Track Or Troll! Creepy Marketing Tales For Halloween

24 Oct 2023

As Halloween approaches, we couldn’t resist the temptation to embark on a spooky adventure exploring the “creepy” side of CXM. These spine-tingling tales reveal the secrets that ghoulish marketers and sinister spirits use to cast their spell over unwitting consumers. Join us in the Martech House of Horrors – if you dare!

Socionics Visual Identification: Spooky (pseudo) science

Spookometer: 👻👻/5

Socionics, like its cousin phrenology, says physical features like facial contours can give away personality traits. By gazing into the abyss of your LinkedIn profile photo, Socionics VI (visual identity) algorithms use the principles to classify you as an “INTJ Vampire” or “ESFP Ghost Whisperer”. Marketers can use this for segmentation, targeting and product development. 

But does it work?

Although entertaining, the underlying science of Socionics is more séance than lab experiment. We recommend approaching Socionics VI with a healthy dose of scepticism and an analytical third eye. 

Speech-to-face and AI avatars: Digital alchemy or spellbindingly scary? 

Spookometer: 👻👻👻👻/5

Speech-to-face tech is an AI that, you guessed it, converts speech into facial expressions. It’s done by analysing tone of voice, pitch and inflection, then mapping them to facial muscle models. 

Let’s be clear: current models can’t conjure your face from your disembodied voice. Yet. 

The tech is trained on large data sets like YouTube. Reconstructions are based on averages rather than individuals, and models are subject to all kinds of hair-raising biases.

But your digital doppelgänger could be lurking in the shadows. 

MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) is already working on a Speech2Face AI model that’s spookily lifelike. 

Synthesia creates high-quality AI avatars that mimic human expression. Current avatars are based on actors, but they’re convincingly natural. And Synthesia is in cahoots with Speechify, a company working on natural-sounding AI voices with aspirations to create a real-time conversational AI.

Then there’s BHuman using AI to create hyper-personalised outreach videos at scale.

On one side, there’s potential to improve training and development, personalised content, customer service and market research. But this is an article about creepy marketing. We’re obliged to imagine an Upside-Down world where powers combine to create better deepfakes that bewitch unsuspecting victims. 

Ethics, transparency, consent and responsible use are going to be critical. And that’s the scariest idea of all.

Psychographics: The possibilities are haunting

Spookometer: 👻👻/5

While demographics (age, gender, location) and behaviours (clicks, purchases) provide valuable insights, they only scratch the surface of understanding. 

Psychographics delves into a consumer’s beliefs, values, goals and intentions to understand them on the psychological and emotional level.

Demographic profile Psychographic profile
Male Enjoys healthy living but also likes socialising
Aged 25-35 Values aesthetics and quality in fashion, home décor and accessories 
Single Friends are all getting married and settling down
No children Trying to balance career aspirations with ambitions to travel the world
Lives in London  Likes the busy city but will take at least one beach holiday every year
Earns £60,000 per year Saving to buy a home with a (future) partner

Netflix, a pioneer in the field, uses psychographic profiling to recommend content. Facebook uses it to target ads, and Amazon uses it for personalised product recommendations. 

Where psychographics gets creepy

Digital marketing is naturally moving towards psychographic profiling, for example by analysing data to identify motivations and intentions. Some martech even offers this as an OOTB feature. 

Still, as CXM consultants, we urge responsible use. You should only have the data you need to achieve your goals.

  • Data should improve the mutual value exchange
  • Best practice calls for transparent and visible data management
  • Consumers should understand how sharing their data improves their experience

The key lies in defining the drivers of data collection and use. Marketers must harness the power of psychographics to create tailored consumer experiences, not manipulate mere mortals.

Worried you’re haunting your audience?

Dark Patterns and the Dark Funnel: A foray into the dark arts of CXM

Dark patterns and the dark funnel aren’t really related. Although, you could argue that both are shortcuts. Which begs the question: where is the line between trick and troll?

Dark patterns

Spookometer: 👻👻👻/5

Dark patterns that intentionally mislead, trick, or coerce users into buying products, signing up for services, accepting cookies or sharing personal data.

  • Pre-checked boxes trick users into signing up for unwanted services or subscriptions.
  • Hidden costs like shipping fees deceive users about the true price of a product.
  • Confusing language makes it difficult for users to understand T&Cs.
  • Limited-time offers apply false pressure, urging users to make a poorly-thought-out purchase.
  • Social proof like a ticker of past customers bewitches users into buying under false pretenses.
  • Confusing interfaces mislead users into completing an action, such as highlighting the “Accept All Cookies” button.

Dark patterns are more misleading than they are mysterious, but it’s still worrying how prevalent they’ve become. Thankfully, the average user is getting better at recognising dark patterns, weakening the spells that sinister marketers use to bind unwitting customers.

Dark Funnel

Spookometer: 👻/5

“Less than 3% [of B2B buyers] intentionally engage revenue teams early in the buying journey. Instead, they conduct online research in stealth mode. But there’s a twist: Buyers create digital “breadcrumb trails” of buying signals as they research. This crucial, unseen data hides in what we call the Dark Funnel.” – 

What martech vendor 6sense calls the Dark Funnel™ isn’t as sinister as it sounds. It merely aims to bring ghostly customers into the physical realm. Essentially, the Dark Funnel™ de-anonymises user data from touchpoints like PR, social media, review sites, podcasts and online videos to give marketers a better understanding of who’s haunting their market.

Augmented reality advertising: Crossing the threshold between realms

Spookometer: 👻👻/5

With more than 1.7 billion users expected to embrace augmented reality (AR) in 2024, more and more marketers are investing in AR ads. 

The better AR ads help consumers make purchase decisions. Virtual clothes try-ons, IKEA’s AR app and interactive 3D product demonstrations are the ‘good spirits’ of CXM.

However, once we start seeing AR glasses in the wild, it’ll be a split-second before contextual advertising appears in front of wearers like the ghost of browsing sessions past. Although we’re all expecting it, there’s still an element of intrusion – and the potential for dark patterns.

Users beware, for ghouls lurk around every corner

Family tree data

Spookometer: 👻👻👻👻/5

You’d have a hard time convincing a digital marketer to take a DNA test. The idea that one company has access to a treasure trove of sensitive personal data should be enough to make your skin crawl. 

From creepy marketing to genetic discrimination by insurers to sharing your data with other parties, the value exchange doesn’t add up.

Pregnancy predictions

Spookometer: 👻👻👻👻👻/5

You might have seen the story from a few years ago about a US retail chain recognising a young woman was pregnant before she’d told her family. 

Whether it was pregnancy or something else hardly matters. The upshot is that predictive analytics crossed a creepy marketing threshold. It teaches marketers a valuable lesson: data should improve the mutual value exchange

Licence plate analysis

Spookometer: 👻👻/5

Petrol companies are using Automatic Licence Plate Recognition (ALPR) tech to offer loyalty discounts, personalised ads at the bowser, invest in electric charging stations, and simplify payments.

But there’s a catch. At least one company we know of is exploring ALPR to build demographic insights, like using car models to estimate income. So the next time you think petrol costs are scarily high, consider the additional costs of your data.

Gait analysis

Spookometer: 👻👻👻/5

A person’s walk is supposedly as unique as their voice. Gait analysis – monitoring and measuring a person’s limb movements – has been mainly used in security and forensics. 

But you can bet there’s a savvy marketer out there designing a use case. It will most likely appear in future retail stores, where gait analysis (together with other -graphic data) will help design shopping experiences, recommend products and predict buying behaviour.

The real horror story is closer to fact than fiction

Take a lesson with you as you leave the Martech House of Horrors. These creepy marketing tricks might seem like the stuff of ghost stories, but many of them have already crossed into the physical realm.


Exchanging data for a better experience is a fair trade. Just remember that if you’re not the customer, you’re the product. Be wary of companies with obfuscated data policies and a taste for dark patterns, because you might be giving up more than you know.


Get the data you need to understand and serve customers, but make it compliant. And keep it clean. Not only is good data management ethical, but dodgy tactics could attract hefty penalties and damage the brand’s reputation. 

Who you gonna call?

As martech consultants and database experts, TAP CXM provides tools and training to optimise the data value exchange without sliding into creepy marketing.

Book a seance with our CXM specialists to communicate better with your customers across the threshold.

Related posts