Bart Dirksen's Playbook for Using AI to Scale Your Partner Desk

AI can take the grind out of partner management. It can also scale your mistakes. Here's how to tell the difference.

Every partner leader has felt the pull by now. Prompt this, build an agent for that, automate the outbound. Bart Dirksen sees that excitement at almost every company he works with. He also sees what usually comes next: AI layered on top of a partner program that was never built to feed it.

Dirksen is the Owner of PartnerImpact, a partnerships advisory practice, and the author of Finding Traction in Partnerships: A Field Guide for Executives. He has spent more than 20 years in tech, starting as a BDR and moving through sales, consultancy, sales management, partner management and partner leadership across MarTech, commerce and SaaS companies. He calls partner management "the best position to be in," and also "a very hard position to be in."

His view on AI is practical: a real way to scale a partner organization, but not a silver bullet. The difference is the work you do before you switch anything on.

Partnerships is a company commitment, not a department

Before talking about tools, Dirksen starts with why companies run partner programs at all. Too often, the answer is coverage.

"Many companies use it as being coverage, right? Let's get a partner program running. Let's get partners on board. Let's leverage their sales organization to do the selling, which is probably the worst reason to do partnerships in general, but it happens." — Bart Dirksen, Owner, PartnerImpact

The better approach, he argues, is teaming up with companies in your space to deliver holistic solutions, which benefits your organization, your partners and your clients at the same time. That changes who owns the motion.

"Partnerships isn't just a partnership department motion, it's a company commitment to generating a different kind of go-to-market motion as a company, which means it's not a one-man job."

That commitment carries real operational weight: partners need to be pruned, onboarded, trained and supported. Dirksen is candid that in many companies, partnerships is still "under-resourced, undervalued," and often "a bolt-on into the revenue organization and not truly embedded into go-to-market motions."

This is exactly where AI earns its place. In his words, it is "a very helpful way to scale partner organizations to put more emphasis on the areas that bring actual value, and reduce some of the overhead that isn't necessarily producing a lot of value, but is necessary to run a program."

Strategy stays human. The legwork doesn't have to.

Dirksen draws a clear line between the parts of building a partner program that AI should touch and the parts it shouldn't.

The strategic questions come first: What's our value prop? What's our ideal partner profile? How do we go to market with partners? AI can be a sounding board, but he expects this work to stay largely manual because it has to reflect the company's culture and strategy.

"That doesn't fit with kind of just prompting something and then expecting the entire company strategy to come out of an AI engine."

The follow-up work is different. Finding partners who fit your ideal partner profile and getting them onboarded used to mean hiring partner managers to do outbound and networking. Once the strategy is defined, AI can do much of that pre-work: researching which partners are active in your space, which ones could add value to your platform or help service your clients, and mapping where you already share customers, whether in the US or Europe.

The upside is speed. You can onboard more partners, faster. But Dirksen flags the trap that comes with it.

"Not all follow-up is equally suitable to do with AI. So you want to prevent that you can onboard hundreds of partners and then not really service them well and not really helping them to be successful."

And underneath that is the problem he sees most often.

"The most common challenge I see with companies is they think AI will accelerate their program while they really haven't built the foundational elements that would feed into AI to be actually valuable in their go-to-market strategy with partners."

Where AI pays off on the partner desk today

How much value you get depends on how much foundation you've built and how much is documented. With that caveat, Dirksen points to a set of use cases he has seen work in practice, several of which he has built himself:

• Co-sell mapping. Mapping account overlaps, pruning the data and "finding nuggets in data that you couldn't find before."
• Joint value props. Automated processes that produce a first draft of a joint value proposition with a partner. "You still need to prune it and make it better," he says, but it gets the work moving.
• Partner profiles. Building them in an automated way, a task that "usually took up a lot of time."
• Data and reporting. The more sophisticated application, which he has applied at multiple companies.
• Trigger-based outbound. Drafting messages to partners based on what's happening across product development, marketing and go-to-market, so the ecosystem gets valuable updates without a manual push each time.

Reporting comes with the biggest warning. To hook AI up to your data and trust what comes out, you need a clear understanding of how the data is organized, what it means and whether it's accurate.

"It always looks good, but is it accurate and does it really represent what's happening?"

→ WorkSpan: Dirksen's point about trustworthy data is where most partner teams stall. WorkSpan connects partner data to the CRM so co-sell overlaps, partner touches and attribution live in one consistent record. That gives AI something accurate to work from, and gives sellers partner intelligence they can act on.

Don't skip the foundation

If there is one thing Dirksen repeats, it is this. Companies want AI fast, and the unglamorous groundwork gets skipped.

"They forget to really dig into the foundation on data, information management, understanding which partners are in your ecosystem and how they kind of help your organization to grow and go to market."

Co-sell mapping is only as good as the account data behind it, and automated reporting only as credible as the definitions underneath it.

Finding the deals partners already touched

The most concrete payoff Dirksen describes is about visibility. At the companies where he has applied AI, a major use has been looking into systems to find "those nuggets where partners actually contributed to either existing deals or have sourced new deals that weren't captured before," along with deals partners have the potential to influence that neither the partner manager nor the salesperson had spotted.

That changes the partner manager's role in the sales cycle, from reactive to proactive.

"As a partner manager, you can become much more proactive to be part of the entire sales cycle and either help sourcing deals, influencing deals, and closing those deals that you didn't find before."

The result, he says, is an uptick in sourced and influenced revenue, a bigger visible contribution to pipeline and closed-won, and a more important seat in the go-to-market motion. For an often undervalued function, that matters.

He also sees AI helping teams scan the market for new partners that close white space gaps they would otherwise miss.

Automate the work that doesn't scale, keep a human in the loop

For Dirksen, partnerships is always revenue related. The metrics he watches are closed-won influenced and sourced deals, sourcing, and attach rates, measured across the customer bow tie, meaning both new deals and expansion of existing ones. (He publishes a full schema of partner metrics on the PartnerImpact website.)

So how do you decide what to hand to AI? Look for the components that eat time without producing much value, or that are very hard to scale as a person: scanning data, finding new opportunities, scanning the market, generating first drafts, and automating some partner outbound and updates.

But automation doesn't mean walking away.

"I do feel in many of those cases, a human needs to be in the loop. I see too many examples of people just using AI and not really pruning what comes out of it, and then it becomes very inauthentic."

The through-line

Dirksen doesn't treat AI as a tool for a single task. It should run through the entire partner process, helping the team speed up and scale. But it amplifies whatever is already there.

"If the foundation is wrong, then you're gonna scale the wrong things. So make sure that you put in the effort on the foundational components, and then use AI to compound and grow and scale your organization as a whole."

The order matters. Get company commitment. Define your value prop and ideal partner profile. Clean up your data. Then let AI take on the legwork, with a human reviewing what goes out.

→ WorkSpan: WorkSpan is built for the partner team Dirksen describes: one with a solid foundation that wants to scale. It puts partner plays and intelligence in the seller's flow of work, ties partner attribution to the CRM, and helps teams run co-sell execution at scale, so AI-driven insights turn into sourced and influenced revenue you can actually see.

Bart Dirksen is the Owner of PartnerImpact, a partnerships advisory practice, and the author of Finding Traction in Partnerships: A Field Guide for Executives. He brings more than 20 years of experience across digital tech, partnerships and consulting. Thanks to Bart for sharing his playbook with WorkSpan.

‍