This article was originally published in the Elevate Your Marketing newsletter and has been repurposed and republished here with the author’s permission. Here is the link to the original article. 

I wanted to cover this topic in one of the future editions of the newsletter. But like everything in marketing and business in general, timing is crucial. So before the GPT-4 buzz fades out, I wanted to share my two cents on how generative AI has been (and will be) changing various domains within B2B marketing.

Before we dig deep into what generative AI means for B2B marketers today, I thought it would be useful to list down all the articles I have written in the past about the application of AI in marketing.

Here are they:

1. How artificial intelligence is changing the way content is written
2. AI-based content writing tools you should explore – Part 1
3. AI-based content writing tools you should explore – Part 2
4. How conversational AI is transforming marketing (and sales)
5. How AI can ‘give voice’ to your brand through Brand Language Optimization
6. Leveraging AI for automated video creation
7. Using AI to improve SEO effectiveness

These articles touch upon a wide variety of use cases where AI is literally transforming the ways in which marketing tasks are executed. Most of the tools I have discussed in them are based on GPT-3.5 (the predecessor of GPT-4). Now with the launch of GPT-4 and the advancements other companies are making in the space, newer use cases will emerge while existing applications will improve in quality and speed.

What is generative AI?

I don’t wish to get geeky about the topic. So let me explain generative AI in simple terms:

Generative AI is a technology that leverages sophisticated deep learning and machine learning algorithms to create content automatically without any human involvement.

When I say, there is no manual involvement, I am referring to the creation process. Manual prompts are required to tell the system what to create.

Tools that are based on generative AI eliminate the need for mundane – and sometimes intelligent – tasks when it comes to content creation and speed up the whole process.

GPT-4 – the talk of the town

Not soon after ChatGPT made a huge buzz in the market, OpenAI launched GPT-4 – its latest language model. GPT-4 is a generative AI system (or model) that can take both text and image as input while providing output in the text format.

GPT-4 vs GPT-3.5

GPT-4 is more advanced than GPT-3.5 in many ways such as the following:

• It can perform better when it comes to consolidating information and converting that to human-like responses (GPT-4 scored in the 90th percentile in the Uniform Bar Exam as compared to the 10th percentile by GPT -3.5). It is also 40% more likely to provide facts and figures accurately.
• It comes with superior multi-lingual capabilities (According to OpenAI, in the 24 of 26 languages tested, GPT-4 outperformed the English-language performance of GPT-3.5 and other LLMs).
• It can take images as inputs.
• GPT-4 is more steerable than GPT-3 (This essentially means that you have better control over the tone and voice in which a system built on GPT-4 would respond).
• It is safer than its previous versions. As per msn.com, GPT-4 is 82% less likely to give responses to harmful content.

With these cutting-edge features, GPT-4 opens a whole new horizon of AI use cases, especially in marketing (we will discuss the use cases in detail in a later section).

Limitations of GPT-4

Much like its alternatives and predecessors, GPT-4 also comes with various limitations:

• Though more accurate compared to GPT-3.5, GPT-4 can still make errors when it comes to giving responses – especially those involving stats and extensive reasoning (As OpenAI states, it “hallucinates” facts and makes reasoning errors).
• The model is trained using data till September 2021. Hence it cannot produce reliable output related to any events that happened post that.
• It does not perform well when it comes to future predictions.

If you are keen to learn more about the features of the model and its limitations, please check out the Open AI article on this.

Also read: Marketing In The AI Era – How B2B Companies Should Adapt?

What could GPT-4 and generative AI advancements mean for marketers?

The scope of generative AI is not limited to GPT-4. For instance, DALL-E 2, another AI model from the creators of GPT-4, is capable of producing images based on text prompts.

Another example is Lensa – the photo app that recently took the internet by storm. It converts your photos into tens (or even hundreds) of pieces of digital art. The tool uses a technology called stable diffusion to make this happen.

Further, OpenAI is not the sole player in the space. Claude, an AI language model (and an alternative to GPT-4) developed by Anthropic, is also finding its way by partnering with popular names like Quora and Notion.

Next, let us talk about what these recent developments in the AI space mean for marketers.

A quick recap of what we have already covered

I linked to the articles I had written on the application of AI in marketing at the beginning of this article. All of them are examples of generative AI use cases in one way or the other. Let’s do a quick recap of those:

1. Content creation and generation: long-form content (like blog posts, video scripts, case studies, etc.) and short-form content (such as ad copy, landing page copy, short descriptions/summarization, social media content, etc.).
2. Conversation marketing: chatbots that can automatically generate intelligent responses to human queries.
3. Brand language optimization: optimizing messaging and copy for a unique and consistent brand voice.
4. Automated video creation: turning scripts into fully finished videos in minutes.
5. SEO: large-scale keyword analysis and automated meta tag updation (and many more).
6. Enhancing search: providing the most relevant information to users (Google has been a frontrunner in this space. Now with Bing integrating GPT-4 into search, the world’s largest search engine is likely to see some stiff competition from its biggest rival).

Where else can generative AI be applied in marketing?

In addition to the applications we have already discussed, there are many more areas where generative AI is already being applied. Let us look at some of them in this section.

1. Identifying patterns and segmenting unstructured data

Analyzing large chunks of unstructured data is a challenge in marketing too. Using AI here for the purposes of analyzing website chat data, CRM data, user behavior information, etc., can make data-driven marketing more effective. An example of a real use case for this is to predict event ticket sales based on past user behavior and other relevant information.

2. AI in Email marketing

Using AI in email marketing happens at three levels.

The first is curating relevant content based on the persona of the recipient (xIQ is a tool that can help you do this).

The second is crafting awesome emails using AI writing assistants – which is something that is covered in my article on content creation and generation.

The third involves personalizing the content you send out. For instance, rasa.io helps you to send personalized newsletters by leveraging AI.

Also read: B2B Email Marketing Trends For 2024 And Beyond

3. Content repurposing

Automating content repurposing at scale using AI can save a lot of time for busy content teams. Using a tool like Automata makes this a reality.

Related: Content Clustering And Repurposing – Building A Scalable B2B Content Strategy

4. Automated image generation

As mentioned before, DALL-E 2 is an AI model that can create images from text-based prompts. Though not highly accurate, this tech is improving. The time when it will be able to eliminate the need for designers is not very far. Midjourney is an example of another popular tool that does this.

5. Out-of-home (OOH) advertising

Out-of-home advertising using AI-enabled digital signage has been picking up lately – especially in the US. AI plays a part in them by helping to show ads based on the demographic characteristics of the viewers. This tech is expected to pick up momentum with the ability to make a traditional method of advertising more targeted and measurable.

6. Translation

Google Translate has been there for ages. But there are many more use cases of translation. A notable player in the space is Duolingo – a language-learning app. According to PC Magazine, the app uses GPT-4 for AI-enabled language learning. More about this here.

Now, if you are wondering how this can help marketers, translation can make it easier to convert content created in English to other languages. Given that SEO is less competitive in other languages like Spanish, French, Japanese, etc., the ability to translate content at scale can be a game-changer.

7. Automated transcription

Transcribing speech to text using AI has helped marketers repurpose content much faster. Tools like Descript make it possible to do this with a high level of accuracy. This can be applied to audio content like podcasts as well.

8. Creating variants for advertising and A/B or multivariate testing

Online advertising is a game of optimizing every element of the ad to achieve maximum ROI. AI-based tools can enable faster A/B (or multivariate) testing by automatically creating different variants of an ad (both the creative and the copy). This helps to identify ads that perform the best at lightning speed.

9. Text-to-speech (TTS) conversion

Automated text-to-speech conversion has improved quite a lot in the last 2-3 years. This technology makes it possible to create human-like voices without having to use a professional voice-over artist in your audio or video content.

One of my favorite players in the space is Speechelo. Speechify is another tool that can do this.

10. Speech and call analytics

Sales intelligence is an area where AI has delivered huge ROI to its users. Speech and conversational intelligence platforms like Gong and Chorus AI can derive valuable insights from your sales conversations. Though not directly related to marketing, marketers can use the insights from these platforms to further finetune their messaging, content, and copy.

11. Website data analytics

Have you heard about a tool called PaveAI? It turns your Google Analytics data into recommendations. While the analysis part is done using machine learning algorithms, the tool uses generative AI to produce recommendations based on the analyzed data.

12. Creative design

The use of AI in design in itself is a separate topic. We already touched upon automated image generation. Some of the other interesting AI use cases in the area of creative design include:

1. Removing video backgrounds (example tool: Unscreen)
2. Creating videos, logos, and banners in minutes (example tool: Designs AI)
3. Image quality enhancement (example tool: Let’s enhance)
4. Image to HTML conversion (example tool: Fronty)

The role of GPT-4 and other AI models in the marketing AI revolution

Many of the use cases and tools I discussed today use either GPT-3 or GPT-4 for building their generative AI engines. Those who have been using GPT-3 are very likely to migrate to GPT-4 soon. This will lead to further improvements in the quality delivered by them.

Also, other models like DALL-E 2, Claude, CLIP, etc., are also expected to play their own role in the AI space – especially with respect to marketing.

The future of AI in marketing

If you noticed, the title of this article said ‘GPT-4 and beyond’. The objective of this was to cover what the future will look like for marketing with AI gaining momentum in the space faster than ever before.

We already touched upon some of these points in the form of use cases and how new models have been making their mark. In addition to them, here are some of my predictions for the future when it comes to the application of artificial intelligence in marketing.

Let’s start with some generic ones:

• AI models will become more accurate, safer, and more reliable with time.
• The adoption and practical application of AI in marketing will rise.
• Right now, AI is mostly about how cool the tool or use case is. But soon it will face the heat of showing ROI in dollar terms.

Now, let us move to some specific predictions:

• The need for stock images will soon die – except for very niche domains that cannot be easily comprehended by AI. As a result, even tools like Canva or Figma (at least a part of their platform) will pivot to a model where users can create designs using prompts and small inputs (like title, description, brand template, etc).
• AI will make its entry into more mundane and niche tasks in marketing. An example of this is revenue attribution. By analyzing pipeline and funnel data, generative AI-based tools will be able to map – with decent accuracy – the channels that contribute to the topline of the business.
• GPT-4 currently takes both text and image as inputs. It is likely to expand this horizon by adding more content formats to the list – such as gifs and videos.
• A new segment of tools that help to validate the output of generative AI-based tools will emerge (in fact, this has already started happening in bits and pieces).
• Just like how consolidation is happening in the SaaS industry now, 3 to 4 years down the line, the same will happen in the AI space as well. It will be characterized by either or both of the two trends – first is mainstream players offering AI solutions along with their core offerings (like how HubSpot developed ChatSpot), and second is dedicated AI companies offering a heavy suite of AI products for marketers.

Final words

Generative AI is a heavy topic (and an ever-evolving space). Its application in marketing is not something that can be covered in a single article. Since I could not go too much into the details, I suggest you do more reading on this in case you are keen.

One of the thought leaders in this space is The Marketing AI Institute. I see them creating a bunch of awesome content related to AI in marketing. It is a great place for you to learn everything related to the application of generative AI (and AI in general) in marketing (I have used the website as a reference for this article as well as a few of my previous ones).

Also, as I always mention in my articles and other forms of content, the tools that I mentioned in this article are curated based on my experience as well as other online and offline sources. If you are considering using any of them, please do thorough research and evaluation on your own to make sure it fits your business use cases.

That’s all I wanted to cover today. Hope this was an interesting read.

Looking to grow your B2B business with marketing? Try Skalegrow

With marketing getting tougher and tougher, every wrong foot you make might hamper your growth. What you need is the right guidance and a helping hand. This is where Skalegrow can make a sea of difference.

Skalegrow helps IT, tech, SaaS, and embedded systems companies leverage new-age marketing tactics to grow their business. Check out the below intro video to learn more about what Skalegrow brings to the table: