Results Require Foundation
Patterns I've noticed while navigating hype to find ROI
You may have seen this video circulating on social media platforms titled “Your Company’s new Agentic AI Workflow”
Mixed Messaging
The internet is full of engagement-bait claims:
“My AI workflow made me $100k.”
“My AI content got 25,000,000 views.”
“My AI messages booked me 50 meetings.”
On the other side, it’s not uncommon to here a different perspective from successful people who aren’t selling an AI course or tool:
“I’m not getting results from it.”
“The work it does just isn’t good.”
“I’m not sure if it’s even positive-ROI.”
There is a massive chasm between hype and real results. And more people are starting to see it.
My Entry Into This Industry
When I first got into the AI consulting world, I had this nagging question I couldn’t shake: where exactly are businesses actually getting results?
The answer, honestly, was mixed. Some saw identifiable wins. Some were more ambigous. And most were doing all the “right” things. They’d watched the tutorials. They’d bought the tools. They’d hired consultants and run projects.
So what was going wrong?
I spent months talking to businesses investing in their AI initiatives watching how teams interacted with the tools they’d been given. I also have met and discussed AI use in businesses with some of the top AI Agencies and consultancies in the world.
I’ve noticed a pattern in the businesses getting results that isn’t being acknowledged enough: The Upfront Work of AI.
AI Agencies Don’t Want To Talk About it
The upfront work is the hard part. It’s the necessary building that sets the stage for everything that comes thereafter.
Building context. Developing real competency. Connecting AI to the tools that actually move the business forward. That’s the work. And it’s a heavy lift.
Especially if you’re new to this, or if you’re a leader with a thousand other priorities. Or if your team is already stretched thin.
The last thing anyone wants to hear is “before AI can help, you need to do more work.” But that’s the truth.
The businesses getting results have taken the time to teach AI how they work. Not just what tasks to do, but how they think. Their priorities. Their decision frameworks. Their voice. The tools they use and why.
The businesses not getting results had skipped that step, or at least not given it their all. They expected AI to figure it out. And AI doesn’t figure it out. AI guesses. And guessing gets you generic outputs that miss the mark.
Three Repeating Patterns
After enough conversations, I started to notice the same failures showing up again and again.
1. Tool-first thinking. Someone hears AI is hot, signs up for five different apps, and tries to make them do something useful. Six months later, they’ve got a handful of subscriptions nobody uses consistently and a growing suspicion that the whole thing was overhyped.
2. Insufficient context given. The AI has no idea who the business is, what they value, how they communicate, or what their priorities are. So it produces generic outputs. The user gets frustrated. They assume the tool is broken or AI just “isn’t there yet.”
3. No connection to real work. The AI is used for surface-level tasks: rewriting emails, summarizing articles, generating social posts. Fine. But it’s another thing entirely to attach AI to things that actually move the business. It stays a novelty instead of becoming infrastructure.
What Actually Works
When we started Infin8 we spent a lot of time doing one off automation or AI tool builds. As time went on, we decided we weren’t going to work like that anymore. We weren’t going to sell consulting reports or tools. We were going to build foundation.
Foundation means three things: Context. Connections. Workflows.
Context is teaching the AI who you are. Your mission. Your values. Your voice. Your team. What you sell, who you serve, and how you operate. Instead of vague “we value quality” stuff, it’s specific enough that the AI can make decisions on your behalf without asking you every time.
Connections are the wires between AI and your actual business tools. Your calendar. Your CRM. Your project management system. When AI knows what’s happening in your business in real time, it stops guessing and starts being useful.
Workflows are the recurring tasks that now run through the OS. Morning briefs. Meeting follow-ups. Weekly reviews. The stuff that used to eat your hours but can now happen while you’re doing other things.
When those three layers are configured, AI stops being a novelty. It becomes an operating system. And operating systems compound.
Then Why The Hype?
I don’t think the people posting “my AI workflow made me $100k” are lying. Some are, sure. But most aren’t.
The fact is that those who are winning the most have done the work. They’ve built context over months or years. They’ve connected their tools. They’ve developed workflows that actually leverage what AI can do.
Using ChatGPT for 10 minutes a day and calling it transformative isn’t good enough. What’s needed is infrastructure. And infrastructure produces results.
The hype comes from people who’ve done the foundation work talking to people who haven’t. The results look magical because the work that created them is invisible.
The Uncomfortable Truth
Here’s what I’ll say to anyone frustrated by the gap between AI promises and AI reality:
The upfront work can be a heavy lift. Building context, developing competency, connecting AI to the tools that impact your business. It all takes time. It takes intention. And it takes someone willing to guide you through it.
But if you do that work and have a vision for the future, AI absolutely delivers: ROI, good results and work your team can be proud of.
It doesn’t take the next set of secret tools. It takes the foundation work that gets less talk because it doesn’t fit in a viral post.
Until next time,
Marshall