It’s ok if I use AI, but not if you do

🤖 This is a Science Says special on the latest scientific research in AI 🎓 Use it to get better, evidence-based results when using AI in your products and marketing 📈

This insight is brought to you by… Modash

Figuring out how much to pay influencers is hard.

The best way to learn? Ask folks who have hundreds of collaborations under their belts.

In this guide to Instagram influencer pricing, 3 seasoned industry pros weigh in on:

• What you can buy from Instagram influencers (Reels vs. Stores. vs. Posts)

• What pricing should be based on

• Ballpark ranges & CPMs for Reels & Stories

• Instagram-specific influencer negotiation tips

• …and more

Want to sponsor Science Says? Here’s all you need to know.

📝 Intro

Imagine you’ve just bought the latest Science Says Playbook and as you open it, on the first page it explains that it was AI-created (e.g. using ChatGPT). Would your opinion of the playbook change? Would you still value Science Says insights in the same way?

Here’s what new scientific research from my alma mater, Rotterdam School of Management, Erasmus University has to say.

P.S.: For the record - we do not rely on AI to analyze research. It’s a non-negotiable for me that Science Says has to be extremely accurate, and we just haven’t been able to get good enough results. It works great for many other things (e.g. helping me produce ideas and repurpose insights for my LinkedIn posts), but not when accuracy is essential. At the moment, it’s unclear when or whether this will change.

That’s why I have a fantastic team of marketing PhDs working with me, and there are at least 3 of us who work on each insight triple-checking each other’s work.

It’s also why there’s nothing else like Science Says.

Want hundreds more insights like these? Explore all Science Says insights here.

People undervalue work others make using AI, compared to their own AI-made efforts

Topics: AI | Messaging & Copy
For: B2C. Can be tested for B2B
Research date: May 2024
Universities: Erasmus University

📈 Recommendation

Be careful when using generative AI for content creation or copywriting. People are likely to value your work less and assume AI played a larger role in the creation than it actually did. 

To reduce this negative effect, specify how you used it (e.g. AI helped provide examples) to make the actual extent of AI usage clear. If ethical, you can also avoid mentioning that you used AI to help you.

People will value your content more highly.

🎓 Findings

• People find their own usage of generative AI (e.g. ChatGPT, Gemini) and AI tools for content creation more acceptable than when others use it for the same tasks.

• As part of a series of 5 experiments, researchers found that people thought:

  • It was 12.6% more acceptable for them to use ChatGPT to prepare a job cover letter compared to someone else using it for the same task

  • It was 11.2% more acceptable for them to use ChatGPT to complete an assignment compared to if someone else did

  • When looking at tasks that were done with the assistance of Gen AI, they:

    • Thought other people were more likely to outsource a task to AI than they were

    • Thought their own (vs AI) contribution to a task was 22.4% higher than other people’s contributions (vs AI)

    • Thought they were (vs others) 38.7% more likely to use ChatGPT for inspiration and help rather than outsource a task to it.

🧠 Why it works

• There’s a disconnect between how we look at our own and other people’s behaviors - especially bad behaviors. When we see bad behavior by other people (e.g. being late), we blame it on internal reasons (the person isn’t punctual), while when we show the same behavior, we consider external forces to be responsible (there was bad traffic on the way).

• Similarly, we also tend to consider ourselves and our own actions more positively. This means when we use generative AI, we assume the best intentions in how we use it, but not in how others do.

• We also think that products influence other people’s performance more strongly than they influence our own performance. For example, if we see someone scrolling through their phone in a group, we’ll assume the phone is making them anti-social, but wouldn’t blame our phone if we’re doing the same.

❓ Are you making business decisions based on nothing but a gut feeling

Successful businesses make decisions based on real data, from real consumers. 

Conjointly is an all-in-one survey research platform with easy-to-use advanced tools and expert support.

Whether you're optimising product pricing, refining claims and messaging, perfecting your product range, or exploring other research questions, Conjointly helps you make data-driven decisions with confidence.

This announcement was sponsored. Want your brand here? Click here.

✋ Limitations

• Given the speed with which AI usage - and capabilities - are expanding, it is likely perceptions of its use will change over time.

• The study focused on controlled, experimental settings - real-world complexities, as well as the reputation of the brand or person using AI, may lead to different results.

• The findings focused on writing-based tasks - it’s unclear but likely the results would be the same for more creative tasks (e.g. designing visuals). For more analytical tasks (e.g. analyzing trends and primary data), it’s likely, but wasn’t tested, that the result wouldn’t be as strong.

🏢 Companies using this

• Companies and marketers seem reluctant to highlight if and when AI has been used in their creative process - perhaps due to (correct) concerns customers would value the work less.

• Though they rarely advertise it, creative agencies have begun using AI to optimize their work. Members of the Forbes’ Agency Council shared some insights on how their firms use AI:

  • Creative agency Isadora Agency uses AI tools to generate the copy for their paid media ads, with the copy then reviewed and edited by professionals to ensure alignment with the correct tone of voice

  • Story Amplify, a full-stack agency specializing in the financial services sector uses AI to generate keywords and write copy for their Google Ads, based on the pages they’re planning to target.

  • Bussler & Co. have developed an in-house AI model to analyze the spending and creatives of competitors’ ads to help optimize their clients’ budgets and targeting strategy.

Agencies like Demo, are transparent in 1:1 conversations with clients about how they use AI to help generate ads and content - but, correctly, don’t advertise it on their website.

⚡ Steps to implement

• Try to train the generative AI you use in your brand language and tone to generate the best results, but remember any AI-produced content should be carefully edited and reviewed by a human. 

• When you mention that AI-inspired or helped in developing content, mention that it’s an AI trained by [your company] to mitigate the perception that it’s generic content without any human involvement.

• Some of the ways you can use AI to generate content - that’s then reviewed by humans - include:

  • Generating topics and details for social media content, like LinkedIn posts and articles

  • Using it to generate SEO-optimized blog posts for your website that are then edited by a human - it’s better and more cost-effective than human-led SEO

  • Suggesting options, or editing human-made options, for ad copy and descriptions

• Be clear with your customers about what role AI plays in generating your copy and content and what is done by humans. Generally speaking, it’s best to use AI for inspiration or to find relevant examples, while using actual humans to edit the content. 

🔍 Study type

Lab and online experiments

📖 Research

Acceptability lies in the eye of the beholder: Self-other biases in GenAI collaborations. International Journal of Research in Marketing (September 2024).

🏫 Researchers

• Begum Celiktutan, Rotterdam School of Management, Erasmus University

• Anne-Kathrin Klesse, Rotterdam School of Management, Erasmus University

• Mirjam A. Tuk, Rotterdam School of Management, Erasmus University

Remember: This is a new scientific discovery. In the future it will probably be better understood and could even be proven wrong (that’s how science works). It may also not be generalizable to your situation. If it’s a risky change, always test it on a small scale before rolling it out widely.

Here is how else Science Says can help your marketing:

• 📈 Join the Science Says Platform to unlock all 250+ insights, real-world case studies, and exclusive playbooks

• 📘 Boost your sales and profits with topic-specific Science-based Playbooks (e.g. Pricing, Ecommerce, SaaS, AI)

• 🔬 Get on-demand evidence to make better decisions. My team of PhDs and I regularly help leading brands in FMCG (e.g. Mars), retail, and tech. Reach out here.

🎓 It took 3 of us 15.5 hours to accurately turn this 17 page research paper into this 3min insight. 

If you enjoyed it, please share it with a friend. If you’re that friend, you can subscribe below for $0 👇