Prompt Engineering Guide
πŸ˜ƒ Basics
πŸ’Ό Applications
πŸ§™β€β™‚οΈ Intermediate
πŸ€– АгСнты
βš–οΈ Reliability
πŸ–ΌοΈ Image Prompting
πŸ”“ Prompt Hacking
πŸ”¨ Tooling
πŸ’ͺ Prompt Tuning
🎲 Miscellaneous
Models
πŸ“™ Vocabulary Reference
πŸ“š Bibliography
πŸ“¦ Prompted Products
πŸ›Έ Additional Resources
πŸ”₯ Hot Topics
✨ Credits
πŸ’Ό Applications🟦 Digital Marketing

Digital Marketing

🟦 This article is rated medium
Reading Time: 2 minutes

Last updated on August 7, 2024

Your favorite AI chatbot is a treasure trove of resources for both the mundane and the nuanced work of digital marketing, from analysis to content generation to tweaking verbiage. The following examples use a hypothetical website about Pez dispensers. Prompts should be prefixed with "Act as a senior digital marketing specialist" or something similar to elicit the best responses from the chatbot.

Keyword Research

Ask your preferred AI chatbot about a particular SEO search phrase (known as a "keyword"): what are its popular variants, synonyms, topic clusters, search intents, and long tail variants?

PROMPT: Generate a list of keywords related to Pez dispensers.

PROMPT: Generate a cluster of keywords related to Pez dispensers, and organize with column headings for the cluster, the keyword, and the search intent.

PROMPT: Generate a list of long-tail keywords related to Pez dispensers.

Try it for your keyword:

Marketing Channel Research

One of the most important concerns for digital marketers is budgeting their time and resources to different marketing channels. An AI chatbot can give you immediate insights into the types of marketing channels that will work best for a particular client and product. Once you've identified the marketing channels, you can drill down into each to identify strategies and content.

PROMPT: For a client wishing to promote their Pez dispenser portraying Youtube star MrBeast, with a budget of $100,000, generate a table showing how much money should be spent on which marketing channels. Use columns for the marketing channel, the budget allocation, and the reasoning behind it.

PROMPT: List some strategies to identify Instagram influencers that can help market the MrBeast Pez dispenser.

PROMPT: Create an introductory email to [influencer name], to ask them about initiating a partnership and referencing some of their tweets at https://twitter.com/[influencer]. [requires a chatbot with access to the Web]

PROMPT: List the top hashtags on Twitter relating to Pez dispensers.

The output for this last prompt ought to something like:

Here are some popular hashtags on Twitter related to Pez dispensers:

#PezCollectors
#PezDispensers
#PezCommunity
#PezLove
#PezCollector
etc.

Tweaking Title Tags and PPC Ads

AI Chatbots are an ideal tool to help generate text that fits within the constraints of SEO and PPC campaigns:

PROMPT: Generate a table with 5 snappy title tag options with SEO power words for a blog article announcing a new Pez dispenser portraying MrBeast. Each title should be fewer than 60 characters in length and incorporate the SEO keyword "celebrity Pez dispenser."  Add a column to display the 5 corresponding meta descriptions, which must also incorporate the same SEO keyword and a call to action.

PROMPT: Generate 5 compelling headlines, each fewer than 30 characters long, for a Google responsive search ad about collectible Pez dispensers

PROMPT: Generate 5 relevant three-word CTAs for a new campaign announcing a new line of Pez dispensers featuring MrBeast

Generating Images for Blog Posts

With the advent of tools like Midjourney, DALL-E, and Stable Diffusion, designers can now use prompt engineering to generate images. If you're unfamiliar with the syntax and techniques of these kinds of prompts, though, you can simply ask an AI chatbot for a set of text-to-image prompts to try out on your preferred AI image generation tool. Note that Midjourney was released after ChatGPT's knowledge cutoff date of September 2021.

PROMPT: Generate 5 concepts for an illustration to accompany a blog article announcing a new Pez dispenser portraying MrBeast.

PROMPT: Generate 5 text-to-image prompts for DALL-E to create dramatic images that convey concept #3 from my earlier prompt.

The output for this last prompt ought to be something like:

Certainly! Here are five text-to-image prompts for DALL-E to create dramatic images conveying concept #3:

1. "A majestic procession of Pez dispenser characters marches behind MrBeast, led by the new MrBeast Pez dispenser, symbolizing the rise of a legendary collection."

2. "Surrounded by a radiant aura, MrBeast holds the new Pez dispenser as a beam of light illuminates a room filled with awe-inspiring Pez characters, signifying the power of a legendary addition."

3. "In an epic clash of Pez dispenser heroes, the new MrBeast Pez dispenser stands tall, embodying strength and courage in the face of adversity."

4. "A storm of Pez candies swirls around MrBeast, forming a vortex of excitement and anticipation, as the new Pez dispenser takes center stage in an exhilarating display."

5. "Silhouetted against a fiery backdrop, MrBeast unveils the new Pez dispenser, its glow radiating hope and inspiration, symbolizing a new era of Pez collecting."

Technical SEO

Just as developers use AI chatbots to generate and debug code, digital marketers can use chatbots to analyze and debug technical SEO issues:

PROMPT: Based on these request and response headers, why isn't Cloudflare returning this web page from its cache? [paste in headers from Chrome Dev Tools]

PROMPT: Based on the rules at schema.org/HowTo, generate some schema in JSON for how to start a Pez dispenser collection.

PROMPT: What patterns can you identify in this data set, which is an export of Google Search Console data for 3 months for a company that sells collectible Pez dispensers? [paste Google Sheets URL if your chatbot has web access, otherwise paste CSV data]

PROMPT: Create Hreflang tags for the German, French and Italian versions of a web page for our retail store in Switzerland, with the base URL https://pezzworld.ch.

PROMPT: What is a regular expression to identify any search query that includes the singular or plural versions of the words "collectible" or "celebrity"

The output for this last prompt ought to be something like:

To identify search queries that include the singular or plural versions of the words "collectible" or "celebrity" using regular expressions, you can use the following pattern:

\b(collectible|collectibles|celebrity|celebrities)\b

Social Media Marketing

Beyond simply generating content for posts, there are a multitude of use cases for generative AI in social media marketing:

PROMPT: The Twitter post at the end of this prompt compliments our company on its Pez dispensers; compose a thank-you reply. Include relevant hashtags and  mentions of [@person1] and [@person2], who posted similar compliments. Incorporate a funny emoji. Offer an incentive for other users to retweet our response. Let the author know that we have reposted his tweet. Make sure the reply has fewer than 280 characters. Their Twitter post is: [paste]

PROMPT: We run a company offering celebrity Pez dispensers. Our newest product is a Pez dispenser featuring MrBeast, a Youtube influencer. Generate 5 ideas for social media contests to run on Twitter that will cross-promote our line of Pez dispensers to fans of artistic round tins of hard candy. For each idea, output a sample post with hashtags and emoji to explain the contest in fewer than 280 characters.

PROMPT: Respond to this negative Yelp review in a polite, respectful and measured tone, but counter the customer's complaint about the quality of our pez dispensers. Emphasizing the care our factory staff take in manufacturing them and the 3 rounds of quality inspections before shipping them. Your response should have fewer than 50 words. The negative review is: "It arrived broken, PezzWorld sucks."

The output for this last prompt ought to something like:

We're truly sorry to hear your dispenser arrived broken. Our factory staff meticulously craft each piece, and it undergoes 3 rounds of quality inspections before shipping. We'd love to make it right for you. Please reach out for a replacement.

Contributed by Prompt Yes!, a company offering instructor-led courses in ChatGPT and Midjourney.

Sander Schulhoff

Sander Schulhoff is the Founder of Learn Prompting and an ML Researcher at the University of Maryland. He created the first open-source Prompt Engineering guide, reaching 3M+ people and teaching them to use tools like ChatGPT. Sander also led a team behind Prompt Report, the most comprehensive study of prompting ever done, co-authored with researchers from the University of Maryland, OpenAI, Microsoft, Google, Princeton, Stanford, and other leading institutions. This 76-page survey analyzed 1,500+ academic papers and covered 200+ prompting techniques.

Copyright Β© 2024 Learn Prompting.