Media Spend: Traditional vs Digital

At least once a week I get asked a question by a client like this one I recently received, from one of the world's largest CPG companies:

How do we spend our Digital Marketing budget more rationally?

Often the question is more focused on the transformation underway as companies move from the "traditional" spending categories (print, broadcast, outdoor...) to digital categories (web, mobile, social), like this one from a computer hardware manufacturer:

What percentage of our marketing budget should be spent on digital?

Full disclosure: My qualifications in marketing (other than having worked in marketing for over two decades) are a bit suspect. While my first marketing job was in a direct mail marketing company in the 1980s, the vast majority of my direct expertise has been in leading marketing organizations for digital brands -- an Internet company, a software company, a mobile operating system...

So if you are a traditional marketer, selling soap (or whatever physical good), you might look at my advice and say something like "sure, that would work if my product was entirely online... but my products are made up of atoms, not bits..."

And I sympathize with you, I really do. The past 100 years have taught marketers to believe that there are two measurable things about advertising: reach and recall. How many people did you reach, and do they recall your brand. After all, the important question for marketers has been whether the consumer will reach for your bar of soap or someone else's when standing in the grocery store aisle.

And the medium that we have had available to us (largely broadcast in dollars spent) has lent itself to a logic about advertising that tells the marketer to focus on repetition of a simple and short brand message. Repetition because seeing something over and over again helps the human brain with that recall problem. Simple and short both because our attention span for advertising is short and also because the cost for each advertisement is high (and higher for longer ads).

But two things are happening that should cause every marketer to pause and re-examine everything they believe about marketing and begin asking a different set of questions. First, our purchase patterns are changing -- we are no longer buying something because the packaging stands out from a store shelf. In fact, we are going into physical retail less and less and this trend will continue to accelerate for most product categories. Second, people are moving from a diet high in traditional media to one high in digital media which is changing how we engage with brands and what we expect from them.

A starting point would be to reconsider what Marshall McLuhan had to say about hot and cool media. McLuhan argued that "hot" media were those that provided little stimulus and thus required an engaged and participatory consumption. Whereas "cool" media were those that offered substantial stimulus and where consumption required very little involvement. From his vantage point at the beginning of the development of 20th century media, McLuhan might have assigned "cool" or "hot" differently than we would today, but the basic model of differentiating levels of engagement is applicable as we think about marketing through traditional or digital means.

Traditional marketing fits McLuhan's "cool" media categorization, as epitomized by the goal of repetitive simple and short brand messages. The expectation by the marketer is that the recipient of such advertising will not expend effort in engaging, but must receive the message over and over again in order to get it to stick.

Digital marketing is more complex because many of the techniques of traditional marketing were seemingly transplated into online spaces but the medium itself is "hot" -- we are engaged with the computer or mobile device, clicking and directing our experience and not just passively absorbing whatever might come next. Even with online advertising (the seemingly transpanted traditional marketing approach) success is measured by the click-through -- did the viewer engage!

For digital the marketing equation has been turned upside down. Instead of repeating short and simple brand messages and measuring reach and recall, marketers should be building vehicles that engage people more deeply and then measuring the degree of engagement achieved. And in digital we have an entirely new capability through this engagement: building our knowledge of people (both individually and as groups) through their interactions with our brands. Finally, digital can allow us to connect what we know about a person throughout their entire experience with our brand -- consideration, selection, transaction, receipt, consumption, satisfaction, return... recognizing when and where we have converted a prospect to a sale, or a purchaser into a loyal fan.

Given my self-admitted bias toward digital marketing, it would seem simple to say that companies should be prepared to move ALL of their media spend to digital over the next decade. But in working with hundreds of companies over the last 20 years as digital has continued to mature, I have come to a different conclusion:

Companies should be prepared to move a substantial amount of their total media spend into digital and away from traditional media. This will require a new set of competencies and even a new organizational structure in most marketing organizations. It is crucial to get to a sufficient amount of digital activities to develop true data-driven insights. Without achieving critical mass, evidence for how digital is impacting sales and customer satisfaction will remain anecdotal. For the largest advertisers I have worked with, the threshold was at about 25% of total spend. Smaller companies will likely need to spend a larger proportion on digital.

Digital media should not be limited to "advertising" but should include all experiential engagement with a customer - websites, mobile applications, social media, even customer service interactions -- anything where you can impact customer experience, measure engagement, and increase your knowledge of your prospects and customers. If you aren't already familiar with the concept of "earned, owned, and paid" media, make it a point to read up on these ideas (click for a good article from Forrester to get you started).

As you move to digital and start generating data-driven insights, the transition of spending from traditional to digital will accelerate. I believe that most companies will stabilize at more than 2/3 of their total budget on digital. But traditional marketing will develop to have a new role - the reinforcement of digital marketing activities. One way is to use traditional ads as traffic drivers to digital destinations. Another is to build positive reinforcement cycles for themes that can appear in both digital and traditional mediums. In any case, traditional will often provide the most value when it drives more digital engagement.

In 2015 the thing I am most surprised at is that we are still having this conversation -- that companies are still blindly spending their marketing dollars in the same way they were spent in the last century. Each company certainly still needs to answer for itself, based on the specific industry, buyers, and products, the kinds of questions I mentioned at the start of this article -- how do we spend our digital marketing budget more rationally, and what is the right percentage of spend to move to digital. But answers to those questions won't lead to greater success until we embrace the new role of marketing, that the marketer is now responsible for engagement and customer knowledge -- not reach and recall.

Tech Bubbles and Escape Velocity

In 1997 I was one of many exuberant entrepreneurs working for a company in this new domain, "the Internet." In telling our stories to investors we all predicted future valuation based on the belief that someday everyone would have high speed connections to the Internet. Its hard to recall today that in December of 1997 only 70 million people had such connections.

By the time we had sold our company in August of 1998 many of these early Internet pioneers, including me, had started to say that we were in a period of gross overvaluation. And when Stephan Paternot and Todd Krizelman took their company, TheGlobe.com, public in November of 1998 and reached a valuation of almost $1 Billion the first day, our fears of a "tech bubble" were confirmed.

But of course, the party continued for quite awhile after that.

An interesting conversation circulates today amongst those of us that knew each other during the so-called dot-com bubble as we watch the current crop of "unicorn" and "decacorn" companies (and the worsening traffic and accelerating rise in the cost of living in the Bay Area). It feels like we've all seen this before and know how it ends -- but we can't decide, is this 1997 or 1998 or 1999? In other words, how much longer will this particular tech bubble last?

But I've begun to think that we are asking the wrong question. Looking back on the crazy predictions that we made that someday we'd go from a mere 70 million people connected to the Internet to "everyone" connected -- we were right. And we were probably even right about the timing. And this is where tech bubbles are fundamentally different from other kinds of speculative asset bubbles -- there is an underlying economic dynamic that actually is in a state of transformation.

Take a step back and examine what economists mean when they describe something as a bubble. The most basic formulation is when assets are valued in a way that becomes disconnected with the intrinsic value of those assets. One way this happens is when supply and demand fall out of balance for what would otherwise be a stable commodity. Everyone wanting a particular toy as a Christmas present might cause speculators to buy up those presents and put them on eBay at a higher price, creating a kind of toy bubble which would ultimately be resolved either by the company increasing production or another toy coming into favor which would decrease demand.

But tech bubbles are different because we are looking at companies extracting value from a transformation of markets rather than from an unbalanced market. When we made our Excel charts of 70 million Internet users growing to billions of Internet users to convince investors of our future value, we were describing a transformation in the way everything would function on our planet. A transformation that has come to pass and is continuing to accelerate today.

The current set of maturing technologies founded on the infrastructure of the Internet -- social, mobile, analytics, and cloud -- are well on their way to reformulate the way our markets operate. How we select, purchase, receive, and consume products and services is being transformed.

Thus the target opportunity for the tech industry is virtually the entire $18 trillion US economy of which technology only represents 7.1% today. And the $80 trillion global economy of which technology is an even smaller component. Shifting cash flows (and thus valuations) by only a few percentage points would move trillions of dollars out of the old economy and into the hands of technology entrepreneurs and investors, thereby justifying the valuations we are seeing today. Why should Uber ($60 Billion) be worth more than General Motors ($50 Billion)? Because Uber has a plan to shift cash flows from old industries to themselves while GM is still making and selling cars (fewer and fewer as well - down to 3 million cars last year from over 5 million back in the dot-com bubble).

So what actually happened during the dot-com bubble? Why did it burst? Will the current tech boom also turn out to have been a bubble and will it burst as well?

Certainly one element of the valuation of assets outpacing their intrinsic value came through the uninteded consequences of tax policy changes (here is a good and brief overview). But capital could have flowed into many different financial vehicles. The reason that Internet companies attracted the most attention was that the underlying trends really did appear to be correct (and in hindsight they were correct). The Internet really was going to shift old economy cash flows into the hands of technology entrepreneurs and investors. But investor enthusiasm outpaced the speed at which this shift could happen. As a result, too much supply was created too fast -- too many companies selling pet food on the Internet. Business models and valuations diverged too far from intrinsic value and investors blinked.

Which brings me to the question that I think we should be asking right now about the current tech boom. The question is NOT "when will the bubble burst" but instead something a bit more subtle. The question is -- this time around can technology transform markets at a fast enough pace to keep up with growing valuations. I think of this as an escape velocity problem -- gravity in this case being investor expectations. Can the current crop of tech boom companies grow fast enough, deliver results fast enough, to outpace the come back to earth force of investor expectations?

Every old economy company on the planet should be frightened by now of the logic behind this process -- shift cash flow away from every other industry to the tech industry. Technology companies are serving as intermediary entities in which the retail (Amazon), entertainment (Apple or Netflix), or transportation (Uber) revenues go first to the tech company and, only after margin has been removed, back to the old economy creator or owner of those assets (often bypassing prior intermediary models). That isn't what technology used to be about. Tech companies used to create new products (personal computers) that created new markets. But increasingly the new products that tech companies create are disrupting old markets instead.

And this time around the tech industry is moving faster and has many more tools at its disposal to capture the $80 trillion global GDP. It may be the case that tech again fails to achieve escape velocity and that this boom also comes to an end. But over time, tech will win. In the future as the less mature technologies of accretive manufacturing, robotics, and artificial intelligence further accelerate market transformation, the tech industry will have an ever increasing array of tools to capture a larger and larger portion of global GDP.

Eventually we won't even talk about a "tech industry" because ALL companies will be tech companies -- that is, all companies will be dependent upon continuous technology innovation to achieve and maintain competitive advantage in their markets.

So next time someone asks you when you think the current tech bubble will burst, ask them instead when do they think tech companies will slow down in transferring wealth from other industries to the tech industry.

My answer: not until they have all of it.