What These Terrifying Images Tell Us About the Limitations of Artificial Intelligence (AI) in Sales

Generative AI is suddenly infiltrating nearly every corner of our digital world. It can write your sales scripts, your marketing copy, your thank you notes, your eighth grader’s research paper; it will make our lives easier, our business tasks infinitely more productive. But before you feel like your laptop is staring you in the face thinking, move over obsolete human, you are soooo 2022, here’s an interesting reality check that says a lot about the importance of real human intervention in our day-to-day tasks---especially sales and marketing.

We recently noticed one of our design platforms had added an AI image generator. Just type in your keywords and the AI will come up with an original, compelling image, with no copyright restrictions. Easy, right?

Well, not quite. It turns out that while AI able to fool us all when it comes to recontextualizing popular images and famous people (such as, the great Pope-in-puffer-jacket fake out), it’s not yet up to par when it comes to dealing with more niche queries. When we used this tool to generate an image of “manufacturing executives,” this is what we got:

Raise your hand if you want to do business with THIS crowd!

Or how about these “energy industry executives”?

Yikes!

Clearly, we’re not looking to use these images for any reason as those are obviously no manufacturing executives that live on this planet. But imagine if there was no human to double check these images. No one to say, “Wait a second, guys, why is Fred’s mustache on his forehead?”

The world would be flooded with terrifying, inaccurate images that serve no purpose.

Just like when you pay a lot for sales leads that promise you the world and use the latest technology to scrape and harvest data off the web. But then you discover that it is filled with errors and inaccuracies—because without human checks in place, technology can only go so far.

To give you an example, we recently checked out our own company profile provided by a competitor who uses automated and AI processes to build what they believe is the world’s largest and most accurate sales database. And do you know what we found? 48% of the contacts listed as working for MNI either no longer work for us or never worked for us, while key decision makers such as our CTO and our Vice President are missing from the profile. It lists our Chief Operating Officer as our Customer Service Manager and also lists a key operations executive who not only doesn’t work for us but works for a company with a similar name – an error a human would see right away but a machine can’t.

As a sales professional, you may be familiar with the frustration of outdated or inaccurate contact data. You dial up Jane from Purchasing, who is really John from Janitorial, and Martha the “CEO,” who was really the comptroller, passed away fifteen years ago. All of your “leads” have led you into a bewildering netherworld of wrong numbers, dead email addresses, fake contacts, and companies that don’t even exist. Suddenly you’re in a twilight zone of bad data, getting ready to dial up this guy:

Just like Brad here is missing half his vest and has an extra eyebrow, your “lead” is missing the right contact name and has an extra letter in an email address, making your important personalized message both pointless and undeliverable.

It doesn’t have to be this way.

While we embrace technology at MNI, implementing our own proprietary technologies for data verification and quality control, we don’t rely on it exclusively. We’ve stuck to our tried-and-true method of contacting each company in our industrial database IndustrySelect directly—meaning real people getting other real people on the phone or via email to personally verify each and every data point we collect on the nation’s industrial companies.

You know, the useful stuff like company address, phone, direct executive email addresses, name and title of executives, key products manufactured, annual sales (we could go on, but we’re talking 30+ data points). So, we discover the new phone number, the executive title change, the new location or product line before you do. You’re back in the real world and all you have to do is grab the data you need and run with it.

How do we collect all those data points on every single manufacturing company in the United States firsthand? First of all, we employ a very large research team. Secondly, we’ve had a century or so to perfect the process so we’re pretty efficient at it. But most importantly, we’re dead serious about it.

In total, we send out more than two million pieces of mail, make one million phone calls, and send three million emails every year to ensure the accuracy and completeness of our company profiles.

So when you have one of our company profiles in front of you, you know you’re not going get something like this:

But something more like this:

See the difference?

Want to learn more about how we collect our data? We’re so proud of it, we made this infographic to show you our entire process. We’re also so confident you’ll love the breadth and accuracy of our data, that we offer a free demo account, loaded with 800 sample profiles, so you can get the full picture of how IndustrySelect can make your life easier—maybe even easier than Bard.