Nik Hengel

More from our series of analytic highlights of May! Let me know what you think!

4

If you like Steve Martin, me, and/or Main Street Brand Strategy - please check out my newest post, part 2 of my series on what I learned about branding from Steve Martin, then drop us a like, comment or share which would really help us out!

10

Many people are talking about how LLMs can't count the number of r's in "strawberry." Most people say it's because LLMs aren't good at math, but that's not the issue. To understand the issue, you need to understand how LLMs work and how the internals interact with the task at hand, which comes down to: 1. Tokenization (how LLMs interpret words) 2. Training data -- Tokenization -- LLMs break words down into tokens. Tokens are small words or parts of larger words, but importantly, they're usually not broken up by letters. Why does this matter? It means they just see a word like strawberry as a single unit as opposed to the letters that make up the word, like we see it. This is efficient for processing (and efficient for training the model). See the screenshot below using the GPT-4 tokenizer. Interestingly, strawberry is a single token when there is a space before it (most of the time) and it is broken into multiple sub-tokens when it either starts a sentence or when a quote comes before it. This tokenization quirk would likely add to its confusion slightly and may be partly why strawberry is picked on. "Blueberry," on the other hand, is broken into "blue" and "berry" tokens in both cases. -- Training data -- It still knows "strawberry" is a compound word made up of the words "straw" and "berry" because that would be in its training data many times (one of those things that just comes up in conversation). But there are too many combinations of words and letters for all combinations to be in the training data enough for LLMs to learn it. Note that LLMs still count the letters in many words accurately; that's because they do have some examples in their training data, and they're pretty good at pattern matching to generalize this to other words.

34

2 Comments

Brands are still in the midst of figuring out how to best integrate #AI into their day to day business. In this @Business Reporter article, my colleague @Eda Cetinok dives into Ipsos’ GET.AI framework that helps organizations prepare for AI adoption. Check it out: https://lnkd.in/gdDAVKW5

14

Benchmark Reports are great ... but ... You never really know what a number represent. Ultimately making the benchmark report more interesting than useful How is the number calculated and more importantly: if I did the same calculation with my data - what would I get? Better or worse than the benchmark number in the report? With Dreamdata B2B benchmarks we seek to change this See your data directly next to the benchmark updated live over time !

19

3 Comments

Ever wonder why some brands can charge a premium while others get commoditized? Our new Blueprint for Brand Growth report sheds light on just that. Our analysis of 6.5 billion attitudinal and shopper data points found that brands that reach “meaningfully different” status, addressing a spectrum of practical and emotional needs in unique ways that competitors don't, have much greater pricing power, higher brand equity and much faster growth. Learn more about how brands can extend their identity and become more differentiated in this @MediaPost article: https://lnkd.in/gHRSmHgw

5

If you want to understand why Meta pivoted back to politics, it's because that's where voters are getting their news and information. In the 2024 post-election survey I did with the Center for Campaign Innovation, more voters now get their news from social media than from their local TV news. Between Facebook and Instagram, Meta has 2 of the top 3 social media platforms that voters use on a daily basis (along with YouTube). While Meta and other social media companies may have tried to downplay politics the past few years, it's clear that their users are still wanting to get news about politics and government through social media. https://lnkd.in/em79gWtp

17

2 Comments

Something I keep hearing from clients, who are looking for a new research and insights vendor, is that they need a partner who understands them and doesn't just flip back data decks. They need their insights partner to weave the data into relevant and insightful stories to have impact. Otherwise the client has to dig through mountains of data to find it themselves - and that's frustrating. This feels pretty obvious right? But the thing is, I keep hearing this over and over. It makes me wonder, how well are research agencies training their people in the art of storytelling? And how much time are research professionals practicing this craft outside of project work? There are ton of storytelling trainings out there - but I want to share one book that I come back to again and again to continually hone my skill in storytelling: The 3-Minute Rule by Brant Pinvidic. The book goes into non-conventional story flow and dissects the elements that makes your work impactful in the eye of your audience and your audience's audience. And it isn't just the same story curve we all already know about :) If your team is getting dinged on report writing and presenting data - this book is a great one to read and practice through together. #Storytelling #Insights #Impact #MRX #MarketResearch #Presentations

67

16 Comments

Join us for our next KEYS webinar as we share findings from our new Ipsos Generations Report and discuss how businesses can foster genuine connections with diverse generational groups and take the right steps to prepare for tomorrow’s population realities. https://lnkd.in/g-bFnr4g

1

Becoming an expert at distilling 50 or 60 pages of data down into a handful of key insights took me a long time. But it is one of my favorite things to do, right up there with presenting! Check out our tips in this great post, or re-watch the webinar we did last year, you might see a familiar face :D

13

“It’s common for companies to treat NPS data as gospel but it’s important to acknowledge that it can be inherently noisy.” So many companies rely heavily on NPS scores for decision making without fully understanding the data. Great article Chris Hauck, thanks for sharing.

5

4 Comments

With the acquisition of Nielsen's MMM by Circana, what does this mean for CPG brands? 1. Brands who were on Nielsen's MMM Its unclear how Nielsen's MMM will get, or will not, move to Circana's additive methodology. MMMs need to be calibrated and Circana will have a lot of work to DO to be able to see how to transition these.... should they take that route. Transitioning MMMs are hard, and explaining these numbers is even harder. 2. Brands who were on Circana/IRI MMM Likely, not much change. Whether IRI plans to use Nielsen model in the long term will be tested out for some smaller brands before they roll out more widely.... lets see. 3. Shop around These partnerships for enterprise clients are extremely long and expensive to setup. But given that you only have one vendor to pick from, its a good idea to shop around and see which other vendor can help with these transitions. I advise clients on picking the right vendor based on their business and needs. The idea that CPG brands need store level data to predict Marketing mix, while ideal, from a statistical point of view, is not that important to marketers. I have seen aggregated models do a fair job of measuring incrementality and ROIs. Most media is not run at store level, so why get stuck on these models. From a Promo POV, which is a significant investment, the Price and promo models, should be a better indicator of your trade investments. Would love to hear some thoughts on folks... #MMM #marketing #nielsen #circana

52

17 Comments

NEW data and perspective from Ipsos North America and #EffieUS showing that brands CAN do good for the world AND good for business. In fact, 74% of people in the U.S. think that brands have a DUTY to contribute to society, not just make profits. While it's not for every brand, having a societal purpose is one of the more impactful ways, I'd argue, for many brands to stand out and have meaning & resonance today. I work with so many brands that invest a huge amount of time, energy and money to try and find ways to connect with their customers on a deeper level. While at the same time, many of these same brands are divesting in their "Post-Covid era" hires and budgets focused on that 'ole brand purpose they established back in 2021. They're slowly, tiptoeing back away from it like a cartoon character hoping that no one will notice and call them out. This piece demonstrates how brands can have business success by giving their customers a reason to be optimistic about making positive change in the world. If you're thinking "This piece is so 2021" then maybe you're part of the problem. We'll keep beating this drum. Interested in learning more? Access the full report here: https://lnkd.in/gyXtaCZE Effie Worldwide Claudia Oberlin Shelley Yang Juliet Haygarth

16

2 Comments

I’d like to share an informative article on the ABC’s of Marketing Mix Modeling, written by Craig Ishill, VP of Analytics at Marketing Management Analytics (Ipsos MMA). In it he calls out the benefits MMM provides brands not just in measuring and evaluating marketing and commercial investments, but also how MMM can be used to better plan and achieve brand financial objectives, improve cross-organization understanding, buy-in and collaboration, better understand important implementation factors and the critical role data plays in increasing the effectiveness and accuracy of the models. As more and more companies adopt MMM to gain marketplace advantages in this rapidly evolving space, the article provides insightful perspectives into some of the keys to understanding MMM and how to get it right. https://lnkd.in/ehJhERgb

33

1 Comment

Ever wonder why companies don’t include the survey methodology in a press release? Simple. The majority of the time the survey is flawed and they don’t want to disclose that to the public. It’s highly likely they went with a cheap survey company and didn’t realize they made a bad decision until after the survey was completed. At that point they had a choice - do nothing with the results or release the flawed results and hide the “source” (typically defined in the press release as a “independent third party research firm” and not the actual name of the market research firm which SHOULD ADD validity). Hiding the name of the market research firm that conducted the survey is a clear indicator the results likely have errors. I feel like Roz from Monsters Inc. “I’m watching you. Always watching. Always.” PS - if you have a different reason why the methodology isn’t included or the name of the market research firm isn’t in the #PR or #marketing materials you might have worked with I would love to hear why! PPS - Roz was voiced by Pixar story artist Bob Peterson! I never knew it wasn’t a grandma until today! 😂

10

11 Comments