Claudeforce and Salesforce: What Sales Leaders Need to Know

Salesforce just made another big move in AI.

On August 26, Salesforce and Anthropic announced Claudeforce, bringing Anthropic's Claude AI together with Salesforce data, workflows, business logic, actions, and governance. The partnership launches with Salesforce in Claude, a plugin with 37 pre-built sales skills that allows sellers to work with live Salesforce revenue data, update pipeline information, and take governed actions directly from Claude. Those skills include things like meeting preparation, deal-health review, and pipeline review. Salesforce in Claude is currently available to select pilot customers, with open beta planned for September 2026.

So what does that mean if you're a sales leader? We get so bombarded with AI announcements that it's tough to know which ones are really significant and which are just run-of-the-mill announcements. I think this one is on the significant side. I think Claudeforce could fundamentally change the way salespeople interact with Salesforce. And if you're a sales leader, there's something I really want you to know: AI doesn't make the fundamentals of your sales process less important. I'd argue it makes them more important.

Salesforce may be changing the way salespeople use Salesforce

For most salespeople, Salesforce has always been a system they have to navigate. You log in, you find the account, you open the opportunity, you look through the history, you update the fields, you add an activity, and you move the opportunity forward. Or at least that's what you want your salespeople to do. Let's be honest. Not everyone is actually doing it.

There's nothing inherently wrong with the traditional Salesforce experience. Salesforce is a powerful system and, when it's designed and used properly, it gives salespeople and managers tremendous visibility into what's happening across the business. But it still requires the salesperson to go into Salesforce and find the information they need.

And this is where I think Claudeforce gets really interesting. Instead of salespeople having to navigate through Salesforce like they've always done to find the information they need, Claudeforce gives them the capability to ask the question directly.

Questions such as: "Help me prepare for my meeting with this customer." "Which opportunities in my pipeline should I be concerned about?" "What's changed in my pipeline since last week?" "Which deals need my attention?"

The AI can pull together the relevant Salesforce information and help the salesperson make sense of it. And that's pretty darn impressive. We're moving from a world where the salesperson has to learn how to use a system to get the information they need to a world where the system can help bring the information directly to the salesperson — and frankly, to you as the sales leader as well. That's a pretty fundamental change.

Here's the catch, though

Now, here's where I think sales leaders need to be careful. All of this sounds great. And it is great. But there's a fundamental question we need to ask:

What information is Claude actually working with?

We've been talking for years about the importance of having good data in Salesforce. Well, that just became a whole lot more important.

Think about it. If your sales stages don't mean anything, how is AI supposed to determine whether an opportunity is healthy? If your salespeople aren't keeping their opportunities up to date, how can AI give you an accurate picture of your pipeline? If next steps aren't being recorded, how can it tell you which opportunities need attention?

This is one of the things we help organizations with when we're designing their opportunity processes. We call them stage gates. A stage gate determines what criteria must exist for an opportunity to move from one stage to the next.

We've always argued that this is paramount. I think the Claudeforce announcement simply reinforces that.

AI needs to know what matters to your organization for a sales opportunity to move from one stage to the next. It needs to know what information should exist, and whether the salesperson has actually captured it.

And that's where I think AI can be really useful. It can do the checking for us. It can validate whether we've captured the information we said we needed. It can identify whether an opportunity actually has the information required to move forward.

But without that playbook, without that framework, AI is going to be doing some guessing on your behalf. And that's where this isn't going to be nearly as useful.

The same thing applies to the Salesforce platform itself. If your Salesforce environment is overly complicated, with too many fields, too many workflows, and too many processes, AI isn't going to magically fix that. In theory, AI may be able to make sense of a complicated system faster than a human can. But it still can't fix empty fields — or worse, fields that have been filled in by salespeople simply because they're there and they're guessing.

Bad data can actually be worse than no data. Garbage in, garbage out.

This is why I think the Claudeforce announcement reinforces something we've been saying for a long time: you can't separate the technology from the sales process it's supporting.

The better your sales process, the better your Salesforce platform, and the better the quality of the information your salespeople are putting into Salesforce, the more valuable AI is going to be.

AI doesn't eliminate the need for discipline.** It makes the consequences of a lack of discipline much more visible.**

This is why Process, Platform and People matter

This is actually where I think the Claudeforce announcement connects very closely to how we think about Salesforce. We talk a lot about three things: **Process, Platform, and People. **And I think AI is going to make all three of them more important.

Let's start with Process.

Before you can expect AI to help you manage your sales process, you need to have a sales process that is actually defined. What does it mean for an opportunity to be qualified? What has to happen before an opportunity can move from one stage to the next? What should a sales manager be looking for when they're reviewing the pipeline? What information actually matters to your sales process?

These aren't technology questions. They're sales management questions.

And the more clearly you define them, the more useful AI becomes.

Then there's the Platform.

Your Salesforce platform is where that process gets operationalized. Your fields, your stages, your workflows, your automation, and your permissions all provide the structure that AI is going to work within.

But I think there's an important distinction here. AI may be able to make sense of your Salesforce platform faster than a human can. That doesn't mean it fixes a poorly designed platform or poor-quality data. If your fields aren't being completed, or they're being completed inaccurately, AI simply has more information to work with — not necessarily better information.

And finally, there's People.

This is the one I think we sometimes forget.

We preach constantly that implementing Salesforce is really implementing change. People are fundamentally not very good at change. Now we're bringing in AI. It's tempting to think, "Well, this is going to make everything easier for our salespeople, so they'll naturally adopt it."

I don't think we should make that assumption.

Salesforce, in theory, was supposed to make the salesperson's job easier than keeping notes and maintaining a paper diary. And yet we all know how much resistance there was to adopting Salesforce. So what makes us think salespeople are automatically going to adapt to an entirely new way of working simply because the technology is more sophisticated?

Yes, AI may make parts of the job easier. But it's still changing what people do and how they do it.

That's why the people side of this is so important. Salespeople need to understand what is changing, why it's changing, what's expected of them, and how their managers are going to support them through that change. AI can identify an opportunity that looks like it's in trouble. It can surface a pipeline issue. It can tell a manager that an opportunity is missing critical information.

But somebody still has to catch it, coach the salesperson, have the conversation with the customer, and make the judgment call. Sales leaders still have to lead.

So what should sales leaders do?

So if you're a sales leader, what should you actually do with all of this? I don't think the answer is to rush out and implement Claudeforce. I think the first thing you should do is take a hard look at your existing sales process and Salesforce environment.

Ask yourself some pretty basic questions.

Do our sales stages actually mean something? Do we have clearly defined stage gates that determine whether an opportunity should move from one stage to the next? Is our pipeline data reliable? Are our salespeople consistently updating their opportunities? Do our managers have a consistent way of reviewing pipeline and coaching opportunities? Are we capturing the information that actually matters to our sales process?

And perhaps the most important question is this: If I gave an AI agent access to my Salesforce today, would I trust what it told me?

That's a pretty good test. Because if the answer is no, I wouldn't start with AI. I'd start by fixing the sales system. Define the process. Simplify the platform. Make sure your people understand what you're asking them to do and why. Then start looking at where AI can take that system and make it better.

That's where I think the real opportunity is.

Don't rush to Claudeforce. Claude-prep your organization.

There's one other point I'd make to any sales leader who's thinking about Claudeforce: you don't have to rush out and get it right away.

This technology is going to be here for a while. If your organization is interested in Claudeforce, I actually think this is a great time to step back and prepare your organization for it.

I would call it Claude-prepping your sales organization.

Before you give an AI agent access to your sales process, make sure you've actually defined that process. Make sure your sales stages mean something. Establish your stage gates. Make sure your managers know what they're supposed to look for when they're reviewing pipeline. Make sure your salespeople understand what information they need to capture and, just as importantly, why it matters.

And that isn't something you accomplish in a couple of weeks.

Getting a management team aligned around a consistent sales process, building the right Salesforce framework, and changing the behaviors of an entire sales organization takes time. In our experience, it's measured in months, not weeks.

That's one of the reasons I think this is such an interesting time for our ProSell™ program. If you're thinking about adopting Claudeforce, the work we're doing with organizations today — defining the sales process, establishing the management system, aligning Salesforce to that process, and helping managers and salespeople adopt the new way of working — is exactly the kind of preparation that can make AI significantly more valuable when you're ready to introduce it.

In other words, don't wait for Claudeforce to arrive before you start preparing for it.

Use the time you have now to build a sales system that AI can actually work with.

The future of Salesforce isn't just about AI

I think Claudeforce is a pretty exciting development. We're moving towards a world where salespeople won't necessarily have to think about Salesforce as a place they need to go to find information. They can increasingly ask questions and have the information come to them. And that's going to change the way people work with Salesforce. No question about it.

But I don't think it changes the fundamentals of what makes a Salesforce implementation successful. If anything, it reinforces them.

You need a sales process that's clearly defined. You need a Salesforce platform that supports that process without creating unnecessary complexity. And you need people who understand the process, know how to use the technology, and are supported by their managers as they adopt new ways of working.

Funny thing is, I could have written that sentence 20 years ago. Just because we're bringing AI into the picture doesn't change the fundamentals. In fact, I think those fundamentals are going to become even more important in the age of AI.

The organizations that get the most value from Claudeforce won't necessarily be the ones that adopt it first. They'll be the ones that have done the work to create a sales system that AI can actually work with.

And that's why I think this announcement is about much more than Claude. It's another reminder that technology alone doesn't improve sales performance.

It's what you build around the technology — your Process, your Platform, and your People — that makes the difference.

Where does your sales organization stand?

If you want a quick read on that, we built a Claudeforce AI Readiness Assessment — twelve questions that score your Process, Platform and People separately, so you can see which of the three is actually holding you back. It takes about three minutes, and it comes with our guide Putting the 37 Sales Skills to Work.