By: Drew Williams, Founder & CEO, Sales Playbook Builder
Everyone in sales is “doing AI” right now.
It’s a checkbox. Teams are trying tools. Reps are using ChatGPT here and there. Leaders are talking about it. Someone on your team probably sent a “top 10 prompts” doc last week. Everyone nodded. Then went back to doing things the same way.
But here’s the reality:
Using AI is not the same as implementing AI.
And that’s where most teams are stuck.
I see a lot of sales organizations experimenting with AI… but very few are seeing meaningful impact or ROI from it. Not because the technology isn’t powerful, it is, but because we’re treating it like software we already understand.
We’re used to SaaS tools. You log in. You click around. You figure it out.
AI doesn’t work like that. It’s conversational. It’s contextual. And it only performs as well as you communicate with it.
Here’s the disconnect:
AI doesn’t speak human. And most of us don’t speak AI.
So, what happens?
You ask it to “rewrite this” → it rewrites it.
You ask it for a strategy → it gives you one.
You ask it for ideas → it gives you something that sounds right.
But it’s guessing.
It’s optimizing for something that feels helpful, not something grounded in your actual business, your deals, your buyers.
If you asked a real person to build a sales strategy, what would they do? They’d ask questions, push back, and try to understand your context.
AI won’t do that, unless you force it to. That’s why most teams are “using AI” … but not getting much out of it. They’re experimenting. Not implementing.
And even when teams get slightly better at prompting, AI still sits outside the actual sales workflow. It’s something you open occasionally. Not something embedded into how work gets done.
Opening ChatGPT once in a while isn’t an AI strategy. It’s curiosity.
Real impact comes when AI is built into your process.
From Experimenting to Implementing
So how do you actually make that shift?
Start with your process. Not the tool. Before you test prompts or buy anything new, map out how your sales team actually works today. Not the version in your playbook. The real version.
- What happens day to day?
- Where does time really go?
- Where do deals get stuck?
- Where do reps slow down, skip steps, or just “wing it”?
Write it out. Step by step.
You can do this as a team, whiteboard, Miro, whatever. Then have reps do the same individually. You’ll usually see gaps between what leadership thinks is happening and what’s actually happening. That alone is useful.
Then run a simple traffic light exercise:
- Red: This is a bottleneck. It’s not working. It’s slowing us down or hurting outcomes.
- Yellow: It works… but it’s inconsistent or inefficient.
- Green: This part is solid. Not a priority right now.
This forces a more honest conversation.
Because most teams jump straight to: “What should we use AI for?”
Yet, a better question is: Where are we losing time, consistency, or momentum?
That’s where AI might help. And I say might on purpose.
Because AI isn’t always the answer. Sometimes the issue is process. Sometimes it’s training. Sometimes it’s that no one truly owns that step.
AI is one tool in the toolbox. A powerful one, yes. But still just one tool. Once you identify the red zones, now you can ask a better question: Is this actually a good use case for AI?
What Workflows Should You Start With?
The first question usually I get: “What should we use AI for first?”
The honest answer is: it depends.
- Who you sell to
- How your buyers buy
- Whether your team is inbound, outbound, or a mix
- Your deal size
- Whether you’re focused on new business or growing existing accounts
All of that shapes your sales process. So use your traffic light exercise as your starting point. Begin with red, then look at yellow.
Where most teams go sideways?
Not using AI in a way that improves the outcome. Right now, most teams are just scratching the surface. For example, every team has call recording software. You get transcripts, summaries, and maybe action items.
Yet, most teams stop there. But that one call transcript could be used across multiple parts of your workflow:
- It could shape your follow-up
- It could inform your next meeting
- It could update your CRM properly
- It could help you prepare for objections
- It could be compared against past calls to spot patterns
And that’s before you layer in:
- CRM data
- emails with the account
- internal notes
- manager feedback
- deal context
This is the shift. It’s about how you combine inputs and think through the process.
The simplest way to think about it: Input → Process → Output
- Start with the output. What are you trying to improve?
Better meeting prep. Stronger follow ups. More consistent discovery calls.
- Then define the inputs. What information should go into that?
Last map the process. What would a strong rep actually do manually?
- Review notes. Look at past conversations. Talk to their manager. Think through the account.
Now translate that into how AI supports your process, that’s when the output improves. The goal is not just faster, but better, more aligned, more consistent outputs.
If It’s Not in the Workflow, It Doesn’t Count
This is where most AI efforts fall apart. If it’s not part of the workflow, it doesn’t count. If your team is using AI when they feel like it, you don’t have implementation. You have experimentation.
Let’s say a rep writes 100 emails a week, and AI helps them do it 20-30% faster. Sounds like a win.
But here’s the question: Did that really move the needle? Or did we just make a noncritical task slightly more efficient? This is where teams get misled. They chase small efficiency gains instead of improving the parts of the workflow that drive revenue.
Real implementation looks different. It means:
- AI is tied to a specific step in the process
- It’s used consistently, not occasionally
- It’s expected, not optional
- It improves a defined outcome
Because the goal isn’t to “use AI more.” The goal is to:
- improve conversion
- increase consistency
- speed up sales cycles
- create better execution across the team
If AI sits outside your workflow, something you open here and there, you’re not improving anything. You’re just adding noise.
A Simple 6-Week AI Pilot Plan
If you want a practical AI adoption roadmap for sales, start with a focused pilot. Not a massive transformation. Not a company-wide rollout. Just one or two workflows.
Week 1: Map the process
Map your sales process as a team. Have reps map their individual workflows. Run the red/yellow/green exercise. At the end of this week, you should have 1-2 workflows you want to improve. Not ten.
Week 2: Define how AI will be used
How exactly does AI fit into those workflows? What part of the process does it support? What inputs does it need? What does a good output look like?
You can brainstorm as a team. You can use AI itself to generate ideas. But give it context, your process, your challenges, your deals.
Week 3: Set guardrails
This is where structure comes in.
- When should AI be used?
- Where in the workflow does it sit?
- What inputs are required?
- What does “good” look like?
Also, train it on your world:
- your sales playbook
- your process
- your buyers
- your messaging
This goes back to Input → Process → Output. If the inputs are weak, the output will be weak.
Week 4: Train the team
Not generic AI training. Use case training. Have reps actually practice the workflow. Refine how they prompt. Adjust how they use inputs. This takes a bit of effort upfront, but that’s what makes it repeatable later.
Week 5: Run the pilot
Now you use it in real situations. Have the team commit to using it. Talk about it in sales meetings. What worked, what didn’t, where did people get stuck?
This is where learning happens.
Week 6: Measure impact
Now you look at results. Not in a vague way. In a workflow, specific way.
How Do You Measure Impact?
AI impact should show up in your sales process. Not in usage, not in how often people open a tool. In results.
Start simple. Pick one outcome tied to the workflow you’re improving and track it:
- conversion rate
- deal size
- sales velocity
- meeting quality
Most teams default to volume metrics:
- number of calls
- number of emails
- number of leads
Those matter. But they’re not where the real leverage is. The real leverage is in conversion metrics.
What happens between each step:
- Lead → First meeting
- First meeting → Second meeting
- Second meeting → Proposal
- Proposal → Close
That’s where AI should show up, because improvements there compound.
If you improve one conversion point, even slightly, it impacts everything downstream. That’s how you measure real impact.
Common Mistakes to Avoid
I see these all the time.
Tool sprawl
You identify a problem… and immediately go buy a tool. Now you’ve got multiple tools, no consistency, and no clear view of what’s working.
Shiny object syndrome
New model. New tool. New “this changes everything” post every week. Constant switching kills momentum.
Trying to use tools that “do everything”
They promise everything. They usually do everything… okay. Nothing great.
Playing with AI instead of implementing it
Trying tools isn’t the same as building a system.
No defined workflows
If you don’t define where AI fits, everything becomes inconsistent.
No ownership
If no one owns AI implementation, no one drives it. It becomes optional. And optional doesn’t scale.
The Bottom Line
AI is overwhelming. It’s new. It’s evolving. And most leaders are learning it in real time.
That’s uncomfortable. You’re supposed to have the answers, and this is one area where you probably haven’t yet.
That’s fine. You don’t need to be the expert. You need to be the one who implements it better than everyone else. Don’t try to transform everything. Be surgical, pick one thing (maybe two.) Improve it.
The more you use AI in a structured way, the better it gets. The more context it has, the better your team gets. The more consistent the output becomes and the technology itself keeps improving in the background. AI amplifies what already exists. Strong processes get stronger, weak ones get exposed.
So, the goal isn’t to “add AI.” It’s to implement it in a way that improves how your team actually sells.
Because the teams that win won’t be the ones with the most tools. They’ll be the ones who implement it best.
To figure out exactly where to leverage AI effectively, take our quick, 10-question AI Readiness Assessment. You will receive an immediate readiness score and a curated list of AI tools designed to solve common sales challenges. Find out if your sales organization is ready to transform and get the exact roadmap you need to enhance your sales process and drive results.
