You would market to an alien civilisation the way a peacock markets to a peahen: with a signal so expensive that it proves itself, because there is no shared language, no product you can ship and no reply for thousands of years. Physics reduces the job to three tasks: be detectable (an effective radiated power near 1017 watts for a galaxy-wide beacon), be decodable (a message that explains its own format, like the 1974 Arecibo message), and be credible (a costly, repeated pattern no natural process makes). Selling is impossible; branding is not.
- Arecibo's 1974 message was 1,679 bits, sent once, in under three minutes, toward a cluster 25,000 light-years away.
- Its effective power was about 1013 W by my estimate; a galaxy-wide beacon needs over 1017 W (Benford et al.).
- Round trip to M13 is 50,000 years; at a normal firm's death rate, the chance your company is alive for the reply is effectively zero.
- Only information is worth trading: moving 1 kg at 10% of light speed takes about 125 GWh of kinetic energy.
- Credibility comes from cost: the handicap principle, the same logic Ambler and Hollier found in TV ads.
This is the hardest marketing brief imaginable, which makes it a useful one. Every crutch is gone: no shared culture, no shared words, no feedback, no payment. What is left is the bare machinery of attention, trust and exchange, and it turns out each part has a law of physics attached.
On the afternoon of 16 November 1974, Arecibo Observatory in Puerto Rico held a ceremony for a major upgrade of the telescope, whose dish is 1,000 feet across. As the finale, the telescope's radar beamed a message at 2,380 MHz toward the globular cluster M13, about 25,000 light-years away and home to some 300,000 stars. The message was 1,679 binary digits, written by Frank Drake and colleagues with help from Carl Sagan. When the audio tone of the transmission began, much of the audience walked out of the tent to stare up at the telescope.
It lasted about three minutes and was sent once. One astronomer later called it a symbolic event, to show it could be done. By 1999, the message had covered about one thousandth of its journey.
Cornell Chronicle: 25th anniversary of Earth's first attempt to phone E.T. (1999)What does marketing look like with nothing shared?
Strip marketing to its fundamentals and it has four stages: get noticed, be understood, be believed, and exchange value. With aliens, each stage hits a hard physical limit, and the table shows which one.
| Stage | On Earth | For an alien audience | The law in the way |
|---|---|---|---|
| Attention | Media reach, frequency | Be detectable above noise | Inverse-square law |
| Comprehension | Shared language | Self-decoding message | Information theory |
| Trust | Reputation, reviews | A signal too costly to fake | Handicap principle |
| Exchange | Product for money | Information for information | Speed of light, kinetic energy |
| Retention | Repeat purchase | Survive until the reply | Corporate hazard rate |
How powerful must the signal be to get noticed?
To be noticed across the galaxy, a radio beacon needs an effective isotropic radiated power (EIRP) above about 1017 watts, according to cost-optimisation work by James, Gregory and Dominic Benford. Arecibo in 1974 managed roughly 1013 W, about 7,700 times less.
Assumptions
- Transmitter power \( P = 450 \) kW (Wikipedia: Arecibo message), frequency 2,380 MHz (Cornell Chronicle, 1999), dish diameter \( D = 305 \) m (the 1,000-foot dish, rounded; Wikipedia gives 306.18 m).
- Aperture efficiency \( \eta = 0.5 \). This is my estimate for a large dish, not a published Arecibo figure.
- Distance to M13 \( d = 25{,}000 \) light-years \( = 2.37\times10^{20} \) m.
- Wavelength. \( \lambda = c/f = 3.0\times10^8 / 2.38\times10^9 = 0.126 \) m.
- Gain. \( \pi D/\lambda = \pi \times 305/0.126 = 7{,}605 \); squared, \( 5.78\times10^7 \); times 0.5, \( G \approx 2.9\times10^7 \), or about 75 dBi.
- EIRP. \( 4.5\times10^5 \times 2.9\times10^7 = 1.3\times10^{13} \) W.
- Flux at M13. \( 4\pi d^2 = 7.0\times10^{41}\ \text{m}^2 \), so \( S \approx 1.9\times10^{-29}\ \text{W/m}^2 \).
- Gap to a galactic beacon. \( 10^{17}/1.3\times10^{13} \approx 7{,}700 \).
The inverse-square law is the cruel part: double the distance and you need four times the power. That is why "reach" in this market is paid for in gigawatts, and why the Benfords found a neat rule: at the cheapest design, the money is split equally between antenna area and transmitter power. It is the same trade-off a media planner makes between frequency and reach, solved with calculus.
The authors ask what a beacon would cost its builder and minimise that cost for a fixed power density at the receiver. With costs linear in transmitter power and antenna area, the minimum comes when spending is split equally between the two, a rule they say industry already uses. A galaxy-scale beacon needs EIRP above 1017 W, emitted power above 1 GW and antenna area above 1 km². Cost-optimal beacons would use narrow, pulsed beams near 10 GHz and revisit any one target rarely. The limit: cost parameters are 2008 values, and the analysis assumes the builders think about money the way we do.
Read the paper (free)See the data as a table
| log10 of power (log10 W) | |
|---|---|
| Galaxy-scale beacon minimum (Benford et al.) | 17 |
| Detectable to a 10-year Green Bank survey | 16.7 |
| Arecibo message, 1974 (estimated EIRP) | 13.1 |
| Largest transmitter ruled out on 3I/ATLAS (2026) | 2.04 |
Source: Benford et al. (arXiv 0810.3964); UCLA SETI via Astrobiology.com (May 2026); SETI Institute (June 2026); Arecibo figure is my estimate from 450 kW and a 305 m dish · Each step of 1 is ten times more power. The 3I/ATLAS limit is for a nearby object, not a distant star.

How do you write a message nobody can read yet?
You write a message that teaches its own format, starting from things any physicist must share: counting, primes and chemistry. Arecibo's 1,679 bits work because 1,679 = 23 × 73, a product of two primes, so there is only one sensible way to lay the bits out as a picture.
The content then built up step by step:
- the numbers 1 to 10 in binary, the only "alphabet" lesson;
- the atomic numbers of hydrogen, carbon, nitrogen, oxygen and phosphorus;
- the chemistry of DNA's building blocks and a double helix;
- a human figure, its height and the world population;
- the Solar System, with Earth marked, and the telescope that sent it.
Mahon proposes a score that splits the shortest description of a dataset into a structured part and a random residual, and counts only the structured part. Random text scores 0; English, German and Irish text score about 13,500 to 14,700; white noise images score 384 against about 500,000 for photographs. Tested on the Arecibo message, the intended 73-by-23 layout scored highest, above 9,000, while most other aspect ratios fell below a random-bitmap benchmark. The limit, which the author states: there is no formal proof the score separates meaning from noise, and the test set was small.
Read the paper (free)That result is the alien version of a brand guideline. A message is readable when it has compressible structure: patterns that repeat and nest. Noise has none; meaning has a lot. Your brand's job, at any distance, is to be the most structured thing in the receiver's sky.
Why would aliens believe anything we say?
They would believe a signal only if it is too costly to fake, which is Amotz Zahavi's handicap principle: an honest signal must carry a cost that a cheater cannot pay. A peacock's tail is credible because a sick peacock cannot grow one. A 1017-watt beacon is credible for the same reason: no natural process and no small civilisation produces a narrow, repeating, prime-numbered pulse at that power.
Marketing research found the same logic on Earth. Tim Ambler and Ann Hollier tested TV ads online in 2004 and found that ads viewers saw as more expensive were more persuasive, mostly because they raised perceived brand quality. Their title says it plainly: the waste in advertising is the part that works. I worked through the biology in the peacock's tail and your ad budget.
To an alien, your power bill is your brand promise. Nothing else you say can be checked.
Three ways to make the signal costly
- Power: the EIRP itself. Cheap to describe, ruinous to fake.
- Persistence: repeat for centuries. A single three-minute burst, like 1974, proves capability once; a schedule proves a civilisation that keeps promises.
- Precision: encode a constant both sides can measure, such as the hydrogen line, so the receiver can check you got the physics right.

What could we actually sell, and when would we get paid?
The only product worth sending is information, and the "payment" would arrive after your company is almost certainly dead. Physical goods fail on energy; services fail on time.
Why goods are out
Moving one kilogram at 10% of light speed takes about 125 GWh of kinetic energy before any losses. With an assumed ₹8 a unit (an example input), that is about ₹100 crore per kilogram, and the parcel still takes 42 years to reach the nearest star. A radio bit costs a tiny fraction of that and travels ten times faster.
Calculated in this post with these example inputs (1 kg, 10% of the speed of light, and ₹8 a unit), from the speed of light (299,792,458 m/s, Wikipedia) and the 4.25 light-year distance to Proxima Centauri (Wikipedia: Proxima Centauri).
Why the invoice never clears
- Lifetime value with discounting. \( \text{CLV} = \sum_t m/(1+r)^t \). With these example inputs (\( t = 50{,}000 \) years, the round trip to M13, and an assumed discount rate \( r = 1\% \)), a reply is worth \( 1.01^{-50000} = e^{-497} \) of its face value.
Calculated in this post from the round trip to M13, twice the 25,000 light-years (Cornell Chronicle, 1999), and an assumed 1% discount rate.
- Survival of the seller. Listed firms die at about 5.8% a year (Daepp et al., 2015), so \( P(\text{alive}) = e^{-0.058\times50000} = e^{-2900} \).
- Nearest case. Proxima Centauri is 4.25 light-years away: an 8.5-year round trip and a 61% chance the firm survives it. That is the only alien market a business plan could contain.
Calculated in this post from the 4.25 light-year distance (Wikipedia: Proxima Centauri) and the 5.8% yearly death rate (Daepp et al., 2015).
Use this to plan your own interstellar campaign.
Plan an interstellar campaign
So what is the alien marketing plan?
The plan is a brand campaign with no sales target: one costly, decodable, repeating signal, aimed at the nearest likely listeners, run by an institution built to outlast any company. In order:
- Target. Pick stars within about 20 light-years, where a reply can arrive within one working life.
- Power. Budget for EIRP near 1017 W if you want the galaxy, or far less for one nearby star; split spending equally between antenna and transmitter.
- Format. Lead with primes and counting, then measurable physics, then chemistry, then us. Test the message with a structure score like Mahon's before sending.
- Proof. Repeat on a fixed schedule for decades. Consistency is the only reputation that survives light-years.
- Offer. Information only: what we know, in exchange for what they know.
Selling to aliens is impossible: goods cost about 125 GWh per kilogram to move at a tenth of light speed (calculated above), and at a normal firm's 5.8% yearly death rate the seller will not exist when a reply comes from M13. Branding to aliens is possible today with changes: a beacon near 1017 W of effective power, a self-decoding message, and a schedule kept for decades by an institution, not a company. The deciding number is distance: within about 20 light-years, the loop closes in a lifetime.
The same skeleton, signal, proof and pattern, is what I use for real campaigns on Earth; the Bazaar world page collects that work. For an Earth-bound version of oversized signals, see what if you wrote an ad on the Moon and a billboard a kilometre tall.
What's new
- SETI Institute scans interstellar object 3I/ATLAS: 74 million signals, no technology found SETI Institute, Jun 2026
- Ten years of UCLA searches limit transmitters to under 0.0063% of nearby stars Astrobiology.com, May 2026
- A new complexity score picks out the Arecibo message's true layout Entropy, Mar 2025
What if next?
- What if every ad had to be 100% true? Aliens can only check physics; humans can check far more, if the law makes them.
- What if AI wrote every ad on Earth? When making a message costs nothing, the costly signal is the only one left that works.
- What if a brand could only say one word, ever? Arecibo said its piece in 1,679 bits; could you do it in one word?
Questions people ask
Has anyone sent an advertisement into space?
Not one that reached another star system with a sales aim. Arecibo's 1974 message was a scientific demonstration of 1,679 bits aimed at M13, and the Pioneer plaques and Voyager records were physical greetings carried on spacecraft. None asked for anything in return.
What did the Arecibo message say?
It showed the numbers 1 to 10, the atomic numbers of five elements in DNA, the chemistry and shape of DNA, a human figure with height and population, the Solar System, and the Arecibo telescope. It was 73 rows of 23 bits.
How would aliens understand our message?
Designers start from things any technical civilisation must share: counting, prime numbers and physical constants such as the hydrogen atom. A message laid out as a product of two primes, like 23 × 73, tells the reader how to arrange it.
How long would a reply take?
Twice the distance in light-years. A reply from Proxima Centauri would take about 8.5 years; from M13, about 50,000 years. For most targets, nobody alive when a message is sent will see the answer.
Calculated in this post as twice the distance in light-years: 4.25 for Proxima Centauri (Wikipedia) and 25,000 for M13 (Cornell Chronicle, 1999).
Sources
- Cornell Chronicle: It's the 25th anniversary of Earth's first attempt to phone E.T. (1999) (source for the story, 2,380 MHz, 1,679 bits and M13)
- Wikipedia: Arecibo message (source for 450 kW, the 23 × 73 layout and the content list)
- Benford, Benford, Benford: Messaging with Cost Optimized Interstellar Beacons (2008) (source for the 1017 W EIRP and the equal-cost rule)
- Mahon: Local Compositional Complexity, Entropy (2025) (source for the structure scores)
- Ambler and Hollier: The Waste in Advertising Is the Part That Works, Journal of Advertising Research (2004) (source for the perceived-expense finding)
- Daepp et al.: The mortality of companies (2015) (source for the 5.8% yearly hazard)
- SETI Institute: Signs of technology in 3I/ATLAS (2026) (source for 74 million signals and the 10-110 W limit)
- Astrobiology.com: Ten years of UCLA SETI searches with the Green Bank Telescope (2026) (source for the 5 × 1016 W detection floor)
- Wikipedia: Proxima Centauri (source for the 4.25 light-year distance, from Gaia DR3)
