I help businesses to get 40% more out of their Martech budget | Martech & MarketingOps Guide | Board advisor | Co-Publisher of MartechMap.com | Speaker & Business University Lecturer | Author
3 reasons why your company needs a BrandLLM. The content supply chain is broken. Many brands suffer from a split personality due to organizational silos. With Generative AI we have a chance to fix that. Here's an idea. Enter the BrandLLM. Here are the three reasons (more in the article). 1. A BrandLLM becomes your company’s most valuable asset Because... 90% of S&P companies' value comes from intangible assets (brand, IP, patents), not from inventory and buildings. 2. A BrandLLM ensures compliance and consistency across markets Because... it aligns with local laws, governance & privacy standards, product variations, and brand guidelines. 3. A BrandLLM unlocks unprecedented personalization Because... it offers the potential for real-time mass-customized visuals and copy. Read my latest article on MarTech.org. I am covering... - What is a BrandLLM? - How to build a content layer with generative AI - How to create BrandLLMs - What your BrandLLM should cover A huge thank you to Rasmus for helping to craft the concept! Link to the article in the comments. #Martech #brand #LLM #AI #MarketingAutomation #DigitalMarketing
I wonder if the bigger issue isn't just about consistency in content creation, but actually creating content worth consuming in the first place. As AI agents increasingly filter and summarize marketing content, simply having more consistent automated content won't help us break through. I'd argue our goal should be creating content so valuable that it can't be automated or easily summarized: ☑️ Original research based on unique company data (like what Gong does with sales conversation analysis) ☑️ Deep insights from actual customer experience ☑️ Real thought leadership that advances industry thinking ☑️ Content so valuable people would pay for it (then give it away free) The opportunity is to use AI to automate the tactical work so our teams can focus on developing truly differentiated insights and valuable content that both humans and AI agents will want to surface.
👍 Nice Frans. I've been working with a similar organization - including via early in-house software - similar to your list but with slight tweaks: You mention: - Brand level - Campaign level - Collateral level - Content level Work in progress, but I'm finding this helpful: - Brand level - Audiences & Positioning level - Campaign level - Asset level - Brand level includes your "collateral level" assets. Ie, brand building blocks: from color pallet, logos, as well as building block assets / universal brand assets like brand imagery and universal templates. - I guess you have equivalent of Audiences & Positioning level outside of Content Layer and inside the Data Layer? Audiences of course are core to existing Data Layers, but this seems less so: the "what are we helping them get done and why it's better?" positioning that ties directly to campaigns. And brand can help multiple segments get multiple jobs done. So each of these audience positionings have their own set of campaigns below. WIP. - Campaign level includes source input data + collections of the assets below. - Asset level includes campaign / content assets - including editable config data (not just the final output/render.) Thoughts?
Great post Frans. I've been championing applying the order of data to the content world since I started at Merkle. I see someone already mentioned Adobe's GenStudio which promises much of what you describe. The two considerations I'd throw in. 1. The human-machine partnership with AI and what I describe as the level of copiloting. AI adoption is often looked at through the lens of AI doing it all. In reality much of today's value is from copiloting or working with AI. With that in mind the nature of that relationship and balance of activities becomes very nuanced between use cases, companies and even at an individual level. The adoption and change management considerations are key to success. 2. The complexity of plumbing it all together and having the UI which is meaningful and usable my business users. A reason why there is still some way to go.
Check out what Adobe are doing with the #GenStudio and #ExperienceDecisioning areas... 👌
And in Europe, you need this all the more for each country you want to localize your content to. Input these LLMs with local culture, memes, and more so that you can audit your content to see if it aligns with local interests and opportunities. Not just the regulations. The LLM should be constantly fed with insights and feedback from your campaigns to help you create content that just works well locally.
I ❤️ the concept of a Brand LLM. Two components that I think are critical to have in a Brand LLM are: (1) An understanding of the customer's jobs-to-be-done (aspirational, functional, emotional, and social) by targeted segments (2) An understanding of the experiential outcomes that a brand is seeking to create along the customer journey Why? Orchestrating experiences that create Brand Connection required an understanding of these two critical components as inputs. My thoughts. Thanks for the insightful post!
It has been a nice thought provoking read. As always a lovely concept! Looking at the direction many vendors of DAM’s, DXP’s, CMS etc. are heading, I trust this will soon become reality for many. E.g. asset variants can already be generated from master and with the introduction of the LLM the variants are becoming smarter. Using data from decision/personalisation engines as input for this process seems like the logical next step. Especially for those vendors that also offer decisioning capabilities. Either from their own platform or through (MACH) alliance. Looking forward to it and the next levels mentioned.
Great overview, there’s so much for organizations to learn in order to understand the potential of a brand LLM. In my experience it’s been easier for leaders to understand ROI when it comes to building the data layer unlocking better segmentation, but much harder to do the same when debating investment into the content layer. Great to see a clear overview of its benifits here.
Frans Riemersma is really really interesting. Do you have any Demo or so to see how this plus Data Layer and Content Layer work in practice? I have a couple of projects at the Moment facing exactly what you describe and i immediately through: yes this might be the Solution!
I help businesses to get 40% more out of their Martech budget | Martech & MarketingOps Guide | Board advisor | Co-Publisher of MartechMap.com | Speaker & Business University Lecturer | Author
1wMartech.org - https://martech.org/why-your-company-needs-a-brand-llm-to-thrive/