Three Ways Generative AI Will Change B2B Marketing

Generative AI—the use of AI to create content, such as images, text, and video—is everywhere. You can't read online without running into some piece about ChatGPT/GPT-4 or other large language models (LLMs).

Generative AI tools have captivated the imagination of millions, but what do they have to offer the B2B marketer? How will we be applying them in our work?

Let's first establish a few basics. The primary way we as humans interface with these tools is using natural language instructions, known as prompts. The art and science of getting AI to do what we want is known as prompt engineering.

An example of a prompt might look something like this:

You will act as a B2B content marketer. You have experience in B2B marketing,?account-based marketing, ABM, ABX, and key account selling. Your first task is to explain the role and importance of email marketing in B2B marketing in a blog post. Write 500 words in a professional tone of voice.

That prompt, when used with a tool like ChatGPT, will create text-based output. Other tools, such as Stable Diffusion and DALL-E 2, take text input and generate images, or even videos; you may remember about six months ago an explosion of AI-generated artwork as the tools hit the market.

Although images of dogs wearing ballet tutus riding on skateboards are great fun, how is generative AI likely to impact the day-to-day work we do as B2B marketers?

Generative AI in B2B Marketing Tools

The first way we'll be using generative AI is in the B2B marketing tools we already use today. For example, HubSpot has announced GPT-based integrations that allow sales and marketing teams to interact with the software using natural language; HubSpot's tool is known as ChatSpot. Salesforce recently announced Einstein GPT as its entry into the field.

A sales professional might use a prompt like this in ChatSpot:

Show me all opportunities in our pipeline with a lead score over 45 that haven't been contacted or had any activity in the last 90 days.

Such natural language prompts are much easier for both technical and nontechnical users, especially stakeholders who have specific questions they want answers to in the data they already have.

For marketing analytics folks in particular, generative AI will be a valuable and powerful addition to their toolkits, helping stakeholders self-serve on many more questions.

But it's not just analytics that will be affected. Common tasks that take multiple, sometimes confusing, steps in today's?CRM software?will be shortened to single prompts.

For example, in ChatSpot, you can ask it to perform a task like this:

Move [email protected] to an opportunity and set the opportunity stage to prospect, the deal size to $100,000, and the deal close date to December 31, 2023.

Anyone who's ever suffered the hours upon hours it takes to update your CRM data will appreciate the convenience of simply telling the software what to do and letting generative AI do the heavy lifting on the back end. (Full disclosure: ChatSpot is still in alpha and very fragile as HubSpot gets it off the ground, in case you want to try it in your own instance.)

Generative AI Integrated Into Productivity Software

The second way we'll be using generative AI is in standard office productivity software.

Microsoft announced recently its?Copilot?integration, bringing the ChatGPT experience into Microsoft Office; Google announced similar with its?PaLM?language model integration into Google Docs, Gmail, and other Google services.

Both tech companies touted generative AI integration as substantial?productivity enhancers. Microsoft's demonstration showed a prompt where a user directed MS Office to make a PowerPoint presentation out of a set of meeting notes in a Word document.

Such generative AI integrations will substantially affect B2B marketers. Any B2B marketer who's just stared at a blank page for an hour trying to come up with ideas will find a ready ally in Copilot.

Though tools like ChatGPT give marketers access to those capabilities today, they're not integrated into our office productivity software. Being able to access a true virtual assistant right inside the office software we use most will be a huge productivity booster.

Imagine sitting down in Microsoft Word and instead of staring at the screen, opening up the Copilot pane and typing a prompt like this:

Generate a sales proposal for Acme Corporation based on the call notes from October 19. Make the terms of the proposal Net 30. Offer a discount for a 50% deposit up front. Make the proposal to the attention of Jane Smith. Make me the signatory on the proposal. Put it on our letterhead.

Productizing Successful Prompts

The third way we'll be using generative AI is taking prompts that we find to be very successful and productizing them.

For example, suppose the blog post prompt above works really well. Wouldn't it be nice if we could scale that? With technical help from our developers (or agency partners) we could make that a system, a process.

At?TrustInsights.ai, for example, we feed our SEO keyword lists to the OpenAI GPT-3.5 model along with predictive forecast data, and it generates blog post headlines as a content calendar for us, dramatically reducing the time it takes to stand up new content.