Marketers are Spending Too Much in Digital and Hyper Targeting is NOT Brand-Building

According to eMarketer data, about 60% of total media ad spending in the U.S. is allocated to digital already. Within digital, nearly 90% of the display ads are bought and sold through programmatic channels. The display, video, and CTV (connected TV) ads in digital are equivalent to billboards, print ads, and TV ads in the offline world -- they are used to generate awareness. Most of the biggest spenders in digital -- CPG, telecom, financial services, automotive, consumer electronics, pharmaceutical, media and online services -- will say they are doing “brand building” with such ads, as opposed to “performance marketing” since they are not trying to drive online sales directly. 

So let’s pressure test this together -- whether advertisers are actually doing brand building in digital and whether the brand building is actually working.

When big brands stopped spending in digital, nothing happened. Why?

Let’s start by looking at what’s not working in digital. There are a few publicly documented cases where the digital ad spending did not accomplish what it was intended to do -- drive more business outcomes and activity. For example, when P&G cut $200 million [1] from their digital ad spending they saw no change in business outcomes (they actually saw an 8% bump in sales in the same time period). Chase cut the number of programmatic websites [2] showing their ads from 400,000 to 5,000, a 99% decrease, and saw no change to business activity -- like new card applications and cards issued. When eBay turned off all paid search spending [3] in the western U.S. they saw no decrease in the number of website visits or the rate of ecommerce sales. Uber turned off 4/5 of their paid mobile app install spending [4], but the app installs kept happening. 

The last case from Uber was most directly related to fraud -- where the mobile exchanges were claiming credit for app installs that had already happened, falsifying reports to hide the fact that ads ran on porn sites; and in some cases fabricating transparency reports to make it appear that ads ran, when no ads were ever placed. In essence the advertiser paid millions for excel spreadsheets that purported to show how many ad impressions they bought, how many clicks they got, and how many app installs -- when none of that actually happened. 

Digital offers big numbers and small prices. But how?

If these cases of digital ad spending yielding nothing keep piling up, why are so many marketers continuing to spend heavily in digital and even raising their allocations to digital? It’s sadly because they are physically, mentally, and emotionally addicted to “digital cocaine” -- super large quantities at super low prices. In what marketing channel can you buy trillions upon trillions of ad impressions? Literally no other channel except digital. But how do long-tail niche websites that no one has ever heard of generate 1,000 times the quantity of ad impressions that mainstream sites generate? Certainly not with humans. They use botnets instead -- i.e. they “buy traffic.”

The vast botnets that were used for DDoS (distributed denial of service) attacks that overwhelmed even the largest sites with traffic, can now be pointed at websites that pay them for such traffic; the traffic firehose helps those sites to make money through ad fraud, lots of money. This is a highly profitable, direct way of monetizing the botnets. When marketers started pouring ad budgets into programmatic channels, in earnest after 2013, the sheer quantity of sites and ads dissociated from reality further and further (red area widening in the chart above). In other words, the number of ad impressions (blue line) could not be justified by more humans using the internet, social media, and mobile more and more (yellow and green lines). Once we dissociated ourselves from the physical constraints of the physical world (e.g. finite number of pages for ads in magazines, finite number of minutes for ads on TV shows, etc.), we entered the digital world where unlimited numbers of websites could magically conjure unlimited numbers of ad impressions. All those ad impressions absorbing all that ad spending had to be generated by something -- hint, bots, because there aren’t enough humans on earth. And that is ad fraud -- ads shown to bots on fake sites, instead of to humans on real sites. 

In addition to large quantities, digital media offered seemingly low prices too. While the following is not a valid comparison, some marketers use CPMs (cost per thousand units) to justify shifting dollars from traditional channels to digital. After all, they claim, if direct mail costs 75 cents per mail piece, that translates into a $750 CPM (for a thousand mail pieces). Broadcast TV CPMs are between $20 - $30; but digital display ads can be purchased for $1 - $3 CPMs (see chart above). This is related to the fraud problem mentioned above. The vast quantities of ad impressions are created by bot traffic on fake websites. That’s how those sites can sell ads for so cheap; they are fake sites that have low costs, unlike real publishers that have real journalists and editors writing real content for real human audiences. So large quantities and cheap prices are all intertwined with ad fraud and bots. Many marketers rushed headlong into digital chasing these large quantities and low prices, thinking they can do more brand building in digital because they could buy lots more ad impressions at much lower prices. But showing large quantities of low cost ads to bots is not brand building. You’d have to show an ad to a human, for that to happen actually. 

Marketers remain addicted to digital. But why?

If marketers already knew about the cases where digital ad spending delivered no additional business activity, and they already knew about ad fraud, why do so many of them continue to spend in digital? It’s because they think they are getting engagement, when they see clicks on their ads. But alas they again missed a key consideration, the possibility of fraud -- clicks by the bots. Bots click on ads to give the appearance of engagement; this is exactly what advertisers were looking for, so they kept pouring more money into digital. In fact, by tuning the click through rates higher on fake sites than the CTRs on real websites, the scammers trick advertisers into allocating even more budget to programmatic channels because it appears that they are getting more engagement. Of course, some humans click on display ads; but these days, the vast majority of the clicks on ads are bot clicks. (When was the last time you noticed a banner ad, let alone clicked on it?) 

While the number of impressions and the number of clicks are easy to measure and convenient to report, they are only vanity metrics. They are not ideal for measuring the success of digital campaigns. Marketers should have been using business outcomes to judge the success of digital campaigns all along. But they were addicted to the quantity metrics -- large numbers of ad impressions, low prices, the illusion of engagement, and the ease of reporting. That is a habit that is truly hard to kick. Marketers are known to suffer physical withdrawal symptoms when they have to stop the use of bot-fueled digital marketing and metrics. 

Hyper targeting in digital is not brand building

With so many marketers still spending furiously in digital, let’s look at what they are actually doing. There’s a lot of data thrown off from digital campaigns; there’s also a lot of data that can be purchased for targeting ads. Marketers have seemingly subscribed to the “more is better” philosophy when it comes to such targeting parameters. They think that buying more and more targeting parameters for their programmatic ads means more relevant ads for users. But they missed a key consideration -- hyper targeting is not brand-building. They are targeting ads at people who don’t want to buy; or they are retargeting ads at people who have already bought, and annoying them.  

The original idea of hyper targeting meant trying to get the right ad to the right person at the right time. This is seemingly possible in digital, because the ad tech companies have sold advertisers on this kind of magic. For example, pharmaceutical advertisers believe that if they can target “high-prescribers” (physicians that already prescribe a lot of their drug) they can get them to prescribe even more. But in actuality, the pharmaceutical marketer is so focused on targeting ads to this tiny segment, they are completely neglecting the far larger segment of doctors who don’t prescribe their drug at all “no-prescribers” or are “low-prescribers.” Using round numbers to illustrate, if 1% of the doctors are high-prescribers, the other 99% are low- to no-prescribers. Wouldn’t it be better to do brand building with the latter segment that is 99X larger than the former? If only a small number of the low- or no-prescribers start prescribing more, the pharmaceutical advertiser would see a substantial increase in business activity. Trying to get high-prescribers to prescribe more is literally like trying to squeeze blood out of the proverbial turnip. 

The exact same concept applies to other industries. Fellow marketing practitioner, Jake Sanders, said it best “loyalty is not a growth strategy.” If all of the digital marketing were focused on getting existing customers to buy more, then there’s no budget left for brand-building -- prospecting for new customers who have not yet purchased. Unless you’re already a complete monopolist in the market, there’s usually a lot more potential buyers than there are past buyers. If you’re spending most of your digital ad budgets on “shouting at existing customers” and not actively doing brand building to win new customers, you’re missing out on a lot of revenue upside, not to mention you might be annoying existing customers with too many ads. 

Marketers learned an unexpected lesson from the pandemic

Prior to 2020, when everything was sailing along smoothly, no one wanted to rock the boat. But when the coronavirus rocked all boats in 2020, the shockwaves created new opportunities and learnings that would otherwise have remained hidden. For example, Airbnb cut their digital spending in 2020 to “save $800 million during the pandemic” [5] but in Q4 of 2020, they saw traffic at “95% of the same traffic as the year before. More than 90% of our traffic was direct or unpaid,” CEO Brian Chesky said on an earnings call [6]. “We don’t intend to ever again spend the amount of money as a percentage of revenue on marketing … as we did in 2019.” Chesky continued “What the pandemic showed is we can take marketing down to zero and still have 95% of the same traffic as the year before. So we’re not going to forget that lesson. The previous KPI was just about how do you buy cheaper media” [7].

Adidas came to the same conclusion. They were so focused on short-term digital “performance” metrics that they underinvested in long-term brand building. “The problem ... is not the metrics per se but a focus on the wrong metrics. Digital technology offers a wealth of short-term measurements, often in real time, which has resulted in marketing investment being misdirected. We [were] overly focused on digital attribution and digital sales but we are improving,” said the sports brand’s global media director, Simon Peel [8].

The “short-termism” that is now pervasive among marketers spending heavily in digital is particularly harmful in certain categories and industries. Matthew Daniell, a media effectiveness specialist, said “In insurance, it is nearly impossible to get people to switch, less than 1% of the market is in-the-market at any given time. Would you rather spend your marketing dollars getting a small piece of 1%? Or would you rather influence 99% over time?” Long term thinking and media spending on brand building yields far better business outcomes than short term thinking which has exacerbated ad fraud and waste — e.g. “can we hit our number for this quarter or this month?” Australian insurance brand NRMA did just that [9]. By reducing spend on digital “performance” tactics that drove what appeared to be a flurry of activity in the short term (clicks and traffic to the website), they could invest more in long term brand building, storytelling, and standing out. If customers don’t know your brand at all, or can’t recall your brand when they are trying to decide which brand to buy, you won’t get that sale. 

As if on cue, Google announced that they are doing away with ads targeted based on web browsing history (otherwise known as behavioral targeting). Of course this will be debated for months to come, but marketing practitioners have long known that targeting parameters derived from “what websites a user visits” is not very accurate, making the ad targeting not very accurate either. By reducing marketers’ reliance on hyper targeting, Google may be inadvertently helping marketers do better marketing -- by doing more brand building. Brand building does not need hyper targeting anyway, especially bad targeting based on bad data. 

Google to Stop Selling Ads Based on Your Specific Web Browsing

https://www.wsj.com/articles/google-to-stop-selling-ads-based-on-your-specific-web-browsing-11614780021

Rethink Channel Mix and Tactic Mix to do Better Brand Building

If marketers fully realize the problems of ad fraud (botnets generating large quantities of ads at low prices), false engagement (clicks generated by bots to trick advertisers into spending more), and hyper targeting as overtargeting, they would rebalance their media investments. First, rethink your channel mix. For example, are you over spending in the digital channel, when you should be spending more in traditional offline channels like TV and billboards/outdoor that are excellent for branding and awareness? If your primary need is awareness, those traditional channels may do far more for brand building than the digital channel, for all the reasons above. Next, rethink your tactics mix. Are you spending too much in programmatic display and video ads, and not enough in other tactics like paid search and content creation? If your brand already has great awareness, but something else is preventing more consumers from buying your product or service, creating content that helps them through their purchase decisions may yield better business outcomes than just buying more ads.

The Unified Marketing Framework below might be helpful in this process. It takes the typical funnel - awareness, consideration, choice, purchase, and loyalty and turns it on its side. On the left side, we have “branding” and on the right side we have “performance.” You can then map various marketing tactics, both offline (green) and online (blue) neatly in one chart. TV is on the left because it’s great for awareness and branding, while digital is on the right side because it’s best for performance. Even within digital, expanding into the blue bar below, some tactics are more for branding like video ads and display ads, while other tactics like search are more “mid-funnel” and helpful when the user is looking for information during the consideration and choice phases. Choose the channel mix that is right for you; and choose the mix of tactics that’s right for you too. 

All in all, at present too many marketers are spending too heavily in programmatic ads and paying extra for hyper targeting parameters in a futile attempt to drive short term “busy-ness” (lots of traffic and clicks). They should rebalance the media spending so they don’t underinvest in long term brand building, even in digital. That way they can drive more actual “business” and not just look “busy.”