For twenty years, an ERP buying journey started with a search and a referral. Today, more and more of it starts with a question put to an AI assistant: "Which ERP suits a distributor my size?" "Business Central or Wiise for a multi-entity group?" "Who implements Acumatica in Australia?" The buyer arrives at your door with a shortlist, a rough comparison and a price expectation already formed — work they never used to do before speaking to a partner.
It's tempting to read that as a threat to how partners get found. It's actually something more interesting. AI is quietly taking over the first half of the ERP sale — and handing the second half back to you, harder than before.
The half AI takes
Everything at the top of the funnel is becoming machine work. Product comparisons, "ten things to know about cloud ERP," the shortlist, even an opening price anchor — a buyer's AI assistant will assemble all of it instantly, before a human conversation ever happens.
Two consequences follow, and both matter.
First, being found and being informative are being commoditised. The comparison page and the explainer blog — the content moat most partners have spent years building — is now the AI's job, and it does it faster than you can. That advantage is draining.
Second, the buyer arrives later and better armed. By the time they're in front of you, they've done what used to be your early discovery. They have opinions, comparisons and a number in their head — and far less patience for a generic needs analysis. They have a sharper eye, too, for whether you actually understand their situation or are just running a script.
So the value migrates. If AI handles the meeting-getting, then deal-winning is all that's left to compete on — and that's the complex, human, high-trust work a model can't do for the buyer.
The half AI hands back
Here's the uncomfortable part. The second half of the ERP sale — the part that's fast becoming a partner's entire competitive advantage — is precisely where win-loss analysis shows partners are weakest.
The patterns are remarkably consistent across the channel:
- Deal-breakers surface too late. The objection that kills the deal — a data-migration expectation, a feature assumption, a budget reality — was usually present early but never surfaced or recorded.
- Stakeholder maps stay too narrow. The deal runs through a single friendly contact while the real decision — often sitting with a parent entity or a head office already moving elsewhere — goes unmapped until it's too late to influence.
- The deal lives in one rep's head. A strong relationship seller who flies solo still leaves winnable deals on the table, because what they know is never captured, coachable or transferable.
- Qualification drifts from the ideal customer. Deals that were never a fit get chased on optimism; price expectations get anchored wrongly; edge cases soak up the resource good-fit deals deserved.
- The cost of doing nothing is never quantified. Without a hard ROI and a clear price for inaction, buyers default to the cheapest decision of all: staying put.
Every one of these was survivable when the top of the funnel did some of the work for you — when volume and being-found covered a multitude of sins in the deal itself. As AI absorbs that top-of-funnel advantage, these stop being occasional leaks and become the whole story. You no longer get to win on being found. You only get to win on the deal.
What this actually means for partners
The instinct will be to answer AI-mediated buying with more AI-mediated marketing — better content, an on-site chatbot, SEO for the machines. Do enough of that to stay legible: make sure an AI can find you and represent accurately what genuinely makes you different. But understand that's table stakes now, not advantage. Everyone will have it.
The advantage is in the half that doesn't scale:
- Assume the buyer arrives pre-informed. Retire the generic discovery script and earn the conversation by going deeper and faster than their AI could — into their operations, their constraints, the things a model can't see.
- Compete where the machine can't. Map the real decision across a group structure. Build trust with a nervous CFO. Quantify value for this business specifically. Make the implementation judgment calls the whole project actually turns on.
- Treat sales discipline as strategy, not hygiene. Documented discovery, mapped stakeholders, surfaced objections and quantified ROI aren't process for its own sake anymore. They are the moat.
The floor and the ceiling
AI is about to raise the floor for everyone in the channel. It will make every partner easier to find and every buyer better informed. That is exactly why the floor is not where deals will be won. The ceiling is — the depth, discipline and trust of the human conversation that AI delivered to your door.
That's the work I spend my time on: helping ERP partners build the sales discipline that turns a well-informed, AI-delivered buyer into a signed deal. AI will get you the meeting. What you do in the room is still entirely up to you.
