July 25, 2019

Want to build content and marketing personalization that doesn't annoy your customers? Start here.

Want to build content and marketing personalization that doesn't annoy your customers? Start here.

This post was originally published in The Content Technologist newsletter on July 25, 2019 with the subject line "Let's Get Personalized."

In the varied and sundry B2B surveys about marketing tactics, more than half of marketers say they are using personalization, which feels wildly inflated considering most marketers don’t have a unified idea of what personalization is.

I’m not talking about ad targeting, when personalization relies on 3rd party “big data” and is generally contextual rather than truly personalized (if you want to talk semantics, hit the comments). Ecommerce is also excluded; product recommendations are a whole ’nother ball game. But for organic 1st party content distribution based on owned data, common personalization examples include, from most to least frequent:

  1. Inserting a name or, at best, job title in an email through basic database segmentation;
  2. Serving location-based promotions, which are not particularly personal;
  3. Serving demographic-based offers, which is mindnumbingly problematic;
  4. Distributing related content through basic segmentation such as “once you clicked on a story about food; here are more food stories.” (This is why I only see restaurant news from my local newspaper and not the more nuanced local reporting I’m interested in and, well, ultimately choose another local news source.)

Whether the four above rely on human-defined segments or determined by predictive machine learning from a large-ish 1st party (owned) data set, they’re wildly biased and rarely feel particularly personal to the user. For example, I am not a huge fan of My Chemical Romance, but for some reason one of my personalization engines wants to serve me every bit of gossip about the former members of My Chemical Romance ever because I searched for the video of one their songs at some pivotal time that now will define my search results forever.

I am large, I contain multitudes, I’m a Sleater-Kinney fan, and look: it’s unlikely that your database is ever going to figure me out. Unless you’re Spotify. Or Netflix. Or, on a smaller scale, Co–Star. So I mean… it can be done. But take baby steps.

The 5 Rules of Content Personalization in 2019

The best writing advice I’ve been given is: write what you want to read. It’s the same with personalized marketing: Personalize how you’d like to receive your personalization. Names and locations can be cute, but they’re the equivalent of small talk. If you’re going to personalize, do it right.

  1. Ask me questions so I know what data you’re collecting. I am happy to provide relevant information as long as you’re paying attention to your privacy policy. (I know we need to have a long discussion about ethical privacy standards in marketing, but we’re going for Idyllic Personalization here.)
  2. Read the room. Know how frequently I engage with your content or whether I’m a subscriber. Use personalization to improve my user experience.
  3. Give me the benefit of the doubt. Assume that I am consuming your content with positive intent, a full schedule, and a nuanced mind.
  4. Keep the conversation going. Asking a new question each time I visit your site or open an email actually makes it more interesting for users!
  5. Don’t box me in. Show me personalized content first, but make everything available so I can find it.

Mostly I want your website to remember that I have already signed up for your newsletter and yes, I visit every day so I don’t need that pop-up or those emails (aiming my stink-eye at Pitchfork and The New Yorker and every other Condé Nast publication).


Audience interaction | personalization | web | email | data | ethics


Want more Content Technologist in your inbox every Thursday? Forever free for the first 1,000 subscribers.