Also True for Humans

Seven Things I've Learned Getting Companies to Use AI

A playbook from the front lines of AI consulting

This post was originally a tweet thread in response to Sam Parr asking how people get their teams to adopt Claude. It touched a nerve, so I wanted to expand on it. I recently joined Every Consulting as the head of tech consulting, where we work with mid-to-large-sized companies on AI training and adoption. Here's what's working.—Mike Taylor

Was this newsletter forwarded to you? Sign up to get it in your inbox.

Entrepreneur Sam Parr asked a question on X the other day: "How is everyone getting team adoption for Claude? I spent a lot of time on Twitter, as do you. We see all this AI stuff popping up. We're on top of it, or at least sorta. But how are all you people getting your team to actually use it effectively without spending all their time on Twitter and learning?"

I hear this question in some form on every single consulting engagement. I know the advice I have resonates in meetings, but I'm short on time. So I dictated this post through Monologue and used Claude to shape it into something readable. (Let me know if this format works for you.)

Here are seven learnings from working with companies through Every Consulting:

1. Buy the model direct, not third-party tools

When you evaluate AI-powered tools, you're also—whether you realize it or not—evaluating the tool vendor's choices and constraints, rather than what the underlying model provider (like Anthropic, Google, or OpenAI) is capable of. It's often faster to build your own Claude/Gemini/Codex skill with your own rules and preferences already built in.

Companies are increasingly building, not buying, AI software on top of models, because it gives you flexibility. I don't know how it's possible for companies that aren't the core model providers to keep up when the big labs know what models are coming, build their internal tools to align with those releases, and train them on how to operate within their own environments. I appreciate the effort that companies like Cursor put into user experience—they're a good product organization. But it's difficult to compete with Anthropic offering $5,000 worth of tokens a month for a $200 subscription.

Third-party tools tend to be less flexible, less cutting-edge, and more expensive. That's not always the case, but as a general rule, it holds. So most companies are better off buying directly from the model providers.

Built for businesses ready to scale. Simplify your tech stack with HighLevel.

If your business runs on scattered tools, disconnected automations, and half-built funnels, you are not alone. Most marketers and business owners only scratch the surface of what HighLevel can do. They use a few features, launch a campaign, and never tap into the full power of the platform.

HighLevel was not built to be just another tool in your stack. It was built to run your entire system. With HighLevel, you can capture leads, build high-converting funnels, automate follow-up, manage your pipeline, and deliver seamless client experiences all in one platform.

Inside HighLevel you can:

Capture and convert more leads
Build funnels that actually perform
Automate follow-up and nurture at scale
Manage pipelines and close more deals
Deliver seamless client experiences
Create predictable, repeatable revenue

Everything works together in one place so you can spend less time duct-taping tools together and more time growing your business. Start your 30-day FREE trial of HighLevel and see how powerful your business can be when everything runs on one platform.

Start your 30-day free trial

Want to sponsor Every? Click here.

2. Raise the ceiling, not the floor

A lot of companies have mandated to their employees, "Everyone needs to use AI now. We bought you AI tools. Adopt it." That doesn't work. Even on pain of death, many people are unwilling to use AI or be told that they have to. It's basic self-preservation.

Instead, use the carrot rather than the stick. Nominate people who are already AI-forward as internal cheerleaders. Maybe it gets other people to come out of the woodwork rather than hiding their AI usage by making it clear that using AI is encouraged. Give those people the support they need to unblock barriers to AI usage (typically IT access to data connectors, approved budgets for coding tools, and removal of layers of bureaucracy)—because someone who's bought in is going to accomplish five to 10 times more work than someone who hasn't seen the magic yet.

You can accelerate adoption by showing that people who use AI aggressively get promoted first or interface the most with senior management. In some cases, we've co-opted those early adopters into being teaching assistants in courses we teach to the rest of the team. When their colleagues see that person advancing in their career, that's a more effective motivator than any mandate.

You also get the productivity boost of enabling someone who's already a believer. It's much harder to convince someone to believe than it is to supercharge someone who already does.

3. Workshops should be at least 50 percent build time

Workshops teaching people how to use AI in a hands-on way are an effective way to teach your team—but they need to be heavy on building tools. No one wants to sit on Zoom and just look at slides.

I learned AI by doing. Guided theory helps orient and motivate people, but the biggest complaint we hear is that they don't have time in their workday to explore these tools and learn something new. If you give them a couple of hours in a workshop where they're expected to build something, and access to the tool and data (either synthetic or actual through connectors like MCPs), that's when the aha moment happens.

4. Assign impossible tasks

An "impossible task" is one that wouldn't have been possible to do without AI. Boris Cherny, a creator of Claude Code, has said something similar—that you should slightly under-resource most teams, which makes employees think, "The only way I can do this is if I use AI."

I think it works better if you are more explicit and strategically choose the tasks so that they can't possibly be done without AI. For example, if your goal is to write one blog post a week, you can likely do that manually. But if your goal is to write one a day, you'll probably need to use AI in research, drafting, and editing (like we're doing here!). And you don't set the goal as, "Starting today, you have to produce one piece a day." Instead, say: "Our goal is to work up to producing one piece a day. What needs to happen for you to make progress toward that goal?" It might take time, but if they know that's where they're heading rather than where they're starting, they'll start thinking strategically about how to use AI to save time, and start experimenting.

5. Mandatory AI note-taking plus MCP connector

Everyone on our consulting team records every meeting with Granola and has the Granola MCP set up in Claude Code, and it's been transformative. You finish a meeting with a potential client, and tell Claude to summarize it and send an email to your colleague. That's 80 to 90 percent of the value of AI: extracting information from unstructured data and structuring it in a way that's useful.

So many times I've come to a task and realized I need context from a meeting, and I can pull that information from the MCP. It's how I create curriculum or put together proposals. Now I can't imagine working without it.

6. Map workflows and systematically automate them

When we do discovery calls with clients about their day-to-day work, we follow a process: We ask them what tools they use, what they do on a daily basis, and what their pain points are. Then we put that information into a Google Sheet with a row for each task we need to solve for, and we systematically work down that list as we automate.

Our goal is to get to the point where nobody on the team ever has to do the same task thrice. If AI can take a first pass at each task type, and we build a skill for each one, that person could be doing five to 10 times more than they're doing right now.

So far, in my experience, this has never led to a reduction in workforce. Instead, either the companies put more effort into each task, or they expand the revenue and throughput of their team without hiring. When we were previously teaching Claude Code workshop-style courses, we used to prepare one project for the whole group to work on. Maybe we could manage one per business unit or team, but the preparation cost quickly added up. Now we can use Claude Code to create an individual project for each person taking part. We're using AI to make each engagement that much more valuable rather than cutting headcount.

7. Train people to be managers of agents

Everyone who was an individual contributor before is now a manager—of AI tools. And they're struggling because they don't have management training. They're not used to context switching, setting up systems and rules, or evaluating whether something that they haven't worked on themselves is any good.

Managers can often adapt to managing AI tools more readily because they don't care how a problem is solved—they just want it solved to their specifications. But the script is flipping: Managers are becoming individual contributors, because managing a team of agents is often easier than managing human teams. It takes a human longer to process reams of information, and to see if they'll be successful. Sometimes it's easier as a manager to vibe code a task using Claude Cowork than it is to brief a human, wait for them to send it to their own Claude instance, and get a response in a couple of days.

The upshot is that companies need more management training. You need to help people understand context switching and teach them how to do evals, develop good taste for deciding what to work on, and train AI in specific skills. How do you systematically write a good PowerPoint skill or a good daily update report skill? That's the work now.

If any of this resonates and you want help implementing it, check out Every Consulting. We've been doing this for a year with a select group of companies and are now open publicly.

Mike Taylor is the head of tech consulting at Every and a co-author of the O'Reilly-published Prompt Engineering for Generative AI. You can follow him on X at @hammer_mt and on LinkedIn. To read more essays like this, subscribe to Every, and follow us on X at @every and on LinkedIn.

We build AI tools for readers like you. Write brilliantly with Spiral. Organize files automatically with Sparkle. Deliver yourself from email with Cora. Dictate effortlessly with Monologue. Collaborate with agents on documents with Proof.