How I learned to love advertising

I always hated advertising. Perhaps that’s not quite true? I feel like I’m supposed to say that. I might have admired the verve of some ads, but I never thought of it as a good thing (imagine, some do!). I was a skater. I tried to feel punk rock. I was anti-yuppie. I did have my consumer side though: Starting in my teens and even into my thirties, I dealt with social anxiety in part by shopping. When I lived on Church Street in high school years I used to go to the record store a lot (OK), but also this particular shop that sold tee shirts in various colors, always checking for a new one (not so OK). I bought cameras, jeans, shoes, shirts, suits (lots of clothes), drinks, airplane tickets, gifts – overall, mostly clothes.

At some point I started to notice this buying impulse and channel it into non-buying. I’d walk myself through the shopping process, see how tiring and pointless it was, mentally spend the money – and be disgusted. I started to see the freedom in not spending the time and money, in not burdening myself with another thing.

In my late twenties, I took a job with a promising tech/media company. I had been working for a bigger, older software company doing consulting work, enjoying that quite a lot, when two things happened. I noticed the internet happening: I had a client, a bank in Boston as I recall, and they wanted their bank “on the internet”. This was 1996. Good for them for asking! The company I worked for said it wasn’t possible. I wrapped their big Unix server app in a perl script and had it print out HTML at the end, showed that to them and they were pretty impressed. I guess I got the picture too.

The other thing was that this job required me to get on a plane every week. There is a certain glamour in that and it wears off as soon as you come to your senses. We’ve all seen the movies about the business traveler trying to make the most of his/her situation, and it’s never that sexy. I started to realize that I felt rather homeless. Already my friends thought of me as that guy who was never around, just showed up for parties. Again, that can be fun, but it’s not a long-term strategy. Since I had already decided that I would never do a daily commute by car, and I was looking for an alternative to frequent airplane travel, I decided to look for a job that I could walk to from my apartment in North Beach.

My friends assured me this was folly. There were no software jobs in San Francisco proper. I think I knew that I didn’t really want to work for a software company. I interviewed with a few. I wasn’t really a programmer, I hadn’t studied computer science, I didn’t know any algorithms. I had seen or bought or read a few copies of WIRED magazine, and I heard that they had offices in the City. I sent someone (Joel Truher) and email. I wish I had that correspondence. We met a few times, often on the roof of the building (why?) and in the end they offered me a job writing code and managing the engineering team on the side.

I wrote a ton of perl that did nothing all that useful and tried to manage the bunch of burners that lingered in the dark with their LSD hangovers, writing code that that continually crashed Apache, Sybase, and whatever else. Everyone was up at odd hours all the time. Computers were still fairly slow; if something crashed while trying to grind through the day’s events, it would often be impossible to catch up – we had to throw away a lot of data. When I think about what got me that job, and what I was good at in software in general, I think it’s this: estimating. I’m good at keeping it in the ballpark. And this is super hard with software. Super hard, especially when you are making everything up.

Advertising just sort of crept into the picture. Someone showed me a script that Crispin wrote that stuffed ads into outbound HTML. Joel showed me some LISP (was it code?!?) that was supposed to decide what ad to shove in. He was ahead of his time. Then we decided to build a search engine: HotBot. This was before Google. We had a deal with Inktomi to provide the back end and the search function itself, so we had to built the front end – and the ad server. By then we had already served “the first ad on the web”, we must have been making some money that way, and ads were going to be a big part of the search engine from the start. We had to figure out how to serve many many millions of keyword-targeted ads per second, and then to produce accurate statistics and reports about what ads were served alongside which search terms and why.

This ad-keyword project lit up my nerd brain. I dove into the spaghetti of (thankfully: perl) code that was the back end of the ad server and made many enhancements. Using some of what I had learned on earlier systems, I managed to turn the all-or-nothing nightly process of consuming the web- and ad-server logs into a continuous parallel process that could be farmed out to as many processors and disks as we cared to throw at it. I also figured out how to make a stew of the the guts of the ad server so that it was able to hold thousands of possible keyword-targeted ads in its little mind while responding to queries, without crashing. All of this fascinated me, at least briefly. Jobs are like that sometimes.

I’m not sure if even then I was aware that I was working more and more on “advertising”. Strange that it didn’t dawn on me. I think it came down to this: the work was engaging, and I was excited to be doing something fairly unique.

AdMonsters was essentially an outgrowth of that HotBot keyword targeting project. I remember discussing it at our first meeting in June of 1999. I was proud of the work I had done – and the uniqueness was proven by the fact that our first group was so small. There were fewer than twenty people working on that crap. And it was crap. But I still didn’t really see that. It just felt good to be doing something engaging and unique.

See how hard it is to see the world from the little road you’re on? I was happy to have a good job, happy to getting paid well, happy to able to do my own thing, again, to have a unique job, feeding my ego basically. I wasn’t paying that much attention.

There were signs. I was offered a job, an up-the-ladder management position, a mega opportunity in a lot of ways, a job that would have made me a lot of money. I turned it down. I didn’t quite know why. I didn’t listen to myself hard enough. I convinced myself that I didn’t want to leave San Francisco, that I didn’t want to travel so much. As close as I got was that I didn’t want the life that I imagined having, if I had that job. That I didn’t want to be that person. This happened twice actually.

I loved putting together conferences. There’s a lot of stress, and it is repeated. That’s not the part I loved, but I didn’t hate it, or I responded well to it. And I loved the gratification of seeing it all come together, so many “moving parts” – people – all timed perfectly to create an experience over just a few days. Clearly, this engaged and enlivened another part of me. And people loved the conferences from the start. We (well, I, to begin with, because it was just me for several years) did a really good job.

The conferences were about advertising. More specifically, it was about the software that web-based publishers (companies that make stuff you read on the web) use to put ads in front of your eyeballs. That’s what a publisher does: you create something that people want to read, you either sell it to the reader directly or you sell ads to companies that want to advertise to your readers (usually both). This doesn’t have to be complicated – you can just charge people $whatever to read writing and have no ads at all, that works fine if there are enough people that will pay. But: in the early days of the web, there simply wasn’t any way to charge readers anything. And so once the media-business types got the idea that software could be used in new and nerd-fascinating ways to (theoretically) show just the right ad to the right person at the right time, a landslide of investment began to pour into ad-serving software. And since, again, there still wasn’t any way to charge readers anything, all of the time and money that would have been invested at a non-web media company in getting readers to buy or subscribe was also invested into trying to generate revenue from advertising.

This is why the “online advertising business” exploded along with online media. Media business types saw media moving online, there wasn’t any way to charge readers to read, therefore advertising was the only possible revenue stream and the magic of software was so alluring that it seemed that it would be soon possible to make advertising perfect. That is: don’t show ads to the wrong people (so: no wasted ad impressions) and only show you ads that you, er, want to see (so: in theory, no annoying ads).

This massive, multi-billion dollar industry grew up around the premise that you could only show ads for men’s running shoes to men who run, and only to men who run who want to buy running shoes soon. Does that sound like bullshit? It should.

I mean, you can try. Anything is possible with software (it’s just a question of how much it costs). You can “target” men, men of a certain age who might run, men who visit running web sites. No problem, although it turns out that even this is much more complicated than those early nerds admitted. Can you figure out if he is intending to buy another pair soon? Can you? Who fucking cares? With a billion dollars of investment funding, you can certainly try. And they did.

Online advertising has been chasing this dream from day one. In the meantime, 1) Google, and 2) it became possible to charge readers to read, so just as we have been paying to read newspapers, books and magazines, we can now pay to read similar things online, and 3) it didn’t really work that well. Basically we spent something like $20B to figure out that the cost of almost-perfectly targeted advertising is more or less the same as not doing it at all. If it costs you $100 in advertising to sell $110 more of your product, it’s not worth advertising. Even if you sell $200 more, it’s not really worth it – you’re just breaking even.

Perhaps it’s interesting that I never used advertising to sell my conference. I didn’t need to. For the most part, people who were interested heard about it, or if they hadn’t heard about it, once they did they were immediately interested. Those who needed it most sought it out and found it – and then told others. That is the essence of a useful product.

At some point I started hating my job – my company – because it was all about advertising. I hated telling people what I did for a living: I produce conferences (bad enough!) – oh, about what? – they’re about online advertising. Dead air. I loved having my own business, I enjoyed the process, the doing, the producing – and I also really loved the positive feedback that I continued to get from my customers – the people that came to the conferences who told me how much they loved it, how it changed them, helped them in their careers. But the bad taste in my mouth for advertising got worse and worse.

Was I doing something morally wrong? Certainly not. I wasn’t even doing any advertising. I was making nerds happy!

My long-standing general distaste for advertising was crystallizing into a more concrete understanding that not only was advertising just sort of dumb, it was unnecessary – really, advertising is obsolete. This started to become clear to me. I wasn’t foresighted enough to conclude: now I must get out, but it did increase my urgency.

In parallel with this I finally developed some new professional relationships, some friends and peers who understood what I was doing from a purely business point of view, understood the conflicts and rewards and encouraged me to set aside my distaste long enough to work hard enough to make a real go of it. I had been distancing myself from the business because I didn’t want to be associated with advertising, and the business was suffering – not quite badly enough to fail, but enough to make the business more of a pain in the ass than it needed to be (and unattractive to potential buyers).

When I finally came to terms with the fact that this was my story, that I knew that I didn’t want to be that guy who took one of those up-the-ladder jobs, I knew that I wanted my own business, I knew that I didn’t like advertising, and I knew that I was neglecting my own business – then I knew what I had to do. Buckle down, set aside my distaste for advertising, run the business well, and find a way to get out clean.

Now that I’ve had time to read a few books, I see that I fucking won! Hugh MacLeod calls this the “Sex & Cash Theory” – a creative person should have two jobs, one that pays the bills, and the other that is the sexy, creative one. “Sometimes the task at hand covers both bases, but not often. This tense duality will always play center stage. It will never be transcended.” In his outstanding book Antifragile, Nassim Nicholas Taleb shows how “a combination of extremes kept separate”, a bi-model strategy of “playing it safe in some areas and taking a lot of risks in others” is central to anti-fragility. This can happen in parallel, or serially (as it appears to be happening for me).

I bought my freedom, and in fact not by paying someone else – I paid off my own house, and now I can live in it. It’s strange that the house was built on advertising, but that’s how it happened. So, as much as I hate advertising, and advertising is obsolete – I love advertising.

Advertising is Obsolete

(This post also appears on here on Medium)

I’ve been saying this for quite a while, and now it’s become painfully obvious: advertising is obsolete.

Advertising is no longer needed for discovery

Advertising originated as a way to spread the word to potential buyers about things for sale. Without a sign with an arrow pointing to that drugstore around the corner, some people with a legitimate need for something sold at that store would not know of its existence. This is economically additive and a good thing.

This evolved into mass-market advertising a way to try to sell more stuff, mostly stuff that wasn’t needed, but might be wanted. This sort of advertising is a big part of the fuel of the consumer economy; it increases GDP but creates nothing, and is not economically additive.

Since the invention of comprehensively useful search engines, we can easily find whatever we might actually need, wherever it happens to be. Advertising is no longer needed to find things that we need. The fact that most search engines happen to be themselves funded by advertising is a historic artifact; there’s no reason that they must be funded by advertising simply because they currently are.

Media does not require advertising

The widely-accepted idea that advertising is an intrinsic part of the media business model is false. Just because advertising is part of most media businesses doesn’t mean that it must be. As Ev Williams outlines in his piece on The Rationalization of Publishing, other forms of media (TV, music, books) do not depend on advertising, and yet we’ve continued to operate under the assumption that for some reason news is different. He sets up three strawman arguments as to why news might be — and isn’t — different. I mostly agree with those points, but I believe that there are two additional (and equally misguided) reasons that we’ve believed that news is somehow different from other types of media.

Software was supposed to be magic

Big ideas have a long legacy. Riding along with the early meme that “information wants to be free” was the technocratic fantasy that software would make online advertising work so well that it would again become an economic benefit — a way for people to discover things that they need (or really want), and that they would have otherwise not been aware of. While this does happen to some small extent, we have treated the personal data that is required to make this possible as a zero-cost externality. If we factor in the real value (any value, really) of this data, the net economic value of the discovery of a product by way of highly targeted internet advertising becomes negative. Many companies and countries are now taking steps that take this negative impact into account, making such advertising much more difficult, impossible, or illegal.

In addition to the hidden cost to individuals of sharing all this data, the cost of processing all of this data was vastly underestimated. For most online publishers, the complexity and cost of building and operating software to ‘serve’ advertising to their readers is more than the return on that investment. Online advertising never has been that profitable for smaller publishers, and it will prove not to be profitable even for the largest (Facebook, Google).

It’s true that in those early days of the ‘net it didn’t seem possible to simply charge people for what they wanted to read. Serving ads seemed like an easy alternative, and once we started down that road, we kept at it, even though it turned out to be much harder than we originally thought. And because of our love of software and of problem-solving and our stubbornness and our reluctance to abandon sunk costs, we kept trying to make online advertising work long after we should have simply reallocated all that effort to finding effective ways to charge readers directly.

As we now know, it’s not at all impossible to charge readers directly, but because we had put so much effort into trying to make online advertising work as the sole form of economic support for online journalism, we back-formed the idea that online advertising is the “only rational model” that can support online journalism. Bullshit. Advertising seemed like an easy way to support online journalism, it wasn’t, it’s not, and it’s time to move on.

News was supposed to be free

The “free press” is a nice idea but somehow along the way we confused free speech with free, as in beer. Again, because of the roots of the internet we became enamored with cheap, or even free news. Ev Williams said it well in the same piece I referenced above:

There is — and probably always will be — a surplus of free content. But that’s like saying there’s a surplus of free food in the dumpster behind the alley.

We’ve been dumpster-diving for news for long enough, and, needless to say, the effects are showing. Don’t think that we’ve seen the bottom either — it’s likely to get worse before it gets better.

Advertising adds no net value to the world

We need to discard the misguided assumption that news is somehow better if it’s free. As with any other product, we will pay for information that we actually need, want, and use. Advertising came to be part of the media business model as a way to support the creation of media not that readers wanted, but that advertisers wanted. Real companies make products that people pay for. News is no different.

Anything one needs to market heavily is necessarily either an inferior product or an evil one. (Nassim Nicholas Taleb, Antifragile)

All of what is spent on advertising could be used to make better things — or not spent at all. Nobody would shed a tear if we turned advertising off, and nor would it harm the economy. I include advertising in my list of things that we’ll look back on as 20th-century ideas whose time came… and went.

Advertising is obsolete.

(This post also appears on here on Medium)