NEW: iContact Case Study

The Role of AI Prompt Engineering in Marketing

 In AI, Business Insights

Marketing professionals are increasingly recognizing the value of effectively using generative AI, and OpenMoves is no exception. By leveraging the power of this technology, we are able to perform our marketing tasks more quickly, efficiently, and with greater effectiveness. Successfully using generative AI relies on prompt engineering, which entails interacting with and instructing the AI in a manner that yields the desired outcome. This involves embedding the task description that the AI is supposed to achieve into the input.

There’s an old saying in the world of computers and programming: Garbage in, garbage out (GIGO). In other words, the output is determined by your input. If you provide flawed, incorrect, or low-quality input data or instructions to a system, computer, or process, the output or results will also be flawed, incorrect, or of low quality.

In the world of Natural language processing (NLP), Artificial Intelligence (AI), Machine Learning, and Large Language Models (LLMs) like ChatGPT, Bard, etc., this phrase is very apt. More simply, the quality of the output of LLMs like OpenAI’s GPT is dependent on the massive datasets from the internet that it is trained on. Moreover, when you interact with these models by asking questions and ‘prompting’ them, the answers are highly dependent on how you ask your questions. It’s important to ask the right questions in the right way if you want desirable results.

Marketing professionals now have a brand new tool that will help make their jobs easier and more efficient. They can take advantage of generative AI in myriad ways and not even have to be technically inclined. 

Generative AI is fast becoming the go-to tool for OpenMoves marketing professionals. It helps us with content ideation and creation, perform market research and glean insights. It helps with creating social media plans and execution, writing email copy, and helping with client SWOT analysis too, just to name a few. And prompt engineering is the way in which we interact with it to help us with these tasks.

Let’s explore first what a prompt engineer is and does, and then look at how it can supercharge a marketer’s output. 

First, a few definitions to help

  • Artificial Intelligence or AI (IBM)

Artificial intelligence is a field, which combines computer science and robust datasets, to enable problem-solving. It also encompasses sub-fields of machine learning and deep learning, which are frequently mentioned in conjunction with artificial intelligence. These disciplines are comprised of AI algorithms which seek to create expert systems which make predictions or classifications based on input data.

Machine learning is a subfield of artificial intelligence, which is broadly defined as the capability of a machine to imitate intelligent human behavior. Artificial intelligence systems are used to perform complex tasks in a way that is similar to how humans solve problems.

  • Natural language processing (NLP) (IBM)

Natural language processing (NLP) refers to the branch of computer science—and more specifically, the branch of artificial intelligence or AI—concerned with giving computers the ability to understand text and spoken words in much the same way human beings can.

  • Large Language Models (Fast Company)
    LLMs are machine learning models that utilize deep learning algorithms to process and understand language. They’re trained with immense amounts of data to learn language patterns so they can perform tasks. Those tasks can range from translating texts to responding in chatbot conversations—basically anything that requires language analysis of some sort. 

Generative AI is the process of AI algorithms generating or creating an output, such as text, photo, video, code, data, and 3D renderings, from data they are trained on. 

What Exactly Is Prompt Engineering?

There’s a whole new job category that has recently emerged from generative AI called ‘prompt engineering’, people skilled at asking LLMs the right questions in the right way to get desired results.

More specifically, prompt engineering refers to the process of crafting effective prompts or instructions for large language models like ChatGPT to generate desired, quality outputs. It involves formulating the initial text input, which serves as a guiding cue for the model to generate coherent and relevant responses.

Prompt engineers need to understand all this and be able to coax the best (and most accurate) outputs from large language models.  

With ChatGPT taking the world by storm, and big tech following suite with Bing’s ChatGPT version and Google’s Bard, companies large and small are scrambling to take advantage of generative AI. As a result, there’s been a surge in AI job opportunities like prompt engineering.  

An easy way to conceptualize prompt engineering is to think of it like a game of Pictionary. The person at the white board provides just enough information for their partner to figure out the predetermined word or phrase.

Think of the model as representing the partner in Pictionary. Just enough information is provided to the model for it to output what’s desired and accomplish the task at hand.

And the prompts and desired output is dazzlingly varied. Here are just some of the types of prompts we can use with these models:

• Q&A

• Questions based on existing knowledge.

• Grammar correction

• Corrects sentences into standard English.

• Summarize for a 2nd grader

• Translates difficult text into simpler concepts.

• Natural language to OpenAI API

• Create code to call to the OpenAI API using a natural language instruction.

• Text to command

• Translate text into programmatic commands.

• English to other languages

• Translates English text into French, Spanish and Japanese.

• Natural language to Stripe API

• Create code to call the Stripe API using natural language.

• SQL translate

• Translate natural language to SQL queries.

• Parse unstructured data

• Create tables from long form text.

• Classification

• Classify items into categories via example.

• Python to natural language

• Explain a piece of Python code in human understandable language.

• Movie to Emoji

• Convert movie titles into emoji.

• Calculate Time Complexity

• Find the time complexity of a function.

• Translate programming languages

• Translate from one programming language to another.

• Advanced tweet classifier

• Advanced sentiment detection for a piece of text.

• Explain code

• Explain a complicated piece of code.

• Keywords

• Extract keywords from a block of text.

• Factual answering

• Direct the model to provide factual answers and address knowledge gaps.

• Ad from product description

• Turn a product description into ad copy.

• Product name generator

• Create product names from examples words.

• TL;DR summarization

• Summarize text by adding a ‘tl;dr:’ to the end of a text passage.

• Python bug fixer

• Find and fix bugs in source code.

• Spreadsheet creator

• Create spreadsheets of various kinds of data.

• JavaScript helper chatbot

• Airport code extractor

• Extract airport codes from text.

• SQL request

• Create simple SQL queries.

• Extract contact information

• Extract contact information from a block of text.

• JavaScript to Python

• Convert simple JavaScript expressions into Python.

• Friend chat

• Emulate a text message conversation.

• Mood to color

• Turn a text description into a color.

• Write a Python docstring

• Write a docstring for a Python function.

• Analogy maker

• Message-style bot that answers JavaScript questions.

• ML/AI language model tutor.

• Bot that answers questions about language models

• Science fiction book list maker

• Create a list of items for a given topic.

• Tweet classifier

• Basic sentiment detection for a piece of text.Create analogies.

• JavaScript one line function

• Turn a JavaScript function into a one liner.

• Micro horror story creator

• Creates two to three sentence short horror stories from a topic input.

• Third-person converter

• Converts first-person POV to the third-person.

• Notes to summary

• Turn meeting notes into a summary.

• VR fitness idea generator

• Create ideas for fitness and virtual reality games.

• Essay outline

• Generate an outline for a research topic.

• Recipe creator (eat at your own risk)

• Create a recipe from a list of ingredients.

• Chat

• Open ended conversation with an AI assistant.

• Marv the sarcastic chat bot

• Marv is a factual chatbot that is also sarcastic.

• Turn by turn directions

• Convert natural language to turn-by-turn directions.

• Restaurant review creator

• Turn a few words into a restaurant review.

• Create study notes

• Provide a topic and get study notes.

• Interview questions

Types of Prompt Engineering in AI

There are several different types of prompt engineering, including assigning a role, Chain of Thought (CoT), least-to-most, zero-shot, one-shot and few-shot prompting.

Assigning a Role
You can tell an AI model to play a role when answering your prompt. For example, you could ask it to play the role of “an expert at quantum physics,” “a college professor” or “an auto mechanic”. This instructs the model to answer as if it were actually in the assigned role.  

Zero-Shot
Where the AI model can understand and execute tasks generates responses without any examples given, relying only on prior knowledge and understanding.

One-Shot
Provides the AI model with only one example to demonstrate the desired task, helping it understand what is being asked and how to output it. 

Few-shot 
Provides the AI model with multiple examples to demonstrate the desired task, helping it understand what is being asked and how to output it. 

Least-to-most
A problem-solving method that involves breaking down a problem into smaller, more manageable pieces, with the results of each piece being fed into the next. 

Chain of Thought (CoT)
Encourages the AI model to explain its reasoning. It improves the reasoning and logic ability of the model to give the best and most accurate answers by prompting it to generate intermediate steps that lead to a final answer of a problem with multiple steps.

Prompt Engineering Tips for Marketers

Prompt engineering is not only a science, it’s an art as well. Crafting the perfect prompt takes creativity and flexible thinking. Marketing professionals ought to bear in mind the following suggestions when prompting AI models.

Specify the output format and length of the desired output. You can affect the formatting of the output text, like get it in tables or lists. And specifying length helps avoid unneeded extra content or allows for more comprehensive answers. 

Ask it to explain things “step by step”. Doing this is using the Chain of Thought tactic explained above and makes the model think through its answer logically. 

Provide the model with examples. Use one-shot or few-shot tactics explained above to help coax out the exact desired output you want. 

When trying to understand a difficult, challenging, or complicated concept or topic, instruct the model to “Explain it to me like I am a 6-year-old”. You’ll be amazed at how effective this is. 

Here’s an example prompt that incorporates a lot of the above concepts:

You’re an expert at writing YouTube videos. Think of catchy titles that have the chance of going viral. The titles should be short and succinct. List the titles in a numbered list format along with the associated search engine keyword, like this [title] [keyword]. Then explain the process you just used step by step.

And here’s the output:

An OpenMoves Generative AI and Marketing Real World Example 

We were tasked with increasing one of our client’s organic search sessions and leads. To continue to learn about its business and competitor landscape, we used generative AI to perform a SWOT analysis first to determine what the client’s strengths, weaknesses, opportunities, and threats are. Then we asked it to name competitors and find content gaps. We then used AI to help determine relevant keywords and topics for lead-generation in our client’s industry that they didn’t have content for already. 

When we decided on one topic that we thought would attract organic visitors who converted into leads, we used generative AI to help do the research, write the copy and to create relevant and original images for our new landing page. 

The amount of time we saved on research, content creation, and image creation alone was astounding. And the final result was a brand new, relevant, original, and robust page that covered the topic comprehensively, making it a traffic magnet that continues to produce leads.

Prompt Engineering and Marketing

Many fields are going to be significantly disrupted by LLMs and generative AI, and marketing is certainly not an exception. Actually, one could even argue that generative AI and marketing are two peas in a pod, as if made for one another. A surprisingly large number of marketing-related tasks can now be successfully carried out automatically using generative AI. And many others can be augmented by it.

What’s more, AI has the potential to change marketers’ jobs for the better, by replacing routine and repetitive tasks with work that’s more fulfilling, strategic, and creative.

Importantly, It will also be the job of marketers to serve as the first line of defense against AI hallucinations, which is when an AI model makes things up when it doesn’t know the answer, something these models are prone to do today. Marketers will also have to edit and improve any text that is generated by AI. They can’t just take the output and post it; it all needs to be vetted. This all makes the job of marketers in the world with AI even more critical.

Why Marketers Need to Know Prompt Engineering

Marketing professionals who are adept at prompt engineering will have a leg up. They’ll be best positioned to be able to conjure the best outputs for many varied marketing tasks. The better they are at prompt engineering, the more efficient they will become in their jobs. 

The ability to create new ideas and content out of nothing but a prompt, and to be able to analyze data and come up with actionable insights from that data will all skyrocket, significantly augmenting their output, recommendations, and deliverables. 

The best prompt engineers are going to have both hard and soft skills. They’ll be able to understand technical topics related to AI while also being able to communicate effectively with the AI and with stakeholders using the output. 

Marketing Tasks AI Models Can Help With

Here are just some of the types of things AI can now help a marketer with.

Generating good content topics. Marketers can ask models like ChatGPT to come up with relevant and search-for blog topics, ones that have the potential of going viral. It will output an endless number of very creative topics and article headlines.

Writing blog content. AI can now write full articles. Marketers still need to re-work, edit, and fact-check the output, but it makes the entire process more efficient.

Writing lead-generation pages. Just like Ai can write blog posts, it can generate relevant and useful content to populate product pages and lead-gen pages. 

Writing product descriptions. Ecommerce stores have lots of products that need accurate and enticing descriptions. AI can help this process immensely.  

Keyword research. Find relevant keywords to target with your SEO efforts. Search engine optimization professionals need to determine which keywords a company’s potential customers use to find them. AI can help determine exactly which keywords SEO’ and marketers ought to target with new content. 

Creating social media content. Ever left stumped when trying to come up with a new social media angle to write about? AI can help with these ideas. Furthermore, it can actually craft the posts too.

Create video scripts. AI can create entire video scripts to use with YouTube videos for example. For example, ChatGPT can generate YouTube video ideas, keywords, video outlines, and tags too.

Writing email content. When marketers use email marketing, the text in the subject line and body of the emails is crucial for getting the recipients to take a desired action. Generative AI can write all this for marketers. 

SWOT analysis. Marketers can ask AI to review a company and deliver a SWOT analysis (strengths, weaknesses, opportunities, and threats). 

Customer influence. Chatbots and virtual assistants can offer personalized recommendations, or assist with product selection. AI can improve customer interaction, streamline the customer service process, and influence buying decisions.

Market research and insights. Generative AI models can review and analyze large amounts of data, including customer reviews, social media posts, websites, and market trends, to extract valuable insights for marketers. They can identify patterns, search behaviors, sentiment trends, or emerging topics, helping marketers understand customer preferences, market dynamics, risks, and potential opportunities.

Creative design. Generative AI can aid marketers in generating unique and visually appealing images, designs for logos, graphics, website layouts, and advertising materials.  

Marketing Prompt Examples

YouTube Video Script Prompt: “You’re an expert YouTube creator. Write a conversational Youtube video script for my [product/service] about the following [topic].”

Keyword Research Prompt: “Give me a list of informational, long-tail keywords that I can use for blog posts for ideas for [product or company].”

Email Prompt: “Help me in creating an email marketing campaign for my [product/service] using the Customer Value Journey framework.”

CTA Prompt: “I’m trying to drive more sales on my landing page. What are five compelling call-to-actions I can use? Then explain how you did it step by step.”

Landing Page Optimization Prompt: “Use the 5Cs framework to help me optimize my pay-per-click landing page.” (The 5Cs are Clarity, Credibility, Conversion, Call to Action, Consistency

Brand Story Prompt: “Help me create a brand story for my [product/service] using the Rags to Riches story framework.”

Create Headlines Prompt: “Create a list of headlines about [topic]. The headlines should be memorable, eye-catching, and exciting. List them along with the best search engine keyword associated with it, like this [topic] [keyword].”

Social Media Content Plan Prompt: “Create a monthly social media content plan for my [product/service] targeting [my target audience] for [social media platform].”

Where to Learn Prompt Engineering

If you’re a marketing professional, it’s a smart career move to learn the basics of prompt engineering. Here are a few places where you can learn:

Conclusion

With LLMs and generative AI, a brave new world is being born right now and marketing professionals are at the vanguard in many ways. Generative AI and marketing are a natural fit and marketers who eschew it will likely be left behind. As performance marketing professionals, every day our output is augmented with generative AI and the future seems bright. 

If you’d like to learn how OpenMoves would use generative AI to help your business, contact us today

Recent Posts