AI business use cases: artificial intelligence sales

In the West, our economy is knowledge based; we transfer information, and its causes or effects, for money. Sales is the fundamental way in which we do this.

But what if I told you that software now exists that can do the job of a traditional salesperson thousands of times faster? The vast majority of customer-centric roles are in danger of being replaced... most people just don't know it yet.

You can take advantage of the upcoming artificial intelligence sales boom. In this newsletter, I lay out a roadmap to implementing AI in marketing and sales process to grow an AI sales business practically overnight.

Using artificial intelligence sales in business

Before we get into how to employ artificial intelligence in sales, let's break down the typical 21-st century sales cycle, so we can see where opportunities might arise.

Locating customers

First, you identify potential customers through means like social media, public databases, or just scrolling through Google. The salesperson then reaches out to these leads and starts a conversation in order to qualify them.

Right now, this is a manual process that costs large sales organizations billions of dollars a year in revenue. Here's how you could use artificial intelligence and automation to do the same thing in a quarter of the time:

• Sourcing customers on platforms like LinkedIn? Use Sales Navigator in conjunction with Make and PhantomBuster to automatically export high-quality, laser-focused lead lists to CSV. Include tertiary information, like where they went to school, or what their LinkedIn summary says, so you can use AI to personalize your outreach in the next step.
• Manually searching B2B for businesses on Google? First, learn how to refine your search using the following search operators. This can improve your lead sourcing efficiency by as much as 90%, saving you or your salesperson hours of scrolling through irrelevant information. Then, use a platform like APIFY to automate your search. Use pre-built workflows like the Google Search Results Scraper in conjunction with a Contact Details Scraper to source massive lists of relevant businesses and then find their contact information. Save their niche, headline, or meta description for the next step.

Qualifying leads

Then, you qualify those leads - i.e you clarify whether or not they possess the qualities that make them likely to purchase from you. This usually takes the form of a few questions, and your lead's answers determine whether you continue moving them through your sales pipeline.

Despite it's importance, lead qualification is one of the lengthiest and most expensive portions of the sales pipeline. Response rates are abysmally low, and it's usually because the qualification stage is heavily templated (and the lead picks up on that).

Here's how you'd increase your success rate by an order of magnitude in moments:

• Decide on your contact method. Most leads are reached out to either directly on platforms like LinkedIn, or via email.
• Using the personalized information from the previous step (whether that's a LinkedIn summary, a company someone worked at, or a meta description from Google), call a natural language API like OpenAI or AI21Labs with a prompt that looks similar to this:

• This is much better than templated, boilerplate emails with no customization. Like I've mentioned before, personalization jumps out at people - the more you can imply "hey, I'm not a robot, and I actually found this interesting", the more likely a lead is to take this seriously.
• That said, you would, of course, build this introduction into a proven email template using something like Mailshake; i.e you'd start with a sentence or two (or a subject line) that's heavily personalized, and then work your way into a standardized pitch. But at that point, your personalization will have fulfilled its purpose and gotten the eyeballs you wanted.

Understanding needs, proposing solutions

At this point, most salespeople will 'jump on a call' with the prospect to try and understand their needs and attempt to sell. This is, understandably, a key part of the sales process, as it's here that you build a relationship with the prospect, grow comfort, and get to know them personally.

The artificial intelligence sales approach I've outlined earlier can't fully replace people yet. Intonation, logic, timing the offer - these are all hard to get right, and the closer we get to robots emulating human mannerisms, the worse the uncanny valley becomes (and the more negative the psychological impact).

For now, we can optimize around this step with the following:

• Writing a sales script? Use artificial intelligence to help boilerplate the majority of it.
• Don't spend time writing your own script or system from scratch - use one that's already been proven. In-depth, advanced B2B sales progressions like Gap Selling can scaffold the majority of your sales cycle.
• Once you have the barebones, you can use natural language models to flesh out specifics. Most big AI writers have read hundreds of thousands of in-depth sales pages and can nail everything up to your unique value proposition with something like this:

Closing the deal

If the lead is qualified and you've got them to the point where they're interested in your product or service, now is the time to close.

Historically, closing is a messy affair. It's rife with inefficiency, from the types of e-contracts you're using to how you collect the money. Artificial intelligence can automate large portions of the process for you:

• Use an automated e-signature platform like PandaDoc or HelloSign. These software tools can be integrated into any workflow, and you can use artificial intelligence to customize each contract to your customer's needs.
• For example, using Make, you could build a simple workflow that triggers a call to the OpenAI API whenever a deal is agreed to. This could turn a list of bullet points into a paragraph-style explanation of what the customer is getting. PandaDoc would then send your client a customized contract with the new addition.

• The great part about these contract tools is they do the heavy lifting of following up for you. Additionally, many have payment functions integrated using a third party like Stripe. You can then attach a trigger upon successful contract signing (again, using Make) that would send a customized onboarding email, or start assigning the client to the various post-close procedures your company is responsible for.

Conclusion: improving the artificial intelligence sales cycle

Okay, that was a lot to take in, so let's summarize. In the modern age, you can use artificial intelligence (and good old common sense) to:

• Completely automate lead sourcing with workflows like PhantomBuster and Make (90% improvement),
• Completely automate lead qualification with AI-customized emails that your clients actually click on (90% improvement),
• Standardize your sales script with pre-vetted systems like Gap Selling, and then customize your boilerplate with AI (30% improvement),
• Eliminate all sales work after an agreement is given with PandaDoc, Make, and a little AI love; i.e, no more following up, no more collecting payment, no more customizing proposals or invoices, no more dealing with onboarding

There are also dozens of other use cases we have yet to scratch the surface of which I'll cover in future posts (how about following up with completely customized, cute cards to a prospect's address with zero additional work? the opportunities are endless!)

In short, human-driven sales are awesome. Artificial intelligence sales are even more awesome. Move over, Jordan Belfort - the future is here.