There's a pattern playing out across every industry right now. Teams are experimenting with AI, leadership is excited about the potential, and from the outside, it looks like real progress.
But underneath most of it, not much has actually changed with how the work gets done.
These experiments tend to follow the same one or two arcs.
- Someone on the team vibe codes a dashboard or an app over a weekend. It looks great, but the data feeding it was manually exported from one system, cleaned up in a spreadsheet, and uploaded by hand into another system.
- Or, someone asks an AI agent a question about a client. They get back a generic answer because the agent can only see what's in one tool. It doesn't know what's in the CRM, the project timeline, or the client’s email from last week.
Both versions feel like progress, but they're actually pulling in the wrong direction.
While the dashboard looks like it has live connected data on campaign outcomes, it’s really a front-end for a static spreadsheet with bad data. The answer the agent gives is from a 2022-dated document, not the latest CRM information, and the person it references is no longer at the company.
What AI impact needs
This points to these two things that AI impact really needs, with appropriate guardrails:
- Universal context: the ability to see across every system the business runs on
- Functionality access: the ability to actually act on what it sees
A prototype built in isolation has neither, and every hour spent polishing it, is one hour not spent building the foundation that would make AI actually work.
This creates a dangerous illusion of momentum.
Organizations come out of a full year of experimentation with a collection of disconnected prototypes that are unaware of each other. The company is further from meaningful AI adoption, separating context and functionality even more.
At MERGE we weren't watching this problem from the outside. We were living it.
The challenge of sprawling tools, siloed AI
An agency is a particularly tough environment for AI because almost nothing happens in a single tool or team. One client engagement touches strategy, creative, media, technology, and account management, and every single team is working somewhere different with a different piece of the picture.
Just like any agency, MERGE runs on a sprawling stack of SaaS tools.
We had the same choice every business faces: keep adding AI to individual tools one at a time and hope it adds up to something, or step back and ask what we actually need to build.
We chose the second option. We didn't need another tool; we needed a foundation capable of connecting our existing systems and giving AI the context to work across all of them.
Applied AI Blueprint in practice
To explore what this foundation looked like and needed, we used our AI Garage to vet the technology for security, bias, functionality, and the like.
It’s crucial that we’re responsible with our AI usage and, to further ensure these safeguards, we run everything by our AI SteerCo that includes representatives from all domains. Their primary purpose is to verify our work is ethical, oriented to human-centric design, and incorporates empathy as the highest value.
The result of our experimentation is MERGE One™ .
- It’s a digital operating system that's connected to everything we run on: project management, creative tools, CRM, finance, and more.
- Now, AI agents can work across the full stack rather than being trapped inside a single product.
In practice, having this universal context means AI stops acting as a siloed assistant and starts providing real intelligence.
The vision of this model is guided by these kinds of real use cases:
- For the strategist: They ask a question in a group chat, and an agent pulls the full context including relationship history, active projects, open issues from Jira, Salesforce, and Google Drive. The agent parses the fragmented data across those silos, analyzes the project roadblocks against the historical account data, and instantly compiles a comprehensive brief.
- For the account lead: They need a budget update, and the agent cross-references live media spend in Meta with billing status in NetSuite. The AI actively translates marketing metrics into financial data, comparing the daily ad spend directly against the client's overarching contract, to provide a fully reconciled financial health check on demand.
That same seamless integration applies to the entire creative lifecycle.
Because everything from the brief, assets, brand guidelines, and compliance reviews live in the same connected environment, a creative director doesn't have to wait for an email handoff or ask what's missing.
Yet, visibility is only half the equation; the AI also needs to act.
Custom FDA compliance tool
Take the custom FDA compliance analysis tool we built for our health team.
- Instead of forcing teams to download a file from Workfront Proof and re-upload for a detached review, we embedded the tool directly into MERGE One™ .
- It pulls the asset, reviews it against regulatory guidelines, and acts as an approver right there in the flow of work.
That is the functionality layer at work.
These apps succeed because they are built directly on top of the foundation, not next to it. The AI isn't an isolated chatbot reading from the sidelines; it has the authorization to execute steps right where the work happens.
What we’re seeing
When AI eliminates the mechanical friction like the exporting, reconciling, and "can someone send me the latest version" messages, context no longer has to live in people's heads or scattered across a dozen tools.
Your teams are finally free to focus on what makes work uniquely human.
Strategy. Creative thinking. Client relationships. The ability to look at something and know whether it's right for this brand at this moment.
We didn't set out to replace anyone's job. We set out to give people their real jobs back.
Get in touch
The AI advisory space is crowded with consultants theorizing about what might work. We took a different path.
We built MERGE One™ because we had to solve the fragmentation in our own growing operations first. We did the hard work of establishing the governance, building the architecture, and figuring out how to move AI from scattered experiments into actual production.
We turned those learnings into the Applied AI Blueprint, part of our Humanity Suite, to help clients solve the challenge of activating AI with full context and functionality across their business.
Are you ready to evolve from managing disconnected AI experiments to working intelligence on a unified system that delivers impact? Let's talk.