Connecting the 5 Ps of Marketing Data: A recipe for gen AI adoption & value
Part 2 in the series on scaling generative AI in enterprise marketing organizations
ICYMI, my last article explored the idea of the U-shaped excitement curve and how it's a great tool for gen AI program leads to drive organizational trust and aid adoption efforts.
Today, I'm exploring another foundational pillar that helps drive adoption but also unlocks the real value of generative AI: Connecting the 5Ps of marketing data.
At this point, you've already used gen AI for the basic "no regrets" use cases. Point it to a long-form case study and out comes a blog post (with some editing) on your corporate blog. Instant productivity and speed boost.
But there is one problem...
❌ The blog post might never have been needed
Does similar content exist already? Does it use the latest company messaging? Are we overloading the audience when looking across all of what the company is publishing at the same time? Question. Question. Question. Question.
This is why more context isn't just helpful — it's essential.
The real power of generative AI? Speed to Insight
Driving value without context is a challenge. Yes, gen AI can create content at speed — but to create differentiated, impactful marketing, it needs to understand the who, what, where, why, and how of your business and your audience.
Let's use an analogy: The chef
Their goal? Make exceptional food that people will love and come back for. Do they have a recipe? Yes. Is that all they rely on? Absolutely not.
The masters watch the color change as the steak sears, listen for the rhythmic bubbling of a sauce that tells them it's reducing properly, smell the browning butter, feel the texture of the ingredients being stirred, and taste the food to balance the flavors.
Each sense is a context clue and ultimately a checkpoint that allows them to improve what makes it to the table. Can they cook without all 5 senses? Yes. But the precision, nuance, and quality are likely compromised. No Michelin star rating here...
It's the same today for marketers. The question is - what are the equivalent senses a marketer needs to be successful in the world of gen AI?
The answer: the 5 Ps of marketing data (not to be confused with the 5Ps of marketing)
What are the 5 Ps of Marketing Data?
Plan Data – The strategic blueprint: campaign goals, audience segments, content calendars, key themes...essentially all the metadata of the marketing plan.
Product Data – Your business offerings: core messages, positioning, pricing, product details, and competitive differentiators. Also, your "content product" parameters like channel and format specifics, quality parameters, brand and writing style guides, etc.
People Data – The skills and experience of your people and their current workload/allocations - both inside your marketing org and across external agencies, freelancers, consultants, etc.
Process Data – The operational engine: specific business processes and workflow steps, historical project timing, real-time project status, etc.
Performance Data – the results for all activity across all marketing goals and objectives. e.g. engagement, conversion, attribution, etc.
Caveat: External data sources can bring even greater context clues - e.g. competitive messaging, social sentiment, earnings, etc. (a sixth sense?). Here, I'm just focused on your secret sauce - your proprietary marketing data.
Let's bring it to life with an example...
Imagine you are the content lead for a large B2B enterprise launching a new service offering. You are tasked to create and promote a blog article that helps contextualize how businesses can use the new service.
⚠️ Scenario 1: One Data Source (1P)
Your AI assistant is plugged into your performance data only. Here, the marketer generates an article with some basic prompts around the service description plus some key client challenges it solves.
Is it helpful? Yes. You get back a first draft of the content in a structured format that has proven to be highly engaging overall and for the audience segment(s) you are targeting.
So why do we need the other 4 Ps? Here's a literal example I've seen unfold...
The AI draft comes back, and it sounds generic and/or similar to what your competitors are saying (no product data).
The craft experts (writer & designer) tasked with amping up the storytelling quality have other priority projects and are not available to get started when you need them, driving a delay (no people data).
It wasn't clear for new service line launches that the CMO was in the approval chain - and now they are in the middle of earnings prep which clearly takes priority over reviewing your article (no process data).
With the delays, your launch window is now in the same window as a very large, newsworthy initiative the brand team is driving, and it will include your audience segments. But their efforts will dominate customer, executive, and employee communications...minimizing the possibility of your content hitting your engagement goals (no plan data).
🔄 Scenario 2: 3Ps of data
Now your AI platform is connected to the Plan, Product, and Performance data. Same scenario as above. But this time…
It recommends a launch date where you are not being outshouted by other company content/messages (plan)
The content draft is superior as it weaves in your unique messaging, differentiators, and style (product)
It's drafted using a particular article format proven to be a high-performer for your audience segment (performance)
Far better...but without the process and people data, execution still lags as the missing execution avails and governance approval steps drag the timing out.
✅ Scenario 3: All 5 Ps are connected
Ahhh, nirvana...
Realistic planned publish dates with whitespace in the editorial calendar. Proven tactics and formats, powered with relevant, differentiated content for the audience. Execution resources ready to go and approvals from all the right people.
AI becomes a truly seamless collaborator with the marketer who is then able to - pardon the pun - cook.
When all 5 Ps are connected, you're not just speeding up execution — you're gaining speed to insight. You understand what to do, why it matters, and when it will have the greatest impact.
So what's holding people back?
Data readiness isn't a new concept - but with gen AI it's exposed so quickly because non-techies rapidly experiment and immediately see where they can't trust it... and that really will slow adoption.
Here is some current research data on the state of data readiness today...
65% of C-suite executives cite building an end-to-end data foundation as a top obstacle to scaling generative AI. Accenture's "New Data Essentials"
61% of C-suites report that their data assets are not ready for generative AI yet, and 70% find it hard to scale projects that use proprietary data. Accenture's 2024 research
Lots of tools with lots of data - the average marketer uses 8 different tools/technologies in their day, creating the conditions for incomplete, incorrect, or missing data. 2025 Salesforce State of Marketing
It's not easily accessible...while over half of marketers say data is available in real time to execute a campaign, 59% need the IT department’s help to do so. 2025 Salesforce State of Marketing
Only 30% of marketers say they have a completely unified view of customer data across channels. This fragmentation directly impairs AI performance. 2025 Salesforce State of Marketing Report
Let's ___ Go!
From my experience, conversations with industry leaders, and available research, it's clear most organizations are still in the early stages of this journey. What's also clear - the ones getting the data right quickly are going to accelerate their competitive advantage.
If you’re leading a marketing org, especially in a G2000 company, there are some immediate next steps that you can take to get on the path to scale and impact...
Map your 5 Ps – Do an audit. Where does each type of data live? Who owns it? Is there a governance model for it? How clean is it? Are the taxonomies similar? Don't forget gen AI can help with the audit itself! It can tell you if there is a common taxonomy and where the delta is. It can highlight where data is missing, or if it's there are cleanliness issues.
(re) Design your data structure for generative AI-readiness – Clear metadata, accessible repositories, consistent taxonomies. Unstructured data is a great source of insight, but you will still need to create the common core approach to ensure a level of accuracy exists and so that the early gen AI skeptics in the org are more comfortable (traceability).
Start small: Goal is 2-3 well-formed "Ps" in year 1 - if your org has a massive martech stack, chances are your audit will reveal a ton of work is needed to bring your data supply chain into good order. My suggestion - less is more. Focus on a clean, accurate, integrated set of Ps that you can trust and prioritize use cases for gen AI around them. I've personally found that plan and performance data are super valuable as a paired set, although Product and Performance data are typically the most well-formed at the start given they are the closest to business outcomes for both sales and marketing. People data will be the hardest given all the security, privacy, legal, and HR considerations.
Final thought
When marketers connect the 5Ps of marketing data, they’re like world-class chefs in full command of their kitchen — not just cooking quickly, but cooking intentionally, with deep insight into timing, ingredients, tools, and the guests they’re serving.
Get it right, and you’re not just accelerating marketing. You’re reinventing it.
If you want to chat about the 5Ps or any other aspect of scaling gen AI in a marketing org, email me or follow me on Linkedin.
