Personalization has been somewhat of an industry buzzword for years; with a number of tech media outlets insistent that this year, is the year of personalization. It’s been that way since 2014, but in the real world, this has never come to fruition…. Why? The fact of the matter is, personalization is hard. The industry has been sold a dream of 1-2-1 personalization; but a severe lack of education in the industry has left a massive blur between optimization, experimentation and personalization. It seems that people expect to plug a tool into their site and hey presto! personalization is live and sales will increase… That’s not the case, and if you’re going to do this right; you’ve got to do the hard work!
So, what can we do about it? I frequently hear people say “isn’t that an experiment, not a personalization?” and whilst I understand the premise behind this statement; it shows why we’re failing as an industry when it comes to personalization. We’ve separated personalization from experimentation, rather than separating personalization into (achievable) stages; where, 1-2-1 personalization comes after a period of experimentation, learning and insights.
We need to start thinking about personalization as creating personal interactions and experiences for core user segments, by dynamically serving content or recommendations based on the data and insights you hold on your user base. Based on over 5 years experience doing this, believe me when I say that if you’re tasked with improving the performance website, your efforts are much better placed on understanding the anxieties, motivations and usability issues on your site with the different user segments, than they are, searching for the right tool to do this for you, based on their super secret algorithm…
Think about it; if you’re working on a clothing site will a user buying a pair of jeans have the same anxieties as a user buying a pair of shoes? No. Even if these anxieties are similar e.g. sizing, the way to address this problem is going to differ… Don’t spend your time setting up simple personalizations in a tool, spend your time (and money!) understanding the differences in your users and creating bespoke experiences based on their needs. I guarantee you, that in the long run the returns will be so much bigger and more sustainable. Not only that, but the knowledge your business gets alongside this, will be so much more powerful than “a red button seems to work on Wednesday, but a purple button works on Friday” (tongue in cheek… but you get what I mean…)
Automated and tool-driven personalization campaigns take content as a starting point (you input a test into the tool) and identify the user segments afterwards, based on the results of the content you provide. At this point, we’re ‘hitting and hoping’ to some extent – i.e. we create something, figure out who it works for and dynamically deliver it to them. If we’re to turn this on it’s head however; and focus on identify your segments first and then focus on delivering tailored content (based on user insights and specific anxieties and motivators), you’ll increase your creativity (because you’ll be thinking about user segment anxieties, not solution first), and you’ll provide a better experience, believe me.
We use a simple 3 step process with our clients; based on delivering effective personalizations. In step 1; we take the crowd sourced / generic data, to understand how the ‘general’ user, uses our clients site and start to make improvements and suggestion based on this data. Think Facebook… if you’ve logged in recently; you’ll see recent stories. If you’ve not logged in recently, you’ll see the most popular stories. It’s all about giving the users, the most relevant content based on what everyone else is doing. This probably ‘technically’ isn’t personalization, but during this time we’re focussing our attention on understanding differences that exist in the core user segments.
Step 2 is about identifying our segments (usually based on category / price point and device type etc.), identifying samples of users where we can run A/Bn tests to understand their drivers and blockers to purchase. We’ll run continuous testing amongst these audiences, meaning that over time the journeys and page(s) evolve away from the ‘base template’ on site, into a personalized journey, specific to that audience type. This then means that within step 3, we can focus on the dynamic content for the small sub-sets of the audience (e.g. gifters vs. non-gifters, new vs. returning), where the page template doesn’t need to change, but the content, recommendations etc. might.
What this delivers is a fully personalized journey for the user; where they feel like the site is tailored to their needs. That’s effective personalization. It’s much more effective at delivering results and insights to the business; and the people making the decisions are seeing continual change and growth on-site, as well as increased ROI from tools such as recommendations etc.
The sooner we start building bespoke journeys to our core segments; the sooner personalization becomes viable. After all, the last thing we want if for 80% of marketers to abandon their personalization efforts.