| | |

Training the Data

Training the Data - preparing for Thought Leadership program on Sorilbran.com

NOTE TO READER: I think in loops and patterns. I prefer to train heavy and run light, meaning I do a lot of the strategizing on the front end for how something should run, taking into account what I see coming down the pike. 

In this log, I’m walking you through a messy middle – how I’m thinking through personal branding before I stand up a thought leadership program within an organization. This is the thought work – the loops, the experiments, tracking how long it takes for AI to shift, from input to next search. Tracking how long it takes search engines to shift when it comes to personal branding. Figuring out which channels have the most impact on visibility, and what types of content are sticky for AIs. 

I do all the thinking on the front end (this is my front end for thought leadership efforts), then I map out repeatable processes (frameworks), build systems to manage outputs, automate as much as I can without compromising the integrity of the output, and then I keep it moving. 

So, if you ever wanna know how I come up with frameworks, this is a pretty solid example of my messy middle.


Quick Context

In 2023, I was fascinated with the bio Google’s SGE pulled together for me when I Googled my name. The summary was built using stuff it found online – old articles, old profiles, new articles,  social media. It was cool. And telling. So, I started talking about it in a newsletter on LinkedIn.

Earlier this year, I conducted similar searches in AI environments  – ChatGPT, Gemini, Perplexity, Copilot. Perplexity had A LOT of info on me. Like, A LOT. 

But it wasn’t all factual. It was a fabrication, but not a lie. By that I mean, the bio was created by AI using pattern recognition and machine logic. 

Facts about my life tied together by probabilities to create the AI-generated story of Sorilbran. Basically, it was AI’s best guesses: Sorilbran probably… she likely… she could have…

For clarity, the info wasn’t wrong. I think it’s important to emphasize that. AI guessed right. But the fact that the information wasn’t wrong is because there’s already so much info about me available online. For years, I would submit proposals with “Google Me” as my CTA. So, AIs have enough information about me to make some pretty solid guesses. 

Still, I was alarmed by the fact that someone could search for me in Perplexity and read a narrative that was 40% fact, and 60% fabrication presented as fact. So alarmed that I created a video series that I posted to Facebook to give my friends, family, former classmates, and friends of friends a heads up about AI. 

I don’t want machines making guesses about me, especially when it’s surfacing those guesses publicly as fact. 

Takeaway here: If I want the right version of me to show up online, I have to teach the internet who I am. Not let it guess.


Phase 1: Identify Your Why

I had a pretty solid idea of the outcome I wanted to see. The first step was getting specific – if I know why I’m doing it, I can align my outcomes to support my why. So, I got really specific about my reasons.

Reason #1: I’ve worked diligently to hone my expertise, and I want the information the internet shares about me to represent me in the right way. 

Reason #2: As I start sharing more insights and tactics around visibility strategy, it’s important for the information I publish now to be substantiated by my existing digital footprint. 

Reason #3: Having an in-house subject matter expert (SME) boosts the perceived value, credibility, and trustworthiness of a brand. Especially as search and AIs start recommending people the same way they recommend products. As I move into thought leadership, it’s a win-win for the brands I’m tied to, my own included.

Takeaway here: Trust is multi-layered. It hinges not just on what we say today, but the track record we’ve built online. 


Phase 2: Decide What You’lL Measure

I started a project in ChatGPT to track the evolution of my online presence. I watch for: 

  • Incremental shifts in visibility, growth, and engagement on LinkedIn 
  • Visibility in search and AI environments
  • Shifts in how the AIs tell my story after I’ve intentionally planted the right information online  

I kept measurement simple – no dashboards. I use before and after screenshots of bios, my LinkedIn Social Selling Index, and any other metrics I’m tracking. Maverick (my AI assistant) keeps track of metrics and highlights any areas that may need more attention from me.

Takeaway here: Measuring doesn’t have to be fancy dashboards. It can be a before and after split screen. And still tell you what’s working and which areas may need more attention. 


Phase 3: The Messy Middle – Thinking Through How to Author a Digital Story

Cross-platform. Multi-modal. For a multi-hyphenate. 

Historically, I’ve intentionally kept the different parts of my output separate, but AI doesn’t do that. I’m starting to see my personal creative endeavors bleed into my online story as a marketer. 

So, I decided to create a home for not just marketing and AI fluency, but also for music and fiction and family – all the things AI sees about me, but doesn’t have enough information to talk about knowledgeably. These, of course, are things I don’t mind other people knowing about me. 

Positioning

First up was positioning. This was the tough part. Because make it make sense! 

How do you tell the story of a songwriter whose current claim to fame revolves around a role as a Visibility + Pipeline Strategist – who spends her days training robots and building automations to do marketing tasks?

On paper? That sounds chaotic. 

But it’s not that. I’ve lived long enough to be really good at communicating – and it just so happens I’ve mastered a couple of different vehicles over the decades – songs, fiction, B2B articles, essays, books, and now machines.

That’s the story I need AI to tell – the story of a communicator. A storyteller. Not a Jackie of all trades.

Your story – like mine – has to feel connected. Intuitive. People need to say, “Well, it makes sense that you Y because you X.”

Not, “How do you even have time to X if you’re a Yer?”

For me, crafting that narrative took a bunch of long conversations with Maverick. It wasn’t branding. It was… building an ecosystem. I had to figure out how all the pieces of me fit together.

(Cue Ledisi)

The Channels

Once I decided on the  topics I would talk about, I had to figure out where those conversations would live. Not just the different categories or disciplines. But also, the different roles I play within each discipline. 

Take marketing as an example…

As the head of marketing at an agency, I serve as strategist, builder, and implementor. I align departmental KPIs with company goals set by leadership, then develop the strategies to get there, build the systems, and do the things. Traditionally, three separate roles. Now, I handle most of it.

Here’s the thing about putting out content: What I focus on and talk about as a B2B writer is different from what I’d talk about as an SEM-focused visibility strategist. And in that role, I have a different focus than I do when I’m building automations. So, where and when do I talk about the stuff I build vs the stuff I write vs the research I’ve come across vs designing a character from a story I’m working on? 

In the olden days, you could pick a lane and stick with it. But in the AI age, if AI spots something out in the wild and you don’t either erase that lane or build it up, AI will hallucinate – it’ll fabricate a story about why and how you probably do a thing. Probabilities. 

So, I had to find a way to strategically organize the different parts of me to make each part seem closely connected to the other parts. 

That was tough to even wrap my mind around. But I landed in a metaphor that actually helped me pull it all together in my mind so that I could implement it:

Think of your public-facing persona as a mall, with each channel where you show up being a mall door. 

  • Some people come in through the food court door.
  • Some through the high end department store
  • Some enter through the doors of a discount retailer
  • Some through the main entrance

People have different reasons for entering the mall where they do. Regardless of their destination once inside, they’re in the mall. And if they find what they need or get some sort of value from their visit, they’ll come back. 

The Modes

Writing is my strength. So, I started there. I’ve been adding video. Slowly. No home YouTube studio  to speak of – just me with crappy lighting trying to drop gems. I can improve the experience as I go. The goal right now is just to get things out. Train the friggin data. 

My workflow right now is: 

Industry insights as input -> filter through my experiential wisdom -> record a detailed voice note -> hand that voice note over to Maverick for a first draft -> refine and finish the article myself -> repurpose as video or audio. 

The Audience

Next, who do I want to reach? 

Is my target audience the same as my ICP? Not at all. Many of the people I want to help can’t afford me. Brands with interesting problems can. I have a couple of target audiences, it turns out.

Peers. Water cooler talk – content I make specifically for other builders. I have to actively resist the urge to just hang out on LinkedIn reading comment threads about some interesting research or some experiment other marketers have done.

Startups. I want to reach founders, slim teams who need a GTM workflow that will allow them to stay lean until they can afford to scale strategically, startups gearing up for funding rounds. 

Solos. But also working moms, consultants, contractors. Folks like me who get pulled by home and work at the same time.

Emerging. Under-resourced, under-visible micro businesses headed up by folks from socially and economically disadvantaged demographics. The tech available today makes it possible for them to remain competitive, even as their industries change. They’re small enough to be agile and I can share strategies with them that’ll keep their costs down.

MLK Blvd. My peeps. I wanna be able to share what I know with folks like me who moved from the west side to 8 Mile and Van Dyke to be closer Woodward and closer to church. 

Community. Stakeholders in organizations who are touching more people than I am, and want support putting together programs, workshops, and opportunities for all of the other groups I mentioned to accelerate their learning and earning.

I have different audiences. And I needed to find a way to tell other people what I’ve learned in a way that’s practical but also useful and actionable to them specifically

So…

I built multiple micro sites, I publish  multiple newsletters, and I started putting out social media content.

And then, of course, there are the machines – AIs and search engines. Gotta train the data after all. So, some of what I create is so that I can leave breadcrumbs around the internet for machines to follow. 


Progress: So Far, So Good

Progress, but also setbacks. I’m seeing progress. The AIs update quickly, so my recent overhaul of my personal website is affecting the narrative right now. To make room for music, fiction, and non-work writing, I tweaked my website to include all four of my primary pillars – Marketing Ops, Essays (You vs You), AI Fluency, and Creative Works. Only, I haven’t dropped any thought leadership content on the site yet (Marketing Ops or IP + Creative Works). But I changed the language on my home page and my About page to tell the story of a storyteller. 

One of the algorithms connected me to the topic of hair. Not sure why. But it’s not unusual for Google’s algos to tie me to hair topics. I’m a black woman who wears vastly different hairstyles. So, as ridiculous as that may sound… sometimes hair shows up as a category with which I’m aligned. 

[She shrugs and walks off.] 

Frameworks. Not up yet, but I’ll prioritize those over the next few days as I build out the library that’s going to host the frameworks. I’ll put walk-thrus of my frameworks on my main hub – the property AIs tend to prioritize when it comes to digging up info about me. And make them available to buy directly from my framework library.

LinkedIn is doing okay. My SSI score is holding steady. Still need to prioritize relationship building. New subscribers to both newsletters increases with every new issue. I’ve experimented with boosting individual newsletter editions for a couple of days. Vanity metrics. I’m better off just leaning in organically for awareness. But I’ll circle back to paid for downloadable assets. 

Here’s what this tells me:  The machines are learning what I’m feeding them. When I drop breadcrumbs in one place – say, a newsletter or a quote in someone else’s article – it often surfaces in a different AI environment entirely. That means the networks are porous. Which also means: I don’t need to publish everywhere, but I do need to seed strategically in places that cross-pollinate.

Lightweight syndication. But also: traceable influence.

I’m sure I can accelerate this with LinkedIn video. Right now, my only video is going to Instagram and it’s about me personally and music. So, I think in the eyes of the machines, I’m prioritizing the storyteller. 

Next steps

Lean into building out content around Marketing Ops and AI frameworks before circling back to songwriter and introducing the Tuesday Universe (my fiction stuff). I need to begin curating that crossover between my strategy brain and my storytelling world so that AI doesn’t have to invent a bridge. It can just follow mine.


Why This Matters

Most people think visibility is about getting seen. But it’s really about being recognized when you are seen.  And those are not the same thing. If you’ve ever had to introduce yourself to the same person more than once, you can probably attest to the fact that visibility is not the same thing as recognition. Visibility is about showing up. Recognition is about being understood. The right version of you showing up, in the right context.

AI doesn’t just echo what you’ve said. It builds a version of you from what it can find, verify, and connect. And that version is what gets surfaced in queries, in search, in summaries.

I don’t wanna go viral. I want to be the person the system recommends when the head of a community program is planning workshops for new businesses in the area and AI comes up. I need the machines to know that powering lean small businesses with AI – my whole jam. 

Being visible doesn’t automatically mean AI can connect the dots. It doesn’t mean it can distinguish what’s core to who I am from what’s just ambient data in the feed. So, my job is to know how AI sees me and seed the right data so that AI can connect the dots. 

I know. Seems like a whole project all by itself. But if you do the work on the front end, you can automate your outcomes. You can train heavy and run light. 

It’s more authorship than branding in the beginning. One post, one article, one framework at a time. Every piece is a connection point. Truth is, you don’t need to be everywhere. Just make sure the places you are showing up have been authored intentionally so that when the machines find you, they recognize you.


Sorilbran in her office edited resized

About Sorilbran

Thinker. Writer. Strategist.

Out here in these digital streets trying to be a good human, write the next backyard barbecue anthem,
and keep a finger on the pulse of innovation.

All while training my girls – and the data – in real time.