Five reasons AI is overhyped for B2B outreach

If you’re like me, you probably noticed that every software platform started asking you if you want to add some artificial intelligence to the mix. It reminds me of going to a restaurant and being asked if I want pepper on absolutely every dish, including dessert.

I wrote this because one of the most common questions I get asked is, “How are you using AI in your B2B sales outreach?”

If you’re here you’re with an agency or B2B service company, you probably want to build to build a business development process that is repeatable.

There are already enough shiny object and distractions without having AI shoved down your throat at every turn.

When it comes to outreach, AI might possibly play a small part in your process, but here is why it’s mostly an overhyped distraction:

First, what this content is NOT:

1. An existential prediction about the future of AI in our lives

That’s beyond scope for now, and for me, the the jury’s out. I get annoyed with the usual binary of AI as either A. the harbinger of the singularity and/or a robot apocalypse, or B. a passing, disco-like tech fad, a typical argument from anti-technology types. As usual, the truth is probably in the middle.

2. An argument that AI has no utility, or is otherwise overrated

Like most people, I used ChatGPT to edit this post and I’m sure there other great uses I haven’t uncovered yet. There will likely be a generation of entrepreneurs who make their fortunes on AI or areas adjacent to it.

All that said, if you’re conducting B2B sales outreach to build relationships with future clients in 2024, here are the top reasons to steel yourself against the AI hype:

1. Mass skepticism to fake personalization

As of now, AI in a nutshell is limited to looking at publicly available data and regurgitating it back at recipients. While I’m sure there are many salespeople winning with this approach, it will become harder and harder to make work as more players enter the game, and as we all grow more skeptical. Because an approach like, “Hey Bob, I saw that you run Blamco and have been there for five years, that’s awesome!” is so abundant, our walls grow higher to tune it out.

2. Reputation risks outweigh marginal efficiency gains

When it comes to small total addressable markets, provenance and trust matter. While scaled outreach is effective, the cost of over-automating it to a small audience can cause reputation and deliverability damage that outweighs the small upside of saving time.

3. Successful campaigns latch onto what customers are becoming, not what they are right now

This is a like how the best quarterbacks throw the football to where the receiver will be, not where he is when at the moment the ball is thrown.

AIs regurgitate ‘winning’ copy from an infinite who’s who of campaigns, where the variables all differ. This like putting a spoiler on your Chrysler and expecting it to drive as fast as a Lambo (or whatever fast car has a spoiler, sorry I’m not much a car guy). AI makes decisions based on inputs from one dimension - the past. This is a problem even if the data was sourced yesterday, much less 1-2 years ago, as is the case with ChatGPT.

In our experience, successful outreach campaigns, like almost all effective sales and marketing, come from pattern interrupts combined with deep marketing understanding.

4. Market incentives toward hype

As Eric Weinstein says, “We look forward our blind spots so we can do business there.”

In a debatably down economy, AI is one of the few areas enjoying funding and a hiring frenzy. Incentives alone are not enough to affirm or dismiss a trend, but they’re worth attention.

Precious few people understand how AI works on a deep level, which makes it hard to dismiss claims as to what the technology is capable of. No one wants to be the laggard left in the dust - if you’re a Chief Innovation Officer in 2024, you’re leaning into AI or you’re losing your job.

5. The human salesperson bottleneck

Let’s say I’m wrong about all of the above and you launch an AI-driven outreach program that generates many qualified meetings. Sooner or later all of those prospects will have to talk to a human through multiple conversations over a long sales cycle.

Until the day that AI can close a multi-six-figure deal for your complex service, which I think is unlikely in my lifetime, that person will have to vouch for the things that the AI communicated on their behalf. This means the salesperson will have to see and approve everything the AI says, which destroys any efficiency that was gained, or they will have to be ok with whatever random stuff the machine spits out. This is a problem because if you’ve ever led a complex sales process, you know that subtle misunderstandings can lead to blowups and lost deals, sort of like how the micro turn of a freighter can put the vessel hundreds of miles off course.

I suspect that, going beyond sales, this human bottleneck will limit AI’s overall progress.

You might be wondering what’s the better way to conduct scaled outreach in 2024? In our experience, the answer is combining the old with the new - look at the prospects who are already likely to trust and respect you as an authority in your space, and sell where you have an unfair advantage. From there, focus on building relationships, not selling, at least not right away. Finally, don’t make exaggerations or false representations - as a rule, if you wouldn’t sent it someone you met IRL, don’t use it in a leveraged outreach situation.

Over and out,

Dan