A kilometre-tall billboard would be visible to the geometric horizon about 113 km away, but in Indian air it could only be read from roughly 3 to 15 km, because haze, not letter size, sets the limit: on a typical 10 km visibility day, letters about 20 m tall are enough, so 98% of the board's height is wasted on legibility.
Calculated in this post for a board 1,000 m tall (the scenario's premise) from the Earth's radius (6,371 km), and from Koschmieder's law with visibility figures from Jaswal et al. (2013): reading range is about 0.7 times visibility.
- Geometry lets a 1,000 m sign be seen 113 km away; Indian haze cuts reading range to about 0.7 times visibility.
- On a typical 10 km visibility day, letters about 20 m tall are enough.
- At Mumbai's design wind, a 1,000 m by 250 m board takes about 940 MN, the weight of roughly 95,000 tonnes.
Calculated in this post from Mumbai's 44 m/s basic wind speed in IS 875 (Part 3), with an assumed board of 1,000 m by 250 m and an assumed force coefficient of 1.8 (my estimates).
- About 1.6 lakh people could read it at once in Mumbai, at roughly ₹35,000 per thousand impressions.
Calculated in this post from Mumbai's 2011 census density of 20,620 per km² and visibility data, with these example inputs: a 7.1 km reading radius, half the circle as land, and 10% of people with a clear view.
- The version that works: an LED wrap on an existing 200-300 m tower.
The board's real job would be clearing the skyline so that a few lakh people in a dense city can see it at once. Structurally it is possible but brutal: at Mumbai's design wind speed, a 1,000 m by 250 m board takes about 940 meganewtons of wind force, the weight of roughly 95,000 tonnes, and it must be shaped to avoid the vortex shedding that makes supertall towers sway.
Calculated in this post from Mumbai's 44 m/s basic wind speed in IS 875 (Part 3) and an assumed 1,000 m by 250 m board; the steps are shown in the wind section below.
Outdoor advertising has always been a contest of size against attention. A highway hoarding in India is typically a few tens of metres wide, and the city of Mumbai caps hoardings at 40 ft by 40 ft.
The tallest structure humans have built, the Burj Khalifa, is 828 m; the Jeddah Tower, under construction again since January 2025, aims for over 1,000 m. So the question is fair: what would a billboard of that size actually achieve, and what would it take to keep it standing? I'll solve legibility, atmosphere, wind and reach in turn.
How big do letters need to be to read from far away?
On 13 July 1923, developers above Los Angeles dedicated what is now the world's most famous billboard. It advertised a housing estate called Hollywoodland: thirteen white block letters, each about 50 ft (15 m) high and 30 ft wide, spread across 450 ft of hillside.
About 4,000 light bulbs flashed "HOLLY", "WOOD" and "LAND" in turn, then all together, with a searchlight underneath. It cost $21,000 and was meant to stand for only a year and a half.
The sign outlived its sales job. The lights went off around 1933, the "LAND" came down in 1949, and in 1978 the letters were rebuilt in steel on concrete. By the legibility rule below, 15 m letters read at about 5.5 km, close to what Indian haze allows on an ordinary afternoon.
Calculated in this post from the Hollywood Sign's letter height of 15.2 m (50 ft) and the 360 highway-design legibility index from FHWA: 15.2 × 360 ≈ 5.5 km.
Wikipedia: Hollywood Sign (the 1923 Hollywoodland advertisement)Sign engineers use a legibility index: each centimetre of letter height buys a fixed reading distance. The US Federal Highway Administration estimates about 40 feet per inch of letter height for drivers with 20/20 vision, about 20 feet per inch for 20/40 vision, and the US highway manual designs at 30 feet per inch.
Convert those to metric and you get a clean rule: reading distance \( d = L \times h \), where \( h \) is letter height and \( L \) is the legibility index.
- 20/20 eyes, FHWA. \( 40\ \text{ft/in} = 12.19\ \text{m} / 0.0254\ \text{m} \Rightarrow L = 480 \): a 1 m letter is legible at 480 m.
Calculated in this post from the FHWA's legibility factor of 40 feet per inch, converted to metres.
- Highway design value. \( 30\ \text{ft/in} \Rightarrow L = 360 \).
- 20/40 eyes. \( 20\ \text{ft/in} \Rightarrow L = 240 \).
- Check against optics. The 20/20 standard is a letter subtending 5 arcminutes, which gives \( L = 1/(5 \times 2.909\times10^{-4}) = 688 \). The FHWA value of 480 is lower because real signs are read at a glance, from moving vehicles, at imperfect contrast.
The most you could ever need is set by the horizon. For a sign whose top is at height \( H \) above flat ground, on an Earth of radius \( R = 6{,}371 \) km, the line of sight grazes the surface at
The case for a giant board, on paper
A viewer whose eyes are 1.7 m above the ground adds \( \sqrt{2R \times 1.7} \approx 4.7 \) km, so about 118 km in total, a little more with normal atmospheric refraction bending light around the curve. To read a word at that distance with the design index of 360, each letter must be \( 1.18\times10^{5}/360 \approx 330 \) m tall. A kilometre-tall board fits three such lines. On paper, that is the case for building one.

Why can't you read it from 100 km away?
Because the air in between washes out contrast long before the letters get too small. Atmospheric haze, not letter size, is the binding limit.
The governing theory is Koschmieder's law, derived in 1924: the apparent contrast of a dark object against the sky decays exponentially with distance as light is scattered by particles and moisture.
Here \( C_0 \) is the contrast at the sign, \( \beta \) is the extinction coefficient of the air, and \( V \) is the meteorological visibility, defined as the distance at which contrast falls to 2%, the threshold for merely detecting a large dark object. Reading needs more contrast than detecting. Assumptions (my estimates): a board with contrast \( C_0 = 0.8 \), and a reading threshold of \( \varepsilon = 0.05 \).
- Solve for reading range. \( C_0 e^{-\beta d} = \varepsilon \Rightarrow d_{\text{read}} = \frac{\ln(C_0/\varepsilon)}{\beta} = \frac{V}{3.912}\ln\frac{0.8}{0.05} = 0.709\,V \).
- Typical Indian afternoon, \( V = 10 \) km. \( d_{\text{read}} \approx 7.1 \) km.
- Letter height needed. \( h = 7{,}100/360 \approx 20 \) m.
- Hazy winter morning, \( V = 4 \) km. \( d_{\text{read}} \approx 2.8 \) km; letters of 8 m are enough.
With these example inputs (V = 4 km, C0 = 0.8, threshold 0.05) this is calculated in this post from Koschmieder's law and the 360 legibility index from FHWA. The 4 km figure is the poor-visibility threshold used by Jaswal et al..
- A rare clear day, \( V = 50 \) km. \( d_{\text{read}} \approx 35 \) km; letters of 100 m are needed. Even then the 113 km horizon is out of reach.
With these example inputs (V = 50 km, C0 = 0.8, threshold 0.05) this is calculated in this post from Koschmieder's law and the 360 legibility index from FHWA.
How often is Indian air clear?
A study of 279 weather stations from 1961 to 2008 found that the share of afternoons with visibility above 10 km fell from 76.1% to as low as 30.6% of days, and morning days with visibility under 4 km rose to as much as 27.3%. The decline was worst over the Indo-Gangetic plain, which is where most Indians live.
Haze, not letter size, is the binding limit: past about seven kilometres on an ordinary Indian afternoon, a bigger letter buys nothing.
The authors analysed visibility records from 279 surface stations, including airports, over 48 years. Morning poor-visibility days (under 4 km) rose from 6.7% to 27.3% across the range of stations, a trend of +3.3% of days per decade, while afternoon good-visibility days (over 10 km) fell from 76.1% to 30.6%, a decline of 8.6% of days per decade. Declines were strongest over the Indo-Gangetic plains and east and northeast India, and tracked rising humidity, falling wind speed, fog and aerosols. The limit: visibility was recorded by human observers at fixed hours, so it samples the day rather than measuring it continuously.
Read the abstract and paper links (free)See the data as a table
| Reading range (km) (km / m) | Letter height needed (m, ÷10) (km / m) | |
|---|---|---|
| 2 km | 1.4 | 0.4 |
| 4 km | 2.8 | 0.8 |
| 7 km | 5.0 | 1.4 |
| 10 km | 7.1 | 2.0 |
| 15 km | 10.6 | 2.9 |
| 20 km | 14.2 | 3.9 |
| 30 km | 21.3 | 5.9 |
| 50 km | 35.5 | 9.9 |
Source: My calculation from Koschmieder's law (C0 = 0.8, threshold 0.05) and the FHWA-based legibility index of 360 · Letter heights are divided by 10 to share the axis: 2.0 means 20 m.
How much wind would push on it?
About 940 meganewtons at Mumbai's design wind, equal to the weight of about 95,000 tonnes, pushing sideways with its centre more than 500 m up. This is the hardest part of the problem.
India's wind code, IS 875 (Part 3): 2015, gives the design wind speed as \( V_z = V_b k_1 k_2 k_3 k_4 \) and the design pressure as \( p_z = 0.6\,V_z^2 \) in N/m². Mumbai's basic wind speed \( V_b \) is 44 m/s. The height factor \( k_2 \) is tabulated only up to 500 m; at 200 m it is about 1.27 for open terrain (InfraLens). Assumptions:
- Board 1,000 m tall and 250 m wide: area \( A = 2.5\times10^{5} \) m².
- \( k_1 = k_3 = k_4 = 1 \). Above the code's table I extend \( k_2 \) with a power law fitted to the 200 m value: \( k_2(z) = (z/10)^{0.08} \). This is my extrapolation, so treat the result as an estimate.
- Force coefficient for a solid flat board \( C_f = 1.8 \), my estimate in the range codes use for free-standing hoardings.
- Wind at the top. \( k_2(1000) = 100^{0.08} = 1.445 \), so \( V_{1000} = 44 \times 1.445 = 63.6 \) m/s.
- Pressure at the top. \( p = 0.6 \times 63.6^2 = 2{,}427 \) Pa.
- Average pressure over the height. Because \( p \propto z^{0.16} \), the mean is \( p_{\text{top}}/1.16 = 2{,}092 \) Pa. That matches the 1.5 to 2.5 kPa range the Burj Khalifa's wind tunnel tests found for most of its cladding, a useful cross-check.
- Total force. \( F = C_f\,\bar p\,A = 1.8 \times 2{,}092 \times 2.5\times10^{5} = 9.4\times10^{8} \) N, about 940 MN, or the weight of \( 9.4\times10^{8}/9.81 \approx 96{,}000 \) tonnes.
- Overturning moment. The resultant acts at \( 0.537H \), so \( M = 9.4\times10^{8} \times 537 \approx 5.0\times10^{11} \) N·m at the base.
Sensitivity: wind pressure goes with the square of speed, so Bengaluru, in the interior Karnataka band of 33 m/s basic wind speed (InfraLens), cuts the force by \( (33/44)^2 = 0.56 \), to about 530 MN. Making the board a 50% open mesh, the way large building wraps are built, roughly halves the force again. Halving the width to 125 m halves it too.
Calculated in this post from the 44 m/s and 33 m/s basic wind speeds in IS 875 (Part 3): (33/44)² = 0.56, and 0.56 × 940 MN ≈ 530 MN.
The subtler problem: vortex shedding
The second wind problem is subtler: vortex shedding. Wind flowing past a bluff body sheds alternating vortices from each side, creating a sideways push at frequency \( f_v = S\,U/w \), where \( S \) is the Strouhal number (about 0.14 for sharp-cornered buildings), \( U \) the wind speed and \( w \) the width. If \( f_v \) matches the structure's natural sway frequency \( f_b \), the motion resonates.
Peter Irwin, the wind engineer behind the Burj Khalifa, gives natural frequencies of 0.1 to 0.125 Hz for 100-storey towers and as low as 0.05 Hz for some supertalls.
With a stiff structure at 0.05 Hz, the critical speed of 89 m/s is above the design wind of about 64 m/s, which is safe. A lighter, cheaper frame at 0.03 Hz would have a critical speed of 54 m/s, inside the design range, and would rock itself apart. The cure is the Burj's: change the width with height so vortices do not shed in step along the whole structure.
The authors describe how wind tunnel tests on 1:500 force-balance and aeroelastic models, plus 1:50 high-Reynolds-number tests in a 9 m by 9 m tunnel, shaped the tower. The 50-year three-second gust at 10 m was estimated at 37.7 m/s; most cladding pressures fell between 1.5 and 2.5 kPa; and predicted peak accelerations for a five-year return period fell from over 37 milli-g to about 19 milli-g by re-orienting the tower, adjusting its shape and stiffening it, with half the gain from better wind statistics. The limit: Dubai's climate is not Mumbai's, and the building was designed to be lived in, a harder comfort target than a billboard.
Read the paper (free)How many people could actually read it?
A few lakh at any one time in Mumbai, set by the haze radius and how many people have a clear view, not by the board's size.
- Reading radius. On a 10 km visibility day, \( r = 7.1 \) km, an area of \( \pi r^2 = 158 \) km².
With these example inputs (10 km visibility), calculated in this post from Koschmieder's law: r = 0.709 × 10 km, and area = π r².
- People in that circle. Mumbai's 2011 census density was 20,620 per km². Mumbai is a peninsula, so assume half the circle is land: \( 158 \times 0.5 \times 20{,}620 \approx 16 \) lakh people.
- People who can see it. Buildings, roofs and trees block most sightlines. If 10% of those people are outdoors or at a window facing it (my estimate), about 1.6 lakh people can read it at any moment.
- Daily impressions. If each person in range catches it twice a day, that is about 32 lakh impressions a day, or roughly 12 crore a year.
With these example inputs (twice a day per person, my assumption), calculated in this post from the 1.6 lakh people above, which rest on Mumbai's census density: 1.6 lakh × 2 × 365 ≈ 12 crore a year.

How many people can read your giant billboard?
| Design | Height | Readable range on a 10 km day | Wind force at Mumbai design speed | Rough cost |
|---|---|---|---|---|
| Legal Mumbai hoarding (40 ft × 40 ft) | ~12 m board | ~4 km for 12 m letters, but blocked by buildings | ~0.3 MN (my estimate) | Lakhs |
| LED wrap on a 300 m tower (100 m × 300 m) | 300 m | 7 km, with clear sightlines over most roofs | ~90 MN, shared with the tower | Hundreds of crores (estimate) |
| Free-standing 1 km board (250 m wide) | 1,000 m | 7 km; the extra height adds nothing on hazy days | ~940 MN | Comparable to a supertall: the Burj Khalifa cost about $1.5 billion |
Could it be built, and would it pay?
It could be built: the Jeddah Tower has passed 430 m on its way to over 1,000 m. Paying for it is harder.

Take the Burj Khalifa's $1.5 billion as a floor for a kilometre-scale structure, roughly ₹12,750 crore at an exchange rate near ₹85 per dollar (approximate). Spread over 30 years with no interest, that is about ₹425 crore a year, or about 0.35% of India's entire 2025 ad market of ₹1,21,339 crore, from a single billboard.
Calculated in this post from the Burj Khalifa's cost of $1.5 billion, at an assumed ₹85 per dollar (my example input): ₹12,750 crore, ÷ 30 years = ₹425 crore a year, ÷ India's 2025 ad market of ₹1,21,339 crore = 0.35%.
The cost per thousand
At about 12 crore readable impressions a year from the reach estimate above, the cost per thousand impressions is about ₹35,000. A YouTube bumper costs ₹250 to ₹400 per thousand.
Calculated in this post from the ₹425 crore a year above and 12 crore impressions a year: ₹425 crore ÷ 12 crore × 1,000 ≈ ₹35,000 per thousand. The ₹250 to ₹400 comparison is from Mega Digital.
The giant board loses by two orders of magnitude, unless it is famous enough to be photographed and shared, which is the only reason anyone builds a landmark.

This is why the nearest real-world version is not a free-standing board. The Burj Khalifa's LED façade, which set Guinness records for the largest LED-illuminated façade in 2015 and 2019, shows that a screen on a supertall works: it gets the height for free, because the building already exists for other reasons.
For Indian marketers the lesson is the one the data on ad spend by medium keeps teaching: outdoor works when it rides on existing attention, a junction, a stadium, a skyline, and fails when it tries to manufacture attention by size alone. The same logic decides whether trade-show booths pay back.
Typography still matters at any size; Artblog's piece on serif versus sans-serif readability covers why heavy, open letterforms win at distance.
A kilometre-tall billboard can be built, since the Jeddah Tower is heading past 1,000 m, but at full height it is mostly wasted: Koschmieder's law limits reading to about 0.7 times the visibility, around 7 km on a typical Indian afternoon, where 20 m letters suffice, while the horizon-limited 113 km range almost never occurs.
The deciding numbers are about 940 MN of design wind load in Mumbai and roughly ₹35,000 per thousand impressions. The version that works is an LED façade or mesh wrap on an existing 200 to 300 m tower, with letters around 20 m tall, a shape that breaks up vortex shedding, and a cost in the hundreds of crores rather than thousands.
Both figures are calculated in this post: the wind load from Mumbai's 44 m/s design wind speed and the cost per thousand from the Burj Khalifa's $1.5 billion cost.
What's new
- Jeddah Tower reaches 430 m and 106 floors on its way past 1,000 m Khaleej Times, Aug 2026
- VJTI report: the collapsed Ghatkopar hoarding sat on a foundation built for 49 km/h winds Free Press Journal, Jun 2024
- India's ad market projected at ₹1,30,416 crore in 2026, with digital heading to 70% by 2027 BestMediaInfo, Feb 2026
What if next?
- What if a marketing team ran only on AI agents? Production cost falls to near zero; the bottleneck moves to taste and review.
- What if you stopped paid ads for 90 days? How fast demand decays when the billboard goes dark.
- What if Google AI Overviews answered every query? The biggest billboard on the internet stops sending clicks.
Questions people ask
How big should billboard letters be?
Divide the reading distance by about 360 (the 30 ft per inch highway rule) to get letter height in the same units. A sign read from 150 m needs letters about 40 cm tall; for drivers with good eyes, dividing by 480 is enough.
Calculated in this post from the 360 and 480 legibility indices in FHWA's legibility testing: 150 ÷ 360 ≈ 0.42 m.
How far can you see from 1 km up?
The geometric horizon is about 113 km, from \( \sqrt{2RH} \). In practice haze cuts that to the visibility distance, which across Indian stations is often under 10 km in the afternoon and under 4 km on many winter mornings.
Why do billboards fall down in storms?
Wind pressure grows with the square of wind speed, and a large flat board catches all of it. The 2024 Ghatkopar hoarding was built for 49 km/h winds against Mumbai's 158 km/h standard, about a tenth of the required strength.
What is the tallest structure ever built?
The Burj Khalifa in Dubai, at 828 m, completed in 2009 and opened in 2010. The Jeddah Tower in Saudi Arabia is designed to exceed 1,000 m and had reached 430 m by August 2026.
Sources
- FHWA: Information as a source of distraction, Chapter 3, Legibility testing (2015; source for the 40, 30 and 20 ft per inch legibility factors)
- Jaswal et al., Natural Hazards: Decline in horizontal surface visibility over India (1961–2008) (2013; source for poor and good visibility day shares)
- Irwin and Baker: The wind engineering of the Burj Dubai tower (source for cladding pressures, gust speed and acceleration figures)
- Irwin, Physics Today: Vortices and tall buildings, a recipe for resonance (2010; source for the Strouhal number of 0.14 and tall-building natural frequencies)
- InfraLens: IS 875 (Part 3): 2015 walkthrough (source for \( V_z = V_b k_1k_2k_3k_4 \), \( p_z = 0.6V_z^2 \), k2 values to 200 m, the 500 m table limit and Mumbai's 44 m/s)
- Free Press Journal: Ghatkopar billboard had weak foundation, VJTI report (Jun 2024; source for the hoarding size, deaths, and the 49 and 158 km/h figures)
- Wikipedia: Burj Khalifa (source for the 828 m height, $1.5 billion cost and LED façade records)
- Khaleej Times: Jeddah Tower rises past half of Burj Khalifa's height (Aug 2026; source for 430 m and 106 floors)
- Wikipedia: Mumbai (source for the 2011 census population density of 20,620 per km²)
- BestMediaInfo: dentsu-e4m India AdEx report (Feb 2026; source for the 2025 and 2026 ad market)
- Wikipedia: Hollywood Sign (source for the 1923 Hollywoodland sign: letter size, 4,000 bulbs, $21,000 cost and later history)
- Mega Digital: YouTube ads cost in India (Jun 2026; source for the ₹250–400 bumper CPM comparison)
