The PEE Framework for Agentic AI — Whiteboard Friday

Click connected the whiteboard image supra to unfastened a high-resolution version!

Happy Friday, everyone, and invited to Whiteboard Friday by Moz. I'm Rejoice Ojiaku, and I americium a Senior Business Content Specialist astatine Wise. And today, I'll beryllium taking you done the PEE model of mounting up agentic AI.

Now, erstwhile we talk astir AI search, we are proceeding position specified arsenic GEO, AIO, LLM optimization, and each of those fancy fancy words. But really and truly, I deliberation we are doing wholly excessively overmuch because we already understand the usefulness erstwhile it comes to SEO content and each of the astonishing tricks that we person learnt complete the years. But erstwhile it comes to AI search, it's fundamentally the aforesaid thing, but we person conscionable analyzable it.

Now, let's really return a heavy dive into what really is the travel of AI search.

When we person our content, our contented gets fed into an AI knowing because AI is seeking to understand what our constituent is, because they are trying to make it extractable.

That gets fed into an AI reply because, ultimately, AI is trying to springiness the users a clear meaning aliases clear points successful what we are trying to communicate.

And because of that AI answer, it improves our brand visibility. We want to guarantee that done being visible successful AI hunt landscape, we are mentioned, we are cited, and we tin beryllium summarized. But really tin we do that? There has to beryllium an easier model for america to travel and for america to understand. 

We are told to make a point. And akin to AI search, the constituent that we're making needs to beryllium clear. It needs to show your users precisely what you're trying to opportunity to them.

But you besides want to supply evidence, and this is conscionable things that support what you're trying to say. This tin beryllium stats. This tin beryllium usage cases. This tin beryllium different examples aliases moreover further context.

Now we request to link the dots wrong explain. So why does your constituent and grounds matter?

If you've not been clever aliases clear capable successful what you're trying to opportunity and you're not connecting the dots, AI will find a measurement to capable successful the gaps, and therefore, we suffer that communicative that we want our users to understand. 

As you tin spot here, the first illustration says, "Good merchandise descriptions are important for ecommerce." This is not technically wrong, but it is simply a spot excessively vague. It's not needfully telling america a why, for example.

But if we look astatine the after, the aft says, "Clear merchandise descriptions thief ecommerce brands summation conversions and trim returns." This gives america a spot much context. This will alteration AI to extract, summarize, aliases mention immoderate the personification has asked them to.

So it isn't fundamentally each astir penning more. It's astir penning clearer. And that is why we request to ever spell backmost to the fundamentals. 

But you're astir apt asking what happens to your existing contented currently.

Well, it isn't a ace reliable hole that you request to benignant of spell through. You tin travel these 3 easy steps successful position of the mini fixes you tin do erstwhile you're looking astatine your existing content. 

1. Lead pinch an answer

So the first measurement is you want to lead pinch an answer. Look done your contented and really inquire yourself: "Are we answering immoderate the personification is trying to know? Are we getting consecutive to the point? Or are we hiding the reply halfway done a paragraph that's very difficult to find?" 

2. Support pinch proof

Afterwards, you want to support pinch proof. Look done your erstwhile content. Do you person immoderate usage cases that you tin usage to support the answer? Do you person immoderate stats, immoderate benignant of reports that you tin connect to it? If so, see those. 

3. Remove unnecessary complexity

Then you want to region immoderate unnecessary complexity. Now I get it. I've helped constitute contented before, and sometimes you tin get mislaid successful the source, you tin get mislaid successful the fluff. But we want to region that complexity because if it's not adding to the answer, past you're conscionable buying words. That's thing that AI will not reward you for. It rewards clarity. It rewards freshness of content. 

But this is not to scare you astir AI search. AI hunt is conscionable taking america backmost to the fundamentals. That's each it's doing. It's not reinventing the wheel. It's not giving america this full caller world. It's showing america erstwhile bully contented becomes very obvious. So you request to consistently wonder: "Is my contented answering the question? Is my contented really starring pinch facts and starring pinch clarity?"

And if it's not, rewrite it because, astatine the extremity of the day, it's not that complicated.