The unfortunate reality is Artificial Intelligence is not going to make content better through personalization. In fact, the majority of AI Content personalization systems are often more beneficial to agency bottom lines than their clients' businesses, much less those entities' customers.
This isn’t to say there isn't merit in these optimization technologies, but here are a few steps to consider before you unleash AI on your content, and more importantly, on your customers.
Crawl, Walk, Run to Content Personalization
It may be incredibly simple, but there is massive opportunity in building content that people actually want to read. The challenge today isn’t that companies don’t want to make content -- they are making tons of it. The challenge is that no one seems willing to edit.
Our recommendation: Run a content audit on all site pages to measure performance. One of the most basic approaches would be to download a spreadsheet of pages using a tool such as Screaming Frog (good tool - horrible name) then using the Google Analytics connector to see what's working and what isn’t.
Second, adjust your URL structure to be more readable (instead of www.moo.com/ThisIsALink, make it www.moo.com/this-is-a-link) The second option is easer for both humans and robots to read, and better comprehension contributes to better SEO.
Finally, look at including header images in the Open Graph Settings for your long-tail articles. This can dramatically increase the performance of any article shared in social channels.
Once we have a baseline set of content elements, we would recommend allotting a micro-media budget to retarget growth members with relevant content, with the aim of incrementally building engagement between readers and the organization.
The idea would be to have a ROI-positive campaign that either directly builds on an organization’s customer base, or leads to a course or conference sign-up.
Then, you'd use the data captured there to utilize hyper-personalized platforms, such as Google and Facebook, to better understand the customer and the content’s effectiveness.
Once we have fantastic content that people love to share and that drives great organic search (earned media), and we have been able to test various personalization scenarios that match content to customer type, we would then turn to current tools and usage patterns to determine how we could make those things even better.
Our recommendation: Start with some basics, like allowing customers to opt-in to our process of personalization. Also, always check your personalization against a rule-set. Then, move onto prototyping against platforms and incrementally rolling out new features in a way that makes members think “they really get me.”