There is no question that generative AI is here to stay. It’s now present in almost every software platform, from Google to Microsoft 365 to Zoom. Reactions to the onslaught of AI tools among the general public is mixed. Some folks love the opportunities and convenience it provides. Some folks are tired of it appearing everywhere they look.
Likewise, the uses of AI are mixed. It can be incredibly powerful, and it can be incredibly dangerous. It can lead to massive productivity gains, or it can create unnecessary complexity and slow you down.
A few weeks ago, I wrote about what exactly AI is. Then, I shared with you some of the best ways to use AI in sales and in business in general. Today, I want to discuss some of the most terrible uses for AI and how to discern where you should (and shouldn’t) use it in your sales organization.
- AI Is Terrible At Strategy
A recent BCG study found that AI is an outstanding tool for creative ideation. The same study found that it is terrible at strategy. If you think about what AI actually is, you can begin to understand why strategy is the wrong task for it. Currently, generative AI does not have the capability to understand broad context or to think strategically about your business. It is simply guessing at potential outputs based on a huge database of information available on the Internet.
Another reason why AI is bad for strategic tasks is that it will, in general, produce a leveling effect among users. If you are using an AI output to determine your strategy, so are a dozen similar companies. AI is not capable of thinking outside the box, and as a result will generate the same essential strategy output for all of the dozen companies. It’s a fast track to sameness.
In sales, this means that while you might use an AI to generate ideas for questions to ask a customer (creative ideation), you need a human to plan the sales strategy and approach, both at the overall level and at the individual opportunity level.
- AI Is Terrible At Helping Customers Improve
I’m going to get some flak for this one, but AI is simply terrible at customer service. It can be useful as a customer support chatbot to answer very simple questions based on a carefully curated data set, but it cannot be trusted to connect with a customer’s unique needs, understand their problem, or help them reach a resolution.
This ties back to the same problem in #1, which is that it simply cannot understand larger contexts outside of its limited dataset, and it has no context for understanding and connecting with the customer. For any problem larger than an FAQ, AI is the wrong tool and can absolutely crater your customer satisfaction.
In sales, this means you shouldn’t trust AI to answer potential customer questions, either. AI is known to “hallucinate”: That is, to produce patently wrong but strongly confident responses to questions. In a sales context, this can mean that your chatbot could give a customer very wrong information about your product, leading them either not to make the purchase or to make the purchase and feel betrayed when it turns out your product cannot fulfill the expectation your AI bot gave them.
Plus, the last place you want a customer to feel that they haven’t been seen and heard is at the sales stage. AI bots are great at answering questions about limited topics and terrible at connecting with genuine human needs.
- AI Is Terrible At Key Content Creation
AI is great at some kinds of content, such as podcast summaries, white papers, conversations, and long-form content. It can also be good at producing social media content based on longer content. However, you shouldn’t use AI to generate your key content, such as thought leadership and in-depth analysis.
Generative AI is good at amalgamating what’s already been said and saying it again in a new way (shorter, longer, funnier, more professionally, more casually, etc. etc. etc.). It is not capable (yet?) of original thought. And even if it were, do you want to delegate your differentiating value to something outside of your own intelligence?
You don’t want your key content to look like everyone else’s. It should be strategic and laser-focused on the differentiating value you bring to the conversation. AI can’t do this for you. You need human intelligence partnered with your strategy to produce content that will effectively enable your sales.
Once you have the key content created, AI can be enlisted to produce different versions of it or extrapolate from it to create social media posts and summaries.
- AI Is Hopeless At Cleaning Up Bad Data
AI has no means to judge whether your data is good or not. If you slap an AI engine over a bad dataset, you’re going to get bad outputs. Garbage in, garbage out. Don’t rely on AI to clean up your sales data.
- AI Is Unable to Replace Human Intelligence
There may come a day when some form of Artificial Intelligence is capable of matching the human brain for true intelligence, but that day has not come yet. What passes for “Artificial Intelligence” at present is actually just highly sophisticated algorithms matched with enormous datasets. Generative language AI basically just guesses at each next word. It’s not thinking, not in a human sense.
In sales, you need humans figuring out human problems. You need humans on the phone with prospects to help them identify their priorities, align them with their needs, and then find the solutions that are best for them. You need humans envisioning what your company can bring to the world and how to do it. AI is many things, but it is not human, so don’t try to replace humans where you need humans.
How To Discern When to Use an AI Tool in Sales
With the market flooded with AI tools for sales, it can be hard to know when you should and shouldn’t use AI. Here are a couple of keys that can help you:
- Think about who will benefit from your use of it
Hubspot has famously and popularly delivered a tool that enables marketing teams to quickly generate content on any topic. This sounds great for inbound lead generation, but if you look deeper, you’ll find that their AI tools have also been trained to link to Hubspot-owned content. So, while it may look like your company is benefiting, in the end, what it’s doing is driving traffic to Hubspot. This doesn’t mean you can’t use the tool, but it does mean you need to pay attention and have a human overseeing what is produced and how it will impact your company, including what links it’s generating.
- Recognize that all AI is not created equal
Some implementations are better than others. Don’t trust it blindly. Treat it like you do every other technology implementation. Think strategically about what you are trying to accomplish, identify your options, and then test tools before you launch them. Yes, even AI tools.
AI is a powerful tool when it’s used well and a dangerous one when it’s implemented poorly. What are some of the worst ways you’ve seen AI used lately?