Ten assumptions underlying the misuse of digital advertising

By Nigel Hollis

The potential power of digital marketing is huge. So why am I left with the feeling that the practice of digital marketing often falls short of that potential? I believe it is because the architects of digital marketing infrastructure, and many that use that infrastructure, have forgotten they are dealing with humans not machines. And from that basic misapprehension flows a series of assumptions when digital marketing is put into practice.

Is digital’s allure waning?

Shortly after I started to write this post, I came across this article in HBR titled, Why Marketers Are Returning to Traditional Advertising. As proof of the return, Christine Moorman, Megan Ryan, and Nader Tavassoli offer the reversal of a ten-year trend among marketers of reporting an annual decrease in traditional advertising spending. They report,

“In contrast to the historical trend, in August 2021 and February 2022, marketers predicted that traditional advertising spending would increase by 1.4% and 2.9%, respectively.”

Not a huge increase, but notable, nonetheless. The authors offer several factors driving this trend, including better breakthrough for traditional advertising, the decline of third-party cookies, and concerns over ad fraud and over-stated digital effectiveness.

No putting the genie back in the bottle

To be clear, the degree to which marketers can return to traditional media is limited. While people still watch TV, younger audiences are migrating to video on demand and watching on social and gaming platforms. Whether it is a good thing or bad, people spend huge amounts of time with digital media of all sorts. Marketers must engage their audiences where they are, but they must do so effectively, particularly when digital represents over half the typical company’s marketing budget.

Will investing in analytics pay off?

The days when digital pennies were replacing analog dollars is long gone, but marketers must spend their dollars wisely, wherever they advertise. To that end, marketers are spending big on digital strategies and tactics, including a nearly 40% increase in spending on data analytics over last year. My concern, however, is that much of the investment in analytics and advertising will be wasted if the data is taken at face value or interpreted based on what marketers would like to believe rather than what the data really represents. It does not matter how sophisticated the analytical tools if the real findings are lost in a sea of wishful thinking.

With this backdrop in mind, here are my top 10 assumptions that undermine the effectiveness of digital marketing. I’ll start with the general ones which lead into some very specific actions that make no sense to me. You may vehemently disagree with some or have others that you want to add. I would love to hear what you have to say.

1) Assuming people have neither agency nor memory

What do I mean by “agency”? I mean the power that people have to think and decide for themselves. People are not machines to be programmed. And they are not the blank slate that many digital marketers seem to think they are. People may not invest conscious thought when making trivial decisions, but that just means they unable to escape the influence of instinct and emotion. In fact, very few people make a purchase without some prior inclination to choose a brand. This is where memory comes into play. Even if memories related to a brand do not consciously predispose people to buy it, they are inevitably drawn to the more familiar and the better liked.

2) Assuming advertising “moves” people to purchase

There seems to be a widespread belief that marketing can move people from awareness, to consideration, to action in a short time frame. Let’s be clear, advertising rarely works that way, at least, not in the timescales on which most digital marketing campaigns are employed.

For instance, I have seen it suggested that you need multiple exposures of the same ad to move people from awareness to consideration, to action. This might be true for a completely new and compelling product offer, but most brands are not new. For most purchases, people buy when they are ready to buy. No amount of advertising exposure is going to get them to do something they are not yet ready to do. If someone notices your advertising and they consider it personally relevant then and there, they will likely act on it. This is why the response to the first advertising impression is usually the strongest. Subsequent exposures may ensure the ad is visible to or noticed by more people, but the number who will respond suffers from diminishing returns as reach builds.

I have also seen a similar rationale applied to the sales funnel, where advertising magically “moves” people though the funnel. Again, this is delusional. For instance, most B2B product categories have long inter-purchase intervals and the proportion of people ready to buy now a small proportion of those in the sales funnel, never mind the potential buyers outside it. However, this paper from LinkedIn highlights the mismatch between the average business sales cycle and how quickly marketers try to measure return on investment. 77% of marketers try to measure their ROI after one month, even though the average sales cycle is 6 months. Sure, if an ad campaign does not show some response now, likely it will not move the needle in future either, but these sales are the easy ones, where the advertising is pushing on the open door of someone already willing to buy.

3) Assuming the immediate advertising response is all that matters

This assumption leads on from the previous one. A huge amount of time an effort goes into analyzing the response to digital advertising, and a lot attention is focused on the immediate response to an exposure, or, marginally better, in the next 30 days. If there is no behavioral response, the advertising is assumed not to have worked.

This is a drunk searching under the lamppost problem. Just because behavioral data is easily accessible for digital advertising does not mean that it represents the whole range of advertising effects. Particularly for brand advertising designed to influence future buyers, immediate behavioral response could be minimal, but it does not mean that the advertising has not worked, only that the advertising has not prompted a response yet. And never mind buyers, what about new hires, investors, or retailers? If most advertising works by seeding impressions the most immediate response is a change in memory structures not behavior.

For decades comparisons of attitudinal responses like ad awareness, brand awareness, or purchase intent have found no correlation with clickthrough rates (CTR being one of the most readily accessible and so widely used behavioral metrics). The evidence that CTR relates to sales is not much better. The Nielsen report states,

“evaluating a campaign solely on the number of clicks greatly underestimates the number of potential consumers influenced by the campaign.”

I assume the meta-analysis properly accounts for differences in brand and category sales elasticity, and I have seen similar findings from dunnhumby and others. And yet, CTR directly influences the performance of paid advertising on Facebook, Google AdWords, and other advertising platforms, and is used to optimize campaigns, irrespective of whether it relates to real world outcomes or not. I get it. Clicks allow advertising actions to be taken in real-time. But buyers? Well, they will buy in their own time.

4) Assuming you need to act like a start-up when you are not

I realize that if yours is a new or small brand then you need to do everything you can to drive sales now. Whatever your marketing might be – influencer, social, or search – you better be selling. But remember, everything you do builds your brand, for good or ill. So, make sure your selling is consistent with your brand. And remember, the people who buy your brand in the early days probably appreciate what you have to offer the most.

However, as a brand scales, it needs to reach beyond the early adopters to reach and influence a larger but less enthusiastic audience, and that requires a different approach. You are playing the long game now and you need to reach as many potential buyers as possible. Switchers, new category entrants, upselling and cross-selling, you need to be front of mind when people come to buy, not just responding on the fly. For established brands, most advertising works by building up a cumulative set of impressions over time, ensuring future buyers know why the brand might be relevant to them, predisposing them to choose the advertised brand, and then triggering behavior when someone is the purchase window. It is a matter of seeding positive impressions, ensuring salience, and then triggering behavior when someone is ready to act.

Is brand advertising guaranteed to work? No. Its effectiveness is hugely dependent on the power of content to gain attention and leave a memorable impression of the brand. Buying behavior is influenced by additional factors, availability being the most important, learning of a new and more meaningful brand during the shopping process probably the second most important. Even more reason to make sure that your brand is easily available when and where people want it. Less time searching means less time to find a better or cheaper alternative. This said, I am willing to bet an effective brand campaign has multiplier effects far beyond an ad campaign targeted only to identified prospects.

5) Assuming efficiency is the same as effectiveness

I recently read an article encouraging small businesses to get in the habit of tracking marketing efficiency and ROI. Except, of the metrics cited, the majority were measures of efficiency. Cost per click, click through rate, and cost per acquisition are measures of efficiency, not effectiveness.

Effectiveness is about achieving an objective. So, incremental sales are an effectiveness measure. How much extra did you sell over and above what you would have sold anyway. Ideally you take this analysis further and identify the likely incremental profit made on those sales and subtract the cost of to get a short-term ROI. Why short-term? Because effective advertising ought to have a long half-life and influence future sales, not just current ones. Maybe you want to figure out the lifetime value of a customer and factor that into your calculations.

So, when it comes to marketing, you want to prioritize effectiveness over efficiency. You want to maximize the absolute response while minimizing the cost. However, generating incremental sales means paying over the odds to get quality reach, clicks, and leads. Prioritizing efficiency over effectiveness is a great way of buying crap inventory and wondering where your money went.

6) Assuming you know how the algorithms work

I am beginning to realize that a lot of the tools and algorithms put into place by Google, Facebook, Twitter and the like do not actually do what marketers might want or think they do. Let me hasten to say that I am not an expert in this area, but from what I read, what the algorithms deliver is driven by flawed assumptions and easily available data. Rather than reaching new buyers, they focus in on existing ones. Rather than targeting new leads, they target those prone to clicking.

Take this example from Bart de Langhe and Stefano Puntoni in a recent HBR article. According to them, when you conduct an A/B test on Facebook you are not getting a true read on the difference between the two executions. Why not? You might start by serving ads to similar group of people, but the machine-learning algorithms start to refine the selection strategy to target the ads to people similar to those that click on them.

de Langhe and Puntoni criticize this procedure on the grounds that it violates the idea of random assignment. True. But there is another problem. Some people are what I call “happy clickers.” They will click on lots of stuff. So, what the algorithms really deliver is an audience biased to people who click frequently. This is the sort of thing that happens when you let loose a bunch of clever people who never met an algorithm they did not like but have little clue how marketing works.

And now, some specific assumptions

The previous assumptions are all general ones, with most founded in the idea that advertising can change people’s minds quickly. But combine that belief with the last one and you get some classic examples of hope trumping strategy.

7) Assuming retargeting is effective

Retargeting is a tactic that assumes visitors have no decision-making capacity and no memory. If a visitor believes your brand meets their needs and is worth the price asked, they will buy it, now or when the time is right for them. Yes, re-targeting after a few weeks seems to have more effect than after a few days. Why? Because if you re-target people quickly they are likely still making their decision. If you wait, the buyer has likely done their homework and decided your brand is worth buying. They probably would have bought your brand without being exposed to your ad. Meanwhile you wasted money re-targeting the many people who had bought elsewhere or decided your brand was not what they needed.

8) Assuming additional frequency will help people understand
I have seen this rationale offered as a reason to invest in advertising frequency, and it is another assumption that flies in the face of human nature. If your ad takes a long time to reach an optimal response rate, then most likely the creative is so boring very few people notice it each time it is presented to them. And, no, hard to understand messages will not suddenly become understandable. People are cognitive misers; they are not going to waste time and brain power on trying to figure out what your ad is trying to say to them.

9) Assuming interruptive ad formats will gain attention

I can remember the golden days of digital when agencies were pitching it as a less interruptive and more engaging medium. It was an argument that made no sense then and makes none now. How many times today have you been forced to click on the X to get rid of an interstitial ad? How many times have you waited 5 seconds to skip the ad before the video you really want to see? All interruptive ads are irrelevant, because they are not what the person really wants to see or read. Interruptive ads are definitely seen, but the vast majority are either ignored, or worse, put people off using the site or brand.

10) Assuming that buying your own brand name prevents poaching

Thanks to Dr. Grace Kite for reminding me of this one, and judging by the comments to her post, we are not alone in believing buying your own brand name in search is a wasteful idea. Yes, it stops competitors buying your name, but 9 times out of 10 that is probably a wasted investment. If people want to find your brand, and the fact that they searched on your brand’s name suggests they do, why would they click on the competitor’s paid ad? If you searched for Nigel Hollis and found this site, is it likely you would click on an ad for Peter Smith? Maybe you lose a few sales by not buying your brand name, but you probably would have lost them later in the buying process anyway. The opportunity cost of buying your own brand name is big to marketers and the only beneficiary is Google.

An ecosystem problem or a marketing problem?

So, the big question is do these assumptions arise because of how the digital ecosystem is structured, or just the way it is used?

If it is the former, then maybe the downside is less because everyone suffers from the same problem. Marketing is less effective than it could be (and maybe once was) but then a sinking tide lowers all boats, and the relativities remain the same. Marketing is largely a market share battle, so if everyone faces the same limitations then gaining share comes down to using the available tools more creatively and effectively, even if those tools are not ideal to the task.

But if most marketers are simply ignoring or misinterpreting the evidence, then there is a far bigger problem, because it compounds the problems imposed by the digital ecosystem. It horrifies me how much time people spend sweating the details of targeting and search optimization, and word-smithing content to rank well on Google, all the time ignoring the single biggest competitive advantage they have, the ability of their content to get someone to stop, pay attention, and remember a compelling and motivating impression. My recommendation is leave all the tactical stuff to an AI and focus human creativity on what really matters. But, hey, that’s just my opinion, because like most people I try to think for myself.