(Yet) More B2B Leads with ChatGPT

In our last article we gave an overview of the possibilities of ChatGPT for B2B lead generation. Since then, a lot has happened. The development of AI follows the well-known pattern of “first slowly, then all at once”. ChatGPT 4 was launched on 03/14/23. In short, the new version is significantly more powerful. It was trained with 100 trillion parameters (vs. 175 billion for model 3). ChatGPT 4 can “understand” the images, is more difficult to trick and “retrain” and can process 8 times more words than the previous version, namely 25,000 words per query. The ChatGPT API was also launched in March. So, if you are considering generating B2B leads with ChatGPT, the current version has even more to offer.

What is ChatGPT?

ChatGPT stands for “Chatbot Generative Pre-Trained Transformer”. It is a product of OpenAI. It is a chatbot that communicates human-like with the user via text input and output. This technology is significantly more powerful than all previous attempts to build AI-based chatbots and the number of use cases is growing every day.

The author of this article has been dealing with AI-based chatbots for a long time and is familiar with various previous technologies. They could be trained with great effort, but even then, the result was rather poor. With ChatGPT, a breakthrough has been made to bring AI to a new level.

Of the numerous use cases that this product offers, we are interested in B2B lead generation. That’s what we’re going to focus on in this article.

Use Case 1: Generate Content

Our blog readers know that content marketing is one of the best sources of B2B leads. If you want to generate more B2B leads with ChatGPT, content creation is the best place to start. You can solve various tasks, e.g.:

• Have an article written on a specific topic
• Generate ideas for the next blog posts
• Create a product description for your website
• Generate newsletters

The number of possibilities is unlimited. For most use cases you don’t even need external software, you can access OpenAI Chat directly. The Plus version costs 20 euros per month.

It is still doubtful whether the ChatGPT can replace the professional copywriter (in case of doubt, this article is written by a human being 🙂 ). Although the technology can analyze and describe the complex scientific topics, the software does not have your own experience, your very personal view of your field and, above all, your style. But for smaller and simpler in-between topics, you can easily use the software. So go ahead and “cheat” and see if anyone can tell the difference.

An example:

And the result:

What to watch out for:

• Sometimes ChatGPT can generate wrong answers.
• The knowledge of the software is limited to the events that happened up to 2021.

Currently, ChatGPT is best used for the general descriptive tasks and less for the fact-based articles.

Use Case 2: Chatbot

Chatbots have been in use for many years. The first hype came somewhere in 2015-2016. Since then, numerous support and customer chats have been switched to chatbots, saving millions of hours of work. But it wasn’t enough for a big breakthrough. Well, no one will voluntarily converse with a chatbot instead of talking to support or an account manager.

The previous chatbots could be roughly divided into three categories:

• Path-based chatbots. The simplest chatbots where you can configure a question-and-answer path. such as B. ask a question 1, to which there are answers yes and no, if yes – question 2, if no – question 3 etc.
• FAQ chatbots. In this case, a list of questions and answers is pre-generated and when a customer asks a question, e.g., B. refers to prices, the chatbot selects an answer from the list of potential answers that best fits the question (here there may already be the simplest AI or machine learning algorithms so that the chatbot can analyze the question).
• AI-based chatbots. Such bots have been around for a long time and are often used. For example, Microsoft Azure’s solution was one of the most capable. But it also had to be laboriously optimized and trained and was not exactly cheap.

ChatGPT puts all old versions of chatbots to shame. The ability of this technology to understand the context, access the large amount of data and generate human-like answers is hard to beat.

An example from an OpenAI based chatbot (but in this example a “sarcastic” bot is used 🙂 ):

The advantages of a ChatGPT bot are obvious:

• The model is already pre-trained based on numerous data. You only must enrich the bot with the necessary and more detailed information about your own product.
• The way the questions are answered is difficult to distinguish from a human answer.
• ChatGPT understands the context very well and analyzes the question in such a way that the customer does not have to rephrase it several times.
• You can predefine the length and level of detail of an answer.
• The API makes it easy to integrate the technology with common tools and software.

With all these benefits, you can take B2B lead generation on your website to the next level with the help of a chatbot and inform (and amuse) the customers in such a way that they will decide on your offer.

If you don’t want to mess with the ChatGPT API yourself, you can use Zapier to connect your current chatbot to OpenAI without coding, just drag and drop.

By the way: LeadRebel plans to release a first version of the bot in April, which is based on ChatGPT and at the same time analyzes the data of the website visitors and generates very targeted answers.

Use Case 3: Enrich B2B Data with ChatGPT

Clean and up-to-date lead data plays a huge role in sales. The author of this article keeps getting cold calls where he is addressed as “Erich Honecker”. This means that somewhere, in some godforsaken corner of the internet, there is a cheap database of phone numbers and fake names.

There are many such databases in the B2B industry as well. The data from the previous century, the t-online e-mail addresses of managing directors who founded the company 30 years ago and have now risen to international corporations.

The problem is currently being solved by a variety of lead sellers and corporate database providers. Prices vary from tens of thousands per specific amount to tens of euros per month.

However, if you want to enrich, manage, and improve your data yourself, ChatGPT is available (more correctly, these are other offerings from OpenAI, but for the sake of simplicity we will ignore this difference). The technology can automatically extract the data (e.g., from the imprint). In contrast to web scrapers, you don’t have to write any code or develop clever algorithms that can read differently structured websites.

An example:

In this example, it is a short text that is manageable anyway. In the next example we make the task for ChatGPT a bit more complex: add a bit more text, but generate additional output, such as company description, search terms and company branch:

The advantage of data mining with the help of ChatGPT:

• The unstructured data is easy to read. This is also where the biggest advantage over various web scrappers These need to understand the structure of a website and the data should be well labeled (e.g., the phone number needs to be preceded by “Phone”, or “Tel:” or one of the many variations, while ChatGPT understands the context like a human and accordingly acts).
• ChatGPT can “conjure up” new information from very limited data. E.g., classify a company in a certain industry, create a detailed company description, although there were only a few lines of information. In this sense, ChatGPT supplements the data based on its own “imagination” and the data with which the technology was previously enriched.

But there are also a few disadvantages:

• Processing may take a few seconds. Sometimes ChatGPT “thinks” too long.
• The software’s imagination can sometimes make up incorrect information and the company description can be inaccurate. However, this factor can be preset, f. e. allow more or less creativity.

The only task you would now have would be to integrate OpenAI’s API and find the source for the data (e.g., LinkedIn, commercial register, imprint, etc.).

In this way you can generate even more B2B leads with ChatGPT or enrich the existing leads with data or refresh the old data.

More B2B Leads with ChatGPT: Conclusion

The possibilities of ChatGPT in many fields, including B2B lead generation, are numerous. We will actively follow the topic, integrate it into our own business processes and present the new use cases to you.