Stage18

We've been reading the headlines too. Please take a look at our take on the recent Hollywood earnings announcements and the risk and opportunities for the measurement providers that call them their clients.

A beautiful endorsement from one of our extended colleagues Stage18

Our reflections on the continued transformation of the Saudi economy and society.

Stage18 is very excited to explore Saudia Arabia together with AdWeek and leading adtech executives from around the world! Bruce Haymes Eric Weinberg

I love this post from one of the nicest, smartest and most empathetic marketer I've had the pleasure of working with, Scott Carter. He does a great job explaining how data driven marketing, and business decisions in general, benefit greatly from an infusion of understanding and empathy. #marketingimpact #empathy Stage18

Pricing is the #1 question SaaS customers ask us. And it's the most difficult to answer. Here's an intriguing solution. ?? ?? Why is pricing so difficult for startup and emerging growth B2B companies? There are a number of reasons. 1. Pricing is opaque. Unless you are a Salesforce or a Zendesk, you and your competitors likely aren't publishing your enterprise pricing. So it's hard to benchmark. Further, value propositions among competitive sets vary greatly, so just because your biggest competitor charges $X, doesn't mean you should charge the same amount. 2. Small data set. B2C companies can test and learn because of the sheer volume of potential customers considering a purchase. But if you have a relatively small pipeline, do you really want to risk scaring off a prospect with a price that's too high? Most SaaS companies default to underpricing. 3. Price discrimination. Most B2B companies essentially offer different prices to every prospect, based on what they think they will be willing or can pay. Again, this tends to create a downward bias in pricing and can leave a lot of money on the table. 4. Research challenges. Many companies I've worked with try to rely on surveys or one on one research to arrive at a price. Getting accurate results around purchase intent and price sensitivity is notoriously impossible using traditional surveys. First, people lie. Second, most people can't accurately estimate their purchase intent in a hypothetical world. That's why this research approach, the Van Westendorp Price Sensitivity Meter, discussed by Chris Chapman is intriguing. It offers a different way of asking the "how much would you pay" question in a way that may discourage gaming and misleading responses. If anyone has had experience with this or other pricing research methods that work, please share! #pricing #saas #growthcompany

Congratulations to my good friend and colleague Raj Echambadi and Illinois Institute of Technology on their launch of Runway 606, an innovative pathway for urban high school students to enroll in and succeed in college, graduating with a STEM degree from IIT. "The Runway 606 program, starting this fall, will allow any CPS student with a 2.5 GPA to apply for dual enrollment with City Colleges of Chicago and a pathway to earn a technology-based degree at Illinois Institute of Technology." “It's not only about math and science, but it is also about mentorship,” Illinois Tech President Raj Echambadi said. “It is about the holistic well-being of a student.” - and that may be the most innovative part of the program as it comes with $25K of mentorship and support services for students. Fantastic program that demonstrates how higher education can partner with government and the community to fulfill its social mission. Stage18

“Ask not what AI can do for you, ask what you can do for AI.” ?? Maybe a more accurate sentiment would be, “Ask what AI can do for you AND START asking what you can do for AI.”???? Why? Last month, Reddit agreed to license its content to OpenAI, makers of ChatGPT, for around $60 million a year. Last week, a trio of online media companies sued OpenAI for copyright infringement for using their articles to train ChatGPT. This follows a similar lawsuit by the New York Times in January. At issue is the fact that Large Language Models (LLMS) like ChatGPT require a tremendous amount of training content and data to work effectively. These media companies realize that their data has real monetary value to AI engines, and they want their cut. If your business is generating significant data, you will soon have some choices to make around if and how you want to contribute (license) your assets to AI. Questions you’ll need to answer include:? “What is my data worth?”? “Is it worth more to license to a public LLM such as ChatGPT, Google or Meta, or am I better off having a private AI model”? “How will I get ‘credit’ for your data being used? “What is the right business model?" “Will it be attributed in an AI generated answer?”? “Am I protecting my data, or can the LLMs currently access it for free?” There are additional factors and risks to weigh, risks that are compounded by the technology's novelty, the lack of precedents and uncertain success metrics. Legal and regulatory concerns, potential impacts on your core business, and financial strategies, including licensing models and liability, must be thought through. Our advice is to take a cautious approach now, but be ready to jump when the timing is right. Take a look at data protections you currently have in place to ensure they are sufficient for this new world. Look for opportunities to test and learn by contributing limited or aggregated data sets to LLMs to better understand the impact and effect. Identify specific use cases that your data can solve. Keep abreast of the evolving technology and market and be ready to enter at the right time. Reach out if you would like to get our perspective on these or your other data monetization opportunities. #datamonetization #aitraining Stage18

Okay Chris Chapman and Kristi Zuhlke I couldn't resist. Here are the survey results on whether #consumerinsights professionals agree with your post ("Research" concerns for LLM applications) and if they will embrace AI-generated survey results. ?? ?? ?? Here is the prompt I used with ChatGPT: ? Pretend you are the director of consumer insights at a company or a market research professional. You have 10+ years of experience. You are not an expert with AI, but you are both interested and somewhat concerned about how AI will change the consumer research industry and how it will impact your employment. Please take the following survey 1,000 times as if you are the persona above. Please provide the summary results in the form of average score per question and a histogram of the individual responses. Please read the passage below: [insert Chris' post here] Answer the following questions about this article. question 1: what is your role at your company? 1. head of consumer insights 2.work in consumer insights 3.perform market research for other companies question 2: How much do you agree with article above (1=completely disagree, 2=somewhat disagree, 3=neither agree nor disagree, 4= somewhat agree, 5=completely agree) question 3: which do you trust more? (1= AI generated results, 2= traditional panel generated results, 3= equally trust both, 4= don't trust either) question 4: How likely are you to use AI to generate quantitative survey results, rather than using a live panel? (1=never, 2=somewhat unlikely, 3=undecided, 4= somewhat likely, 5=definitely) And the results: ?? ?? ?? Question 2 (Agreement with the Article): The average score is approximately 4.00, indicating a general tendency towards "somewhat agree." The histogram shows a distribution of responses clustered around this value, reflecting a consensus with the concerns and observations made in the article about AI's limitations and the importance of human-generated data in research. Question 3 (Trust in AI vs. Traditional Results): The average score is about 2.10, showing a strong preference for traditional panel-generated results over AI-generated ones. The histogram reveals a significant leaning towards option 2 ("traditional panel-generated results"), with some respondents also feeling equally trusting of both AI and traditional methods. Question 4 (Likelihood of Using AI for Survey Results): The average likelihood score is around 2.54, suggesting a cautious approach towards the use of AI for generating quantitative survey results, leaning more towards "somewhat unlikely." The histogram displays a spread of responses, indicating varying degrees of openness to AI but with a general trend towards hesitation or uncertainty. And there you have it!! Stage18

B2B marketing’s most important customer? Your sales team. Stage18 #gotomarketstrategy #saleseffectiveness #accountbasedmarketing