Artificial Intelligence Use: A Framework For Determining What Tasks to Outsource To AI [Template].

AI Framework Template for AI Use
With any new technology, there are benefits and unintended consequences. Often the negative outcomes happen without thought or planning. We get caught up in the “new shiny object” mesmerized by its “magical capabilities.” That happened with social media. We can’t go back on that technology, but we are in the early stages of AI. In WIRED Rachel Botsman called for frameworks to do more to avoid the negative of tech developments.

Before jumping all in, ask, “What role should AI play in our tasks?”

Just because AI can do something doesn’t mean it is good or it should. AI’s capabilities are both exciting and frightening causing some to be all in and others to be all out. Being strategic takes more nuance. Be intentional about planning the role AI could and should play in your job or business with the AI Use Template below.
AI Framework Template for AI Use
Click the image to download a PDF template.

First, make a list of common tasks and the goal of each.

List tasks you perform in your job, on client projects, or in daily business operations. Then describe the goal of the task. Understanding the goal can help determine the human versus AI value in it. If the goal is to build a personal relationship with a customer or client, AI outsourcing may save time but undermine the task objective.

Recently a university outsourced their commencement speaker to an AI robot. Students started an unsuccessful petition for a speaker who could offer a “human connection.” The AI robot’s speech was described as weird and unmoving. Without any personal anecdotes, The Chronicle of Higher Education reports, “Sophia … delivered an amalgamation of lessons taken from other commencement speakers.”

Second, determine which type of AI Function each task requires.

The six AI functions (Generate, Extract, Summarize, Rewrite, Classify, Answer Questions) are modified from Christopher S. Penn’s AI Use Case Categories. Can the task be performed by one or multiple of these AI functions? If yes, you still want to consider how well AI can perform the function compared to a human and consider benefits that may be lost outsourcing to AI.

In my ad career clients often asked why a certain phrase or benefit was in the ad copy or ad script. Because I wrote it, I could explain it. It could be human insight from research (which AI can summarize), truths from lived experience, or talking with customers. If AI wrote the copy or script it may be missing and I wouldn’t know why AI wrote what it did. If you ask AI it often doesn’t know. Scientists call this the “unknowability” of how AI works.

Third, categorize the level of thinking each task entails.

The six levels of thinking (Remember, Understand, Apply, Analyze, Evaluate, Create) are modified from Oregon State’s Bloom’s Taxonomy Revisited. Bloom’s Taxonomy categorizes levels of thinking in the learning process. It was revisited to consider AI’s role. In each level determine the level of the task and discern AI’s capabilities versus distinctive human skills.

I had a student create a situation analysis of Spotify with ChatGPT. It was good at extracting information, summarizing, and suggesting alternatives (AI Capabilities of the Create Level). It wasn’t good at “Formulating original solutions, incorporating human judgment, and collaborating spontaneously” (Create Level Distinctive Human Skills). GPT’s recommendations lacked the nuanced understanding I’d expect from professionals or students.

Fourth, review the legal and ethical issues of outsourcing to AI.

Does the task require uploading copyrighted material? Are you able to copyright the output (copy/images) to sell to a client or protect it from competitor use? Does your employer or client permit using AI in this way? Are you sharing private or proprietary data (IP)? What’s the human impact? For some AI will take some tasks. For others, it could take their entire job.

Many companies are adding AI restrictions to contracts for agency partners. Samsung and other businesses are restricting certain AI use by employees. There’s concern about performance or customer data uploaded into AI systems training a model competitors could use. Some agencies and companies are developing Closed AI versus Open AI to run local AI storing data on local versus cloud servers. For a summary of main AI legal concerns see “The real costs of ChatGPT” by Mintz.

Fifth, employ human agency to produce desirable results.

We shouldn’t be resigned to undesirable outcomes because AI change is complex and happening quickly. Penn’s TRIPS Framework for AI Outsourcing includes “pleasantness.” The more Time consuming, Repetitive, less Important, less Pleasant tasks that have Sufficient data are better candidates for AI. Don’t give away your human agency. Decide on your own or influence others to save the good stuff for yourself.

A post on X (Twitter) by author Joanna Maciejewska struck a nerve going viral, “You know what the biggest problem with pushing all-things-AI is? Wrong Direction. I want AI to do my laundry and dishes so I can do art and writing, not for AI to do my art and writing so that I can do my laundry and dishes.” She later clarified that it wasn’t about actual laundry robots, “it’s about wishing that AI focused on taking away those tasks we hate and don’t enjoy instead of trying to take away what we love to do and what makes us human.”

Marketers are getting this message. In a survey of CMOs most are using AI for draft copy and images that are refined by humans. And over 70% are concerned about AI’s impact on creativity and brand voice.

It’s easy to get overwhelmed and afraid of the AI future.

As Tech leaders sprint forward in an AI arms race and regulators woefully lag behind, the rest of us shouldn’t sit back and wait for our world to change. Unlike the Internet and social media, let’s be more intentional. Don’t fall prey to The Tradeoff Fallacy believing that to gain the benefits of AI we must give everything away.

In Co-Intelligence, Ethan Mollick says it’s important to keep the human in the loop. It’s not all-or-nothing. Some warn of a future when we don’t have choices in what role AI plays in our lives. It’s not the future. Today we can choose how to use AI in our professional, educational, and personal lives.

What keeps me hopeful is breaking my job and life down into tasks and making intentional decisions on what to outsource to AI. Using this framework allows me to get excited about the possibilities of AI taking over my least favorite or most time consuming tasks. In my next post, I’ll give some specific examples using this framework.

This Was Human Created Content!

Much of Social Media Strategy Is Pay To Play. Use This Guide and Social Media Advertising Analysis Template To Improve Your Social Ads.

Social Media Advertising Analysis Template

When I first posted my Guide to Paid Social Media it was still very new and I was educating myself about what social media platforms had advertising available. Today every major social platform offers social ads and paid social has become a significant part of most social strategies. Social media ad spending has surpassed newspaper and magazine ads to become the third-largest ad channel, behind TV and paid search.

Why have social media ads grown? Social media feeds have become so crowded you often need to pay to reach your audience. Every minute millions of pieces of content are created on social media. Marketing consultant and author Mark Schaefer calls this “Content Shock” and it is only getting worse.

Nearly three-quarters of B2C and B2B companies and non-profits used organic social media to promote content and 49% expect to increase their content marketing spending. Just over half use paid social media as part of that content mix.

Source: Stephanie Heitman,“What Happens in an Internet Minute In 2022: 90 Fascinating Online Stats,”, May 5, 2022.

Social Ads Increased as Organic Reach Decreased.

Organic reach is the number of unique people who see a social media post through unpaid distribution. Getting someone to “Follow us on Facebook” doesn’t deliver the exposure it did in the early days of social strategy. As each platform became more flooded with content, organic reach dropped.

Between 2013-2014 Facebook adjusted the number of posts people saw in their feeds from more than 1,500 to only 300. They did this by adjusting the algorithm and the average organic reach for some business pages dropped 40%.

Cotton Delo, from AdAge described the switch in social strategy, “the main reason to acquire fans isn’t to build a free distribution channel for content; it’s to make future Facebook ads work better.” By 2016 average organic reach rates were 2.27% for Facebook, 3.61% for Twitter/X, 20% for LinkedIn, and 20% for Instagram.

Social Ads Are Efficient and Effective.

eMarketer found social media ads are one of the top ten most effective marketing tactics. A way to compare media types is cost per mille (CPM). CPM is the cost to reach 1,000 people and is used to compare the cost-effectiveness of media vehicles. Topdraw collected average CPM for forms of traditional and digital ads and shows how economical social ads can be.

Average CPM Per Media Type Shows the Efficiency of Social Media Ads.

Google Search Ads $8.60 CPM Google Search Ads $38.40 CPM
Instagram Ads $8.96 CPM Network TV Ads $20-$30 CPM
Twitter/X Ads $6.46 CPM Magazine Ads $140-$1,300 CPM
LinkedIn Ads $6.59 CPM Direct Mail $500-$1,000 CPM

Buying Paid Social Ad Posts.

While social ads vary by platform, most have similar processes. First, select a campaign objective. Social ad platforms require awareness, consideration, and conversion objectives that represent stages in the buyer’s journey. Below are common objectives under each category.

Example Options for Selecting Paid Social Media Campaign Objectives.

Awareness Consideration Conversions
Brand Awareness Traffic Catalog Sales
Reach Engagement Store Traffic
App Installs
Video Views
Lead Generation

Next, determine your budget. Most set a daily budget that adds up to the social ad budget for a campaign period such as a week or month. Be sure to account for content creation costs such as purchasing photos, graphics, videos, and writing. For a guide to determine your larger budget and categories see my Social Media Budget Template.

Then, select the audience based on your target audience/buyer persona. With organic social media posts, you determine specific days, times, and post frequency. With programmatic ad buying, the AI algorithm automatically places social media posts in specific users’ news feeds based on target audience specifications. It spends your daily budget to reach that audience at optimized times, days, and frequencies.

Most content requires videos or images plus headlines, text, CTA, and link URLs. The specifics such as size, shape, and length vary per platform and type of ad. For a guide on writing and designing content see my post on Best Practices For Social Media Content.

Analyzing Social Ad Posts Results.

Once you run your ad campaigns for a period such as a week or a month collect results from all platforms based on key metrics related to objectives. Set up those results in a spreadsheet or dashboard in social media management software to analyze. The social media advertising analysis template below gives you an example.

(Click on the template image to download a PDF)

Social Media Advertising Analysis Template

Look at budget spending per social platform as a percent of the total budget and compare key metric results as a percent of the total per platform. See what you spent the most money on and how much it contributed to key metrics. Make a list of highest to lowest results and highest to lowest costs to analyze what worked best and what worked less. If you’re targeting multiple audiences note which targets/personas are contributing to objective metrics better than others.

To optimize, lower spending on low-performing platforms, posts, and target audiences, and increase spending on high-performing platforms, posts, and target audiences. But also reserve a small percentage of your budget to experiment with new content, post type, audience, or platform. Over time you’ll optimize spending and results as you learn what works and what doesn’t. Don’t know what your key metrics are? First start with my Social Media Metrics Template.

This Was Human Created Content!