Why the AI Hype Misses the Mark: 3 GTM Hot Takes with Abdul Habboub

AI is making it possible to create and communicate at a scale we’ve never seen before. But does more equal better?

Ross Rich, CEO and Co-Founder of Accord, recently sat down with Abdul Habboub, Senior Director of Sales Operations at Vena, to talk about exactly that. 

As Abdul says, “Cool stuff didn’t get us to where we’re at. ROI got us to where we’re at.”

So, before you add another AI tool or hand another process over to an LLM, here are three hot takes worth considering.

Hot Take #1: Stop Sprinkling AI on Bad Processes

“Where can we add AI to our processes?” asks every GTM organization today.

That’s the wrong question.

Abdul flips the question: “If you were to build this company today, would you build that process and then sprinkle some AI on it? Or would you design the process with AI in mind from the beginning?” 

Because making a bad process faster doesn’t suddenly make it a good process. So, before automating something, look at the process itself, and ask yourself: 

• Why does it exist? 
• Which parts require human touch? 
• Which parts could AI handle better? 
• And, perhaps most importantly, would you build it the same way if you were starting from scratch today?

If you wouldn’t build it the same way if you were starting from scratch today, don’t automate it (at least not yet). Fix the process first. Then decide if using AI adds values, or just adds AI. 

Hot Take #2: Don’t Dump Your Data Into AI & Expect it to Understand Your Biz

A lot of executive teams are dumping their data into an LLM and asking it what to do next. 

In a perfect world, this would work great because AI would understand the context, fill in the gaps, and know which data to trust. But it doesn’t. And expecting it to guide decision making without that context is a big risk.

Here are a few examples where AI can really screw up: 

• Flagging a territory as underperforming without knowing it was recently reassigned
• Prioritizing the wrong deals because reps aren't consistently updating the CRM
• Writing off a rep or segment without knowing about one-off circumstances
• Misreading sales calls because it doesn't know what your team considers a strong champion
• Treating an exception like a trend and recommending changes based on it
• Mistaking correlation for causation and confidently recommending the wrong strategy

To avoid making decisions based on incomplete or incorrect information, Abdul recommends starting with context before asking AI to tell you what to do. That means defining what good looks like, which signals actually matter, how your sales process works, and what information can (and can’t) be trusted.

Hot Take #3: Your Prospects are Still Humans & They Want Human Interactions

LinkedIn is littered with AI-generated garbage. So are countless emails, business websites, support chats, and sales messages. And people are tired of it.

And instead of making the interaction more efficient, it makes them trust the company behind it a little less.

“If I like the message, but I know that a human being didn't take the time to write it, I hate it,” Abdul stresses. 

Before a prospect communication goes out, have someone on your team review it and ask: Does this sound like something I would send? Is it relevant? Does it make sense? Would I want to receive it?

Keep in mind, your prospects aren’t evaluating your AI workflow; they’re evaluating your company. And that human interaction not only matters, it’s becoming an advantage for companies willing to preserve it when everyone else is racing to automate it away.

A Word to the Wise, Before You Add More AI…

Remember, AI won't solve all of your problems, and it won't replace the human-to-human interaction that's so important in B2B.

And it also doesn't need to be everywhere.

So, before you add AI to another part of your GTM motion, get really clear on whether it’s necessary or not. Will it add value? Help your team make better decisions? Improve the experience for your buyers?

If the answer is no, you probably don't need more AI. You just need a better process.