and the dinner market lives in the inner suburbs.
Durability clusters where rent pressure is low
The two longest-running venues here are Temasek, at thirty-four years, and Tan Viet, at more than thirty. FACT: both are in western Sydney, and both sit at or below the Sydney average on rent pressure — 3 out of 10 in Cabramatta and 5 in Parramatta, against a Sydney mean of 5.4. INFERENCE: with a low fixed cost base, a bad year is survivable. That is not a glamorous advantage, but over three decades it is the one that compounds.
Every one of them knows which two days matter
The weekend-to-weekday ratio ranges across this set from 0.30 to 2.30 — a factor of nearly eight. LOCATALYZE ANALYSIS: this single number determines the roster, the wage budget, the delivery schedule and the break-even calculation, and it is free to obtain for any site inside the City of Sydney. Most operators sign without it.
Five things operators get wrong about Sydney
“The CBD is the strongest market”
The CBD is the strongest daytime market and the most expensive one. FACT: our model scores Sydney CBD 64 for restaurants, a CAUTION — its demand is a perfect 10 out of 10, but rent pressure is 9 and competition 8, and the composite reflects all three. FACT: Darlinghurst scores 73, Newtown and Surry Hills 72. LOCATALYZE ANALYSIS: maximum demand at maximum cost is not the same thing as the best site, and the model is not being contrarian when it says so — it is doing arithmetic.
“The CBD has recovered”
Parts of it have doubled and parts of it have halved. FACT: George Street at Wynyard averages 30,994 pedestrians a weekday, down 58% on 2019. FACT: York Street is down 51%, Erskine Street 55%, and both Margaret Street blocks around 30–38%. FACT: over the same period Young Street at Circular Quay is up 167%, Alfred Street 61%, and the George Street block at Grosvenor 91%. LOCATALYZE ANALYSIS: the city centre did not recover evenly — it redistributed, north and south, away from the Wynyard office core. Our read on which CBDs are worth opening in covers the national picture.
“Pitt Street Mall is the best retail address in the country”
It may well be the most expensive, and it is not obviously the busiest any more. FACT: Pitt Street Mall averages 50,139 pedestrians a weekday, down 21% on 2019, with weekend counts down 40%. FACT: the George Street block at Liverpool now counts 94,426 — nearly twice the Mall. LOCATALYZE ANALYSIS: for a food business the Mall's problem is not its volume but its shape; at an evening-to-lunch ratio of 0.42 it is a lunch and shopping street, and dinner rent bought there is largely wasted.
“We'll be fine, it's a famous strip”
Fame is measured at the strip; trade is delivered at the block. FACT: King Street, Newtown ranges from 4,646 to 20,480 people a weekday across five counted blocks. FACT: Glebe Point Road ranges from 2,774 to 5,102. FACT: on Darlinghurst Road, two blocks roughly 250 metres apart are down 88% and 10% respectively since 2019. LOCATALYZE ANALYSIS: the strip name is the least informative thing on the lease.
“A good suburb score means a good site”
It does not, and this is the point our own product exists to make. A suburb score is a base rate — the average conditions across a whole postcode. FACT: two of the ten venues here, Saké and Chat Thai, are in suburbs (The Rocks and Haymarket) that our Sydney suburb layer does not cover at all, because neither is a walkable residential catchment in the sense the model is built for. LOCATALYZE ANALYSIS: we would rather say that plainly than extend a number into a place it was not designed for. FACT: Manly, one of Sydney's best-known dining destinations, scores 59 and returns a RISKY verdict on rent, competition and seasonality together — which is a statement about the average Manly lease, not a prediction about any particular restaurant in it.
One change worth watching
FACT: Martin Place's counted weekday footfall is up 42% on 2019, to 71,146, after the Metro station opened there on 19 August 2024. FACT: its weekend ratio is still 0.31. INFERENCE: the Metro has added commuters to an office block, not customers to a weekend. If the surrounding towers convert to residential over the next decade, that ratio is the number to re-check — it would be the first sign that the corridor has changed what it is, rather than simply moving more of the same people.
Read a Sydney street's shape yourself
Nine counted blocks, with the real figures behind every case study above. Switch between weekday and weekend to see how the same footpath changes hands.
Counted footfall · City of Sydney
Who does this street belong to?
Nine counted blocks, averaged across the three most recent council surveys. Pick one to see when its pedestrians actually arrive — and switch between weekday and weekend to see the street change hands.
Martin Place, between Castlereagh and Pitt
- Pedestrians per weekday71,146
- Weekend ÷ weekday0.31
- Evening ÷ lunch0.46
- Peak hour8am
Darker bars are the 6pm–11pm dinner window, which carries 16% of this block’s weekday. Bar heights are relative to this block’s own busiest hour, so compare shapes here and volumes in the figures above.
| Hour | Pedestrians |
|---|---|
| 6am | 1,138 |
| 7am | 1,150 |
| 8am | 10,142 |
| 9am | 7,820 |
| 10am | 2,472 |
| 11am | 2,584 |
| 12pm | 7,444 |
| 1pm | 9,642 |
| 2pm | 3,366 |
| 3pm | 3,414 |
| 4pm | 4,556 |
| 5pm | 6,358 |
| 6pm | 4,344 |
| 7pm | 2,420 |
| 8pm | 1,686 |
| 9pm | 1,146 |
| 10pm | 980 |
| 11pm | 484 |
The purest office corridor in the dataset. It peaks at 8am — the single largest hourly count anywhere in these counts — again at lunch, and is spent by seven. Weekday footfall is up 42% since 2019 with the Metro station, but the weekend ratio has not moved. More commuters, not more customers.
Locatalyze calculation from the City of Sydney Walking count surveys. Figures are the mean of the autumn 2025, spring 2025 and autumn 2026 rounds. Each round is a single fair-weather day counted from 6am to midnight, so shapes are far more reliable than small differences in level. The counts measure footpath movement, not entries or spend, and they stop at the council boundary.
The Sydney location checklist
Everything below is either free or costs a weekend of your own time. None of it requires a consultant, and all of it should happen before you sign.
Twelve checks before you sign a Sydney hospitality lease
Allow two weekends and one full working day
Applying this to your own site
The method in this article is not proprietary. It is: find out who walks past, when they walk past, what they came for, and whether your concept and your roster can convert them at the rent you are being asked to pay. Inside the City of Sydney you can do most of it from a laptop. Everywhere else you do it with a clipboard, which is slower but not harder.
What Locatalyze adds is the rest of the picture at one specific address — the competitors already trading within walking distance, the catchment's size and composition, the rent pressure relative to comparable frontages, and the demand signal behind them — assembled into a verdict with the evidence attached to every number. A suburb score narrows a shortlist. It is the address that gets signed. If you want to see how we build a score before you use one, the methodology is public, every suburb score in this article is on its own page under Sydney restaurant locations, and our Sydney restaurant location guide covers the suburb layer in more depth.
Check the block, not the suburb
Suburb scores narrow a shortlist. Locatalyze reads one specific Sydney frontage for demand, competition, catchment and rent pressure, and shows the evidence behind every number.
Analyse a Sydney address freeMethod, evidence and limits
The pedestrian data. All footfall figures in this article are our own calculations from the City of Sydney's Walking count surveys, downloaded on 6 September 2026 — 5,453 records across 130 counted sites, from spring 2013 to autumn 2026. The council counts at roughly 100 locations from 6am to midnight, in fair weather, on one weekday and one weekend day each March and October. Current figures are the mean of the three most recent surveys (autumn 2025, spring 2025, autumn 2026). Comparisons to 2019 use the mean of that year's surveys as the baseline.
Why we average. Each survey is a single day's count, so individual readings are noisy: at one George Street site the two 2019 surveys differ by a factor of two. Averaging three recent surveys against the 2019 pair is the least misleading comparison available, and we have quoted change only where the direction is consistent across the whole recent record. We also found and excluded an apparent counter fault affecting several sites between spring 2015 and autumn 2017, where readings drop to under a tenth of their neighbouring years and then recover; those years are not used anywhere in this article.
What the counts cannot do. They cannot distinguish a person arriving from a person leaving, they do not say who anyone is or what they spent, and they stop at the City of Sydney boundary. Three of the ten venues here — Temasek in Parramatta, Tan Viet in Cabramatta and Icebergs in Bondi — sit outside the counted area, and a fourth, Firedoor, sits inside the boundary but with no counted block adjacent to it. We have not estimated figures for any of them.
The scores. Suburb scores quoted here are live outputs of the Locatalyze scoring engine, computed from five 1–10 factors and the published business-type weights, with GO at 69 and above and CAUTION from 60. They are not hand-entered, and we have not adjusted any of them for this article. We have not scored any individual venue, and no score in this piece should be read as a rating of a business.
The venues. Address, trading days and operating status for all ten were checked against each venue's own published information on 6–7 September 2026. Nothing about revenue, rent, covers or profitability is asserted anywhere in this article, because none of it is public.
Locatalyze is decision-support for location choice. Nothing here is legal, valuation, financial or investment advice, and no venue named in this article is a customer, partner or sponsor of Locatalyze. Last checked 7 September 2026.
Sources
Primary and industry sources
Related Locatalyze research
Frequently asked questions
About the author
Locatalyze Research Team
Location intelligence, Locatalyze
The Locatalyze research team builds the location-scoring models behind the platform and writes up what the evidence shows for Australian operators.
Tools first — then a full report for your address
Free rent, viability, and break-even checks. Upgrade when you are ready for competitors, map, and numbers for a specific site.
No signup required for tools
