From AI hype to pipeline: 5 lessons from building an AI go-to-market

Almost every services and consulting organization I talk to has AI capability: smart people, promising pilots, and a partner or two. Far fewer have a repeatable way to sell it. The gap between “we can do AI” and “we have AI pipeline” is a go-to-market problem, not a technology problem.

I’ve worked on that gap from several angles: building a unified AI go-to-market framework with the Europe West CTO office at EY-Parthenon, expanding AI partnerships at Softura, and teaching generative AI to faculty and staff at Joliet Junior College. Here are five lessons I keep coming back to.

1. Start with the buyer’s problem, not the model

Customers don’t buy “GenAI.” They buy faster claims processing, better citizen services, or fewer security incidents. That’s why sector activation mattered so much in our framework: translating AI capability into the specific outcomes each industry already cares about.

2. One framework, many markets

At EY-Parthenon, the goal was a unified AI go-to-market framework that could be adopted across Western Europe, while letting country-level leads adapt it to their markets. Consistency makes you scalable. Local flexibility makes you credible. You need both.

3. Arm the sellers, or nothing moves

A strategy deck doesn’t create pipeline. Sellers and pre-sales teams do. The work that moved things forward was the reusable stuff: an AI Playbook, sector activation materials, and GTM assets that people could pick up and use the same day. I learned the same lesson at Microsoft, where a compete playbook and a focused “blitz” for about 80 sellers turned strategy into wins.

4. Partner for what you can’t build fast enough

No services firm has every AI capability in-house. At Softura, I built the partner ecosystem strategy, signed and onboarded Vantiq and Palantir, and revitalized the Microsoft partnership by aligning it to high-growth AI offerings. That increased co-sell activity and helped drive a $1.1M net-new professional services partner deal. The right partners don’t just fill gaps. They open doors.

5. Adoption is the product

Teaching generative AI to college faculty and staff reminded me that the biggest barrier to AI value usually isn’t the technology. It’s confidence. People need hands-on, jargon-free time with tools like Microsoft Copilot, NotebookLM, ChatGPT, and Claude, and permission to experiment. Done well, AI can also make work more accessible for people who think and work differently. That’s a benefit we should design for on purpose.

AI value isn’t created when a model works. It’s created when a customer buys it and people actually use it.

If you’re leading a services or consulting organization and want to turn AI capability into pipeline, let’s talk. You can also see how I approach problems like this in my case studies.