How the coupon was invented, and how tech propels marketing

Why AI, VR, web3, and future tech will beget new marketing strategies

Technology propels marketing and vice versa
OK I think there should be a new law:

Any widely adopted technology will ultimately be used by marketers to sell stuff

In other words, as AI, VR/AR, web3, and other technologies guide us towards the future, part of that future will inevitably be to promote other products/services. (For better or worse!) We’ll see new variations of loyalty incentives, user acquisition programs, cross-sell mechanisms, and other endless tactics, but facilitated by LLMs and the AI companions, or in 3D environments, or on the blockchain. And these marketing strategies and tactics will be unique, native to its underlying technology, and when they work they will be important to culture and society.

The coupon was created at such an intersection, but instead of AI/web3/VR, it was created at the intersection of the steam-powered printing press, mass newspapers, and innovations in the US Postal Service. And now coupons and their digital-first cousin “promo codes” — play a deep role in commerce and culture.

Why the coupon was invented
You might know that Coca-Cola is credited with creating the first coupon in 1887. I know this because I’m a bit of a coke fanboy, if there is such a thing. But I do drink 2-3 diet cokes every day, with every meal. But perhaps you already know this?

Maybe you read this bit of trivia during a late night Wikipedia binge, and you stashed it in a “fun facts” bucket in your brain. Or maybe you’re reading this for the first time. Either way, great - and I have more Coca-Cola fun facts for you!

• The creator of the recipe is actually different than the founder of the company. John Pemberton, a pharmacist and former Confederate soldier, invented the original formula after sustaining a knife wound and becoming addicted to morphine. Pressured by near bankruptcy and addiction, he eventually sold the recipe to businessman Asa Candler for $2300, roughly $77k of today’s dollars, and it was Candler who eventually started The Coca Cola Company

• Side note: Yes, the original recipe had coca leaves / cocaine. But funny enough, it was not unusual at the time. Brands like Mrs. Winslow's Soothing Syrup, marketed for teething children, contained morphine. Bayer Pharmaceuticals manufactured and sold heroin at scale, as a pain reliever and cough suppressant. Thus, less weird than it sounds

• Yes, Asa Candler invented the coupon, but the real story is that the tactic helped solve the cold start problem. It wasn’t just stimulating consumer demand — it’s more complicated than that. Coke was always B2B2C — it sold syrup to pharmacies, general stores, and so-called soda fountains. (Note that this was in the years before it was sold in a bottle/can, of course) Soda fountains were an important type of partner, as they were sort of a late-1800s sparkling water bar that was a hang out spot, serving various syrups in with mineral waters. Coca Cola competed with dozens of other alternatives. Candler would initially give free syrup to these retailers, then distribute coupons at the same time to everyone in the area — this was a way to stimulate both supply and demand simultaneously. (A nice solve to the cold start problem)

• The Coca Cola Company paid for the free drink, not the retailer. The retailers would collect the coupons, send them in, and the company would reimburse them. In later years, entire companies (called coupon clearinghouses) would be formed just to collect, process, validate, and pay for all of these coupons. The company had to handle all of this itself in the early days, having invented the marketing tactic. In the first 25 years of the coupon, Coca Cola is estimated to have redeemed them for free drinks 8.5 million times, when the US population numbered ~60-90 million people. Incredible scale for that time.

I could go on and on here, there’s so much interesting history as this was the turn of the century when a lot of these marketing strategies were being invented. In fact, I think of tech and marketing as intimately intertwined — let me name a few examples to convince you.

Tech begets marketing
Coupons were built on the major tech innovations during the time of the mid/late 1800s. This included the widespread availability of the steam-powered printing press which begat newspapers, plus chromolithography for multi-colored printing, and nation-wide direct mail. You can draw a direct line from each of these technologies to the result of Candler’s invention.

Technologies build on each other, and marketing opportunities arise. Just take direct mail as an example, which was an essential ingredient in all of this:

• Widespread use of trains → expansion of railroads across the US → cheaper transportation costs → cheaper mail costs → everyone gets free home delivery → marketing via direct mail

You can make the same case within more modern technologies — the advent of the TV commercial, soap operas, billboards, and many others all come from underlying tech:

• Invention of the TV → widespread creation of TV shows → TV commercials, and “soap operas” to target moms in the household

• Invention of the car → significant construction of roads → Interstate highways → billboard signage and OOH (plus drive-throughs, Los Angeles, Walmart, etc)

• And let’s do a modern one: the personal computer → smartphone → massive proliferation of front camera video content → influencer marketing

• … and many other obvious examples (email, sponsorships, etc) that I leave to the reader to figure out!

But let’s get more detailed — to dive into Coca Cola’s coupons specifically, you’ll note that they are commonly distributed within newspapers and magazines — sometimes to be cut out and sometimes as inserts. And it was the steam-powered printing press the industrial paper making machine (both invented/available in the 1880s), that enabled the mass market newspaper. These so-called “penny press newspapers” were 1 cent newspapers that were much cheaper to buy than the premium newspapers they eventually drove out, and their revenue primarily derived from ads not subscription. Coupons could be included in these newspapers, to be cut out and used as needed.

And of course, the printing press was also needed to actually print millions of coupons as well, which is an even more direct use of the technology. This was particularly true of advertisements and coupons, which used color to maximize the effects of its marketing.

The coupons then needed to be delivered to consumers’ homes. Sometimes this was through the newspaper, as mentioned before, but Candler/Coke also made use of the US Postal Service. For the first decades of the USPS, mail was usually delivered from Post Office to Post Office — they didn’t deliver to your home. But in 1863, the Postal Service started a program called “Free City Delivery” that sent mail directly to peoples’ homes. I can only imagine this was like giving everyone cell phones where they could directly text/call from, rather than going to shared telephone booths. And in 1890, free rural delivery was implemented widely across the US as well.

The coupon was built on all of these amazing innovations — and the next generation of marketing will come from the technologies we’re working on today.

Coupons and promo codes in the modern age
The use of selective price discounting lives on today, even as the practice of clipping coupons fades away. Most software products seem to have embraced “promo codes” as the spiritual successor to coupons. After all, you can attach them to an email, or a landing page, or a coupon. You can attach a referral program to them, such that every user has a personalized code that benefits them as well as whomever they refer. They can appear on an ad, or you can have it shouted out by your favorite content creator.

Yet some of the same problems that plagued Coca Cola and consumer goods companies still remain:

• Do promos actually help you attract new customers? Or are you just “pulling forward” demand from existing ones — thus incurring additional expenses for purchases you would have gotten anyway?

• How often do promos lead to fraud? A renegade soda fountain could have counterfeited Coca Cola’s coupons, and a modern bad actor could add promo codes to a bunch of transactions, get reimbursed, then charge back the sales. How do you avoid that?

• Who are the right people to target with coupons to drive the desired effect? If you send coupons to low-intent customers, will they be a one-hit wonder, purchasing once and then never using the product again? Will it screw up the CAC/LTV calculations you’ve meticulously created over the past year?

• Have you created loopholes such that customers can “stack” discounts, using a promo code on top of a holiday sale on top of other incentives? Are you driving top line revenue that’s highly unprofitable and undesirable?

If these problems sound familiar to you, it’s because they are the modern dilemmas of marketing teams across the tech industry. Many of you know that before my career in venture capital, I worked at Uber in the hypergrowth years, and we spent upwards of a billion dollars a year on various forms of rider and drive incentives to try to alter their behavior. The above questions are exactly the ones that teams of hundreds of people worked on, given the enormous spend.

After all, it’s easy to launch a new discounting campaign, get a big bump in revenue, but then realize the revenue is low-quality. And you’re monetizing customers at a discount when you could be getting them at full price if you just had patience. These dynamics are all tricky. But the good news is that coupons can be targeted. They can be split tested. And so you can create various targeting systems and try to measure “Cost Per Incremental Trip” by A/B testing different types of discounts. The key is to measure the cost of coupons and try to compare it to a base rate, to figure out if it’s really worth it. (I say a lot more about this in my book, but won’t belabor the point here)

Even though coupons have many problems, I think we’ll see some form of them evolve and survive in the coming years.

Marketing in the age of AI
I recently wrote How AI will reinvent Marketing as a brainstorm of all the ways we might see the marketing discipline change in the coming decade. We are just now starting to accept a world of companions and AI-driven UX across our day-to-day life. But I argue that these agents/companions/advisors/friends/lovers are really just the surface area where new marketing techniques are bound to intersect. And they will intersect in many ways, not just to help new products acquire customers, but also to reward them for loyal usage. And to cross-sell additional products.

Just as coupons are one marketing tactic that sits on the “surface area” of newspapers and mail, I think we are bound to see the surface area of “conversations with my companion friends” become host to new marketing interactions as well. This might seem weird right now, but I think we will simply have various flavors of companion friends, some who help us with specific tasks (and who are obviously going to recommend products— as ads in a podcast commercial might), and friends who are highly engaging and mostly non-commercial in nature. Sometimes I talk to a gym trainer companion who I know will plug the occasional nutritional products, and other times I talk to my funny gamer companion who tells me about the best new games. Is this any more weird than watching my favorite YouTube creator talk about a product they love (and that they are being paid to talk about)? These are all parasocial relationships, but the future AI ones will simply be interactive, more informative, and more engaging.

And the marketing techniques will evolve and grow over time. To go back to coupons for a second, following Coca Cola’s example, the entire industry of consumer goods companies embraced coupons. And coupons themselves evolved. Sometimes they were highly targeted to certain geographies and sometimes combined with holiday campaigns. But also coupons weren’t just used to acquire new customers — they also became loyalty programs. Famously, General Mills printed a rewards program on top of their boxes of cereal, pasta, and their multitude of others products. These coupons had points printed on them, which could be redeemed for products in the Betty Crocker catalog they’d send you. Thus you’d always be buying and saving points as part of their program, creating a sort of proto-loyalty program.

If interactive companions are the inevitable surface area for marketing in the future, I think we’ll see the same thing:

• Ads will appear, whether they are marked or not, in AI-created conversations. It’s easy to think they will be a sidebar of ads as Google does them today, but somehow I think it’ll be more subtle, and more native

• If an analog to SEM exists within LLMs, I think we’ll also see a form of SEO. How will companies and products create content that is most attractive to becoming incorporated into the model data. You could imagine in a world of Open Source LLMs that companies will try to reverse engineer how to get themselves into the answer when people ask, “what are the best hotels in XYZ?”

• Companions could facilitate loyalty programs. If they keep long-running state and have large context windows, they can cross-sell products based on an understanding of your likes/dislikes, just as a travel concierge might keep you on the same chain of luxury hotels based on price and otherwise. If they are able to share some of your data with other AIs, perhaps they can do a real-time negotiation for you that keeps you sticky as a customer

• We might also see AI companions that are directly and obviously sponsored by companies. Today, we have horizontal products like Google that help us research, but we also have Booking and Yelp and other vertical search products for very distinct questions. Perhaps the same will exist of AI

• Once we are fully multi-modal, we might see video ads or interactive experiences becomes answers on commercial queries. Just as Coke’s ads in the early 1900s pushed the boundaries of color printing, if consumers respond best to video, we’ll see more commercial video content — perhaps generated on the fly, in response to their needs and prior preferences

• The most common business models for LLMs today are subscription. I have to say, I doubt that will last. People want free. And they have demonstrated again and again they prefer free plus targeted ads rather than anything else

Needless to say, as long as products/services cost something, and they have prices, then they will be able to be discounted. Whether it’s a piece of paper doing the discounting, or a 500 IQ fully-interactive AGI, who knows? Does it matter?

And thus the story of the coupon continues.