4 Unpopular Truths About Digital Marketing for 2023
This post was originally published on Nov 21, 2022, in my Medium blog.
The past decade was glorious for most marketers. As transactions of all kinds increasingly moved online, social media grew exponentially, smartphones and apps became ever more capable, and the amount of data available to us exploded, it seemed like the party would never end.
Budgets seemed set on getting bigger and bigger, even as increasing automation would keep making our jobs constantly easier. But the reality is now setting in.
Access to data is weakening, buyer behaviors are changing rapidly, and new methods are going to be needed to convince consumers in the future.
While some marketers have already recognized that things have changed irrevocably, the unprepared majority is in for a rude awakening.
So, to help you prepare for the future, here are four unpopular truths about digital marketing for 2023.
1. Direct Response Performance Marketing Still Works Great — But the Glory Days Are Over
No, performance marketing isn’t dead! However, the era where you could (almost) just increase your Facebook or Google search ad budget by hundreds of thousands of dollars per month — and still get a corresponding increase in sales — is over.
With Apple’s App Tracking Transparency policies in place, and Google working on something similar for 2023–2024, the amount of data available for ad optimization keeps declining. This reduces the effectiveness of e.g. social platforms’ advertising algorithms. It is already especially apparent for advertisers relying on mobile app revenue.
Things might not be quite so gloomy in the end though. People’s purchasing behaviors won’t be entirely upended even if access to data is. They will still buy when they see a great ad — what will get weaker is your ability to attribute the results to it. It will be harder to find the right people as effectively as before — but it won’t be impossible.
Marketers will need to get better at the core of their job
The fact is that marketers will need to get a lot better at doing the core of their job: delivering the right message, to the right audience, at the right time in order to generate revenue.
Have things gotten more complex? Yes and no. The key principle of crafting marketing campaigns that perform is still valid: test many different things, learn what works, then iterate and scale based on that.
You will need to increasingly test new marketing mediums, channels and creative formats to find effective means of reaching your target audience. And no, that doesn’t just mean blasting your existing video ads all across TikTok!
For example, partnering with influencers is a great idea for most brands as they can reach the brand’s target audience without granular ad targeting capabilities. Advertising can also be made more convincing and relatable to broader audiences by leveraging user-generated content more.
Influencers can be a great way to effectively reach your target audience. Photo by George Milton on Pexels.com.
As granular ad targeting strategies become less valid, you need to focus on the core job of delivering the right message to your potential customers.
This means you must put effort into crafting better ad creative and messages, something that has been overlooked by many in the past few years in favor of more complex bidding or targeting strategies. Get it right, and results will probably follow.
Which leads me to the next point. Even if the effectiveness of performance marketing is going down, it’s not necessarily a bad thing. You were probably already overspending on it because of the way attribution tends to be done.
2. Most Companies (Still) Use Last-Touch Attribution as a Crutch
I remember struggling with last-click (or more accurately, last indirect click) attribution already in the early 2010s. Unfortunately, many companies still favor these attribution models to an insane degree.
This makes marketing channels very close to the purchase look far better in analytics and attribution software than they actually are. It also causes companies to overspend on direct response advertising because it is (or at least was) easy to make the results look good.
While this kind of attribution may be accurate for certain low involvement consumer purchases or new product categories where very few solutions exist, most purchase journeys are actually far longer and more complex than most people think.
The reason for this last-click addiction is often managerial. People with ultimate decision-making power in a company — such as the CEO and the board of directors — are used to these models and prefer them because they (overly) simplify reality and seemingly give a yes or no answer on whether something is working or not.
Even if you work for a CMO (or are the CMO) who is convinced of more modern approaches, but answer to people who are only used to looking at results based on last-touch attribution models, the temptation to optimize towards these models that make you look good can be irresistible.
If you are in this situation, the most critical thing to do is either A) convince the decision makers to adopt a more holistic approach or B) switch jobs.
This sounds extreme, but if you are in a situation like this, you will eventually set yourself and the marketing efforts of your company up for failure (and in the worst case, you spent millions of dollars on the way). It might not happen in the next few months, or even a year, but it will most certainly happen at some point. And you certainly don’t want to be around when that happens.
This doesn’t mean last-touch attribution (or attribution software in general) is not helpful at all. However, it’s time we recognized that we need to move away from it as the single source of truth, and embrace other methods that take into account the full purchase journey.
So, what kind of model does work?
First of all, no single attribution model is perfect or fully representative of reality. You usually need to look at several models and use additional external inputs to determine how each channel is performing.
Based on these inputs, you can then build an estimate of how much each marketing channel is contributing to revenue, and at which part of the purchase journey.
Data driven attribution models that have become more robust in the past few years can help, but even they do not accurately capture many behaviors.
For example, both of the following conversions would probably be claimed as direct traffic no matter what kind of attribution model or software you use:
A person watches a video from you on their phone and then buys from you later on their laptop by directly going to your website.
A person sees your TV ad, goes to your website on their phone and makes a purchase.
Analyzing the most common customer journeys (through e.g. data and interviews), building organic uplift estimates, and further analyzing the potential true sources of direct traffic can provide you with very valuable insights that can more accurately capture behaviors in different marketing channels and allow you to give them the credit they deserve.
Customer and purchase journey mapping can be immensely helpful for your marketing efforts. Photo by UX Indonesia on Unsplash
If you have more robust analysis resources available, you should also probably look into things like Media Mix Modeling, which can help you build a holistic picture of which marketing activities actually drive revenue.
In addition, self-reported attribution is something you should explore, especially in the B2B space. You can start as simply as providing a mandatory open-ended text field asking “how did you hear about us” in your contact form (use open-ended text specifically, never use a dropdown). You can also have your sales team ask about it during sales calls.
For even more detailed results, you can hire a company specialized in customer journey analysis to conduct interviews with your customers (or do it yourself) to build a better picture of the most common purchase journeys.
While this answer can seem vague (or even depressing), the truth is that attribution of any kind isn’t perfect. Different attribution techniques and models are simply inputs for you to build a holistic view of what works and what doesn’t.
This sounds like hard work, right? It is, but the rewards far outweigh the time investment.
3. Customer Experience and Time to Value Affect Your ROI a Lot More Than Your Ads
Every marketer loves to say that their recent ad campaign has been a resounding success. The greatest ad campaigns can indeed have a staggering impact on sales. For example, Apple’s 1984 campaign resulted in 50% higher sales than the company’s most optimistic scenario for the 100 days from when it first aired.
However, most of the time the reason for success is not your ads — it’s the product and the experience of buying it. But because of how well performance marketing worked in the past decade and how scalable it has been, people have been getting away with under-investing in other areas.
Ads, especially great ones, can attract a lot more people to you cost-effectively, but what ultimately convinces them to convert, make a purchase, and become a recurring customer, is:
Great customer experience (including e.g. the user experience of your site and potential customer service interactions)
A great product, and an attractive presentation of it and the key benefits
The time to value from that product (for a great breakdown of time to value, see this article from Paddle).
Improving your ads can help improve sales (and you should definitely do it), but improving these critical aspects can help more — exponentially more.
You should constantly be asking yourself how you could improve what matters the most for your company (it’s usually more revenue). A lot of times this won’t be achieved by simply spewing out more and more ads with updated copy or creative.
Here a few examples of what you should consider:
Could you improve the load time and information of landing pages where you send people?
Could you ungate your ebooks so people get value from you faster?
Could you offer free express delivery if a customer makes a big enough purchase?
Could you offer a customer service chat to people to answer questions they might have before a purchase?
Could you work with sales to provide a faster way for leads to talk to your team?
Does a customer absolutely need to talk to a sales rep before getting software access? Why couldn’t a customer just provide their payment info and access at least some areas of your product immediately?
Why doesn’t everyone do this? A lot of times it’s the result of working in silos and lack of empowerment in the marketing organization. If you don’t collaborate closely with other departments, it can feel difficult to improve things not directly under your control. But does that mean you shouldn’t even try?
What you are ultimately responsible for as a marketer is generating as much revenue for your company as profitably as possible. Not leads, not website visits, but actual revenue. So maybe it’s up to you to take that first step?
4. TikTok Poses an Insane Level of Societal Risk — But You Are Going to Advertise There Anyway
Photo by Solen Feyissa on Unsplash.
TikTok’s growth is fast, mind-bogglingly fast. The app exceeded a billion users already in 2021 and the growth looks set to continue. And where users go, advertising money follows.
As Eric Seufert recently highlighted in Mobile Dev Memo, Apple’s ATT policy seems to have hit TikTok harder than previously anticipated, but they are still predicting that they’ll reach around 10 billion in ad revenue this year.
However, widespread use of TikTok poses an insane level of societal risk for most countries that aren’t called China. For example, despite claims to the contrary, Chinese employees have accessed US user data on several occasions.
The US entity is technically separate from the Chinese parent company, but as discussed for example in Wired, that might not mean much.
And what would happen in a situation where a government actor wanted to affect things a lot more? For example, what happens if there is an armed conflict in Taiwan and they want it downplayed? What if they want to sway opinions in an election in another country? If government employees show up at the parent company Bytedance’s offices and “politely request” data is handed over, or the functioning of algorithms changed, do you honestly think anybody would refuse — at least for an extended period?
But hey, I’m sure we’ll be fine because they promised they wouldn’t do it right? And we can totally just take their word for it, right?
I don’t want to sound like an alarmist, but I feel this risk is so massive that willful ignorance can’t last forever. A backlash, and potentially an acquisition of some kind of the US-based TikTok entity, is probably inevitable.
But until that time, there’s all those people using TikTok, so you will be tempted to fork over your advertising dollars. Most likely you have already done so, or will end up doing so — and the results will probably be pretty good in the near term.
Just be ready for the unexpected, don’t make yourself overly reliant on TikTok and if (or rather, when) there’s a backlash, don’t claim you didn’t see it coming.
Closing thoughts
So there you have it, four unpopular truths about digital marketing for 2023.
While accepting them can be difficult, and successfully navigating the challenges even more so, I believe embracing them ultimately makes marketing, and marketers, better in the long run.
While this post might have set a fairly negative tone, I’m actually looking forward to everything 2023 has in store for marketers. Marketing channels and strategies will keep evolving, and it will be exciting to see how the best brands and products take advantage of them.
Thank you for taking the time to read through this post (the first one on my Medium blog)! I would love to hear your opinions on these four aspects, and especially if you think there are any other unpopular truths out there.
Until next time!
