The Real State Of AI - An On-The-Ground Look

8 Mar

I created “The Real State Of On The Ground AI" report, as many leadership conversations were around the same themes, feeling behind, seeing many completed automations but struggling to make them work and competitors appearing to be leaps ahead in the AI race, but hard to see any real significant progress away from performance ads...

I provide an assessment of the current state of AI adoption, future trends, and offer strategic actions for leaders.

I have made this multi-format: there is a video walkthrough, the deck to click through, and you can download the deck to share with colleagues, available to download here.

The report's main topics include:

The Current State of AI: Where companies and customers are really at in their AI journey.
Customer Usage: What customers are truly using AI for.
- For instance, I asked 100 C-Suite leaders where they are currently with AI
  Just Starting - 40%
  Using Tools - 42%
  Automations - 17%
  Agentic - 1%
- COOs suggested they were the most advanced - startups suggested they were at the automation phase, some are, some really aren’t.
Fears and Risks: The concerns and risks currently playing out.
The Near Future: The importance of really understanding the AI landscape and why AI hackathons are critical to internal success.
The 3T's of AI: Why Time, Trust, and Truth are critically important.
Phases of AI: The move from "Ask & Answer" Assistant, and then to "Agentic".
The Future: What the AI future looks like and how to build towards the agentic future we are being promised.

The Real State Of On The Ground AI - March 2026

Key Takeaways from the Report:

Adoption is Low: A section survey of 5,000 knowledge workers found that 26% have no work-related AI use case, and 59% of reported use cases are basic task assistance, with only 2% being advanced use cases that benefit the organisation.
The Three Phases of AI: The report outlines three phases of AI adoption:
- Ask & Answer (Today): Think of asking a question and getting an answer. This is where the everyday person is.
- Assistant (Near Future): Think of research leading to a recommended list. This is where some marketers and smart operations people are - but way fewer than most can compare to on LinkedIn and on X demos.
- Agentic (Future): Think of a chat leading to a checkout, or an agent performing research, comparison, negotiation, and purchase.
  This is only for a tiny percentage of early adopters and developers today.
The 3T's of AI: AI use should be considered through the lens of:
- Time: Will AI save you time and energy and help with talent re-pointing?.
- Trust: Do you trust the answers, the platform, and will they handle private data correctly?.
- Truth: Is the output truthful, and can it be validated and verified quickly?.
Focus on Problems, Not Tools: Most companies are focusing on tools (e.g., "What tool should we use?") rather than the problems they need to fix with AI (e.g., "We struggle with analysis - how can we scale this effectively?").
AI is Change Management: AI is as much a shift in change management and operational excellence as it is a technical shift, requiring operational rigour and an understanding of the second and third-order effects. If you do not tackle this like a transformation project your business will operate in large silos and some with just start their AI journey while others are optimising their performance and not in a place to help coach their colleagues.
Map & Build The New User Journeys: User journeys are evolving from a Traditional Journey (Search → Click → Browse → Purchase) to an Assistant-Mediated (Ask AI → Receive Recommendations → Verify → Purchase) and a Fully Agentic one (Agent researches, compares, negotiates, and completes checkout).
Action Items for Leaders: The report provides several action items, including:
- Building a detailed problem list. No problem list = no progress - if you struggle to write down the problems, there will be no progress for the business - individuals will progress but those around them can struggle with adoption and often are not inspired to replicate. Cross-functional change is critical - this is why AI workshops, AI hackathons and dedicated AI training are critical to your success.
- Creating a “Company's Context Document” is my living, breathing recommended document for all teams to use when prompting an LLM.
- Rethinking your Marketing plan is critical, AEO is a new channel not an add on to SEO, you have to adapt to how AI is answering queries (creating new "AEO" or AI Engine Optimisation pages), not just tagging onto the SEO teams’ work - the skills are similar, but the output and design are very different. SEO will be key for many businesses for the foreseeable future.
- Map out current buyer journeys and add in how AI Assistants and Agents will be part of and optimise the experience. This new “UX” has to be designed.

NEED AI HELP? BOOK A CALL NOW

My 6 Steps To Take To Make AI Work For You

1. Build out a detailed problems list, the problems you are facing and where you hope AI can help address these - this should be revisited every few weeks

2. Run an AI workshop (hire an external facilitator like me to add weight) to bring departments or the company together and create a level of understanding and use

3. Budget & Invest - Invest in the right tools, right now. There's no one-size-fits-all tool.
You will need the list of problems and create a list of tools to trial.

4. Complete the recommended action items throughout the deck - including AEO as a new channel recommendation, creating new agentic flows and designing for assistant and agent users.

5. Go deeper with an AI hackathon (I’m always happy to help facilitate), build 2-3 solutions to mid effort mid reward problems and integrate into your existing roadmap. Create prompt libraries and use cases to continually improve adoption - this can be in 1x morning or afternoon session

6. Run monthly check-ins with AI, create an AI roadmap and assign a captain and champion to AI. Captains are rotating people who want to help drive adoption and problem solve, and champions are those who help to train colleagues on use cases

Download Deck