3 Things Marketers Should Always Order Off the AI Secret Menu

We know everyone loves their tried-and-true staples when they’re ordering off a menu—there’s comfort in what we already know.

If you saw our most recent AI blog post, you probably understand two types of AI: predictive and generative. And although neither of those produce Thomas Keller quality memories on their own– you know what to expect and have seen them in action.

But you should be asking for something better. We’re introducing you to the AI capabilities secret menu– functions that you may not have known existed– or didn’t realize you could even ask for. It’s everything you’ve always wished for, but haven’t found yet.

When it comes to AI capabilities, here’s what to ask for - even if it's not on the menu yet.

It’s time for your starter, and nothing will light your AI taste buds on fire faster than brand context capabilities. With Big Data, you have subscriber context. Where current AI functions could be falling short is brand context. If you're operating a Shopify business, your branding matters– it’s likely a major competitive asset that you're investing a ton in.

It’s reasonable for you to require that your AI-generated marketing communications are on-brand. Brand context should be a prerequisite before leveraging AI for content generation. Otherwise, you’re stuck with random copy that doesn’t sound like it came from your brand.

Our point is, existing tools allow you to harness subscriber data with predictive AI, but you’re missing the brand voice layer. You should expect that AI-powered software enables you to add and adjust your brand voice so it can generate on-brand content.

You’re probably familiar with generative AI (AI that uses unstructured data, like from publicly available sites on the internet, to produce content). In our space, or on the regular menu, this could be presented as “Message Generation” or “Subject Line Generation”. And predictive AI– which on the menu could show up as “buying propensity” or “ideal send time.”

Don’t get us wrong– these use cases are important. But you should be ordering AI for revenue generation. 

A cheeseburger simply being listed on the menu doesn’t mean that cheeseburger is going to be delicious. It just means it’s offered. Analogy aside, just because a function (or platform) is marketed as “AI” doesn’t mean it’s going to get the job done. 

An effective AI solution should ingest data about your customers, data about your brand, then layer predictive AI AND generative AI over the top so you can see optimal results.

You should expect AI to help you measurably make more money, proven by holdout testing so that you can trust, but verify that the results are real and incremental to your business.

When results are generated and when you start making more money as a result of the AI, you should expect to be informed about the elements that are generating revenue. 

Hopefully, you saved room for the dessert, because the cherry on top is AI that gives you actionable insights. Your AI should tell you something about your business that you didn't know before.

The right AI may even give you the ability to lightly influence the model to go in different directions and see if you can drive additional results. Ideally, your AI partners with you to make sure you still have your hands on the wheel, can trust, but verify results, and partner with you on both strategy and brand context. 

Closed mouths don’t get fed. Ask for what you want and expect more from AI. When we say the future of AI is at Postscript, we’re serious. Register for our next webinar to stay in touch with what we’ll be launching soon.