A Value-First Guide to Surviving the SaaSpocalypseContents
Part II — Context
Chapter 4
The Missing Ingredient
There's a conversation going on in boardrooms and Slack channels and all-hands meetings. It happens in nervous one-on-ones between CEOs and their technology leaders, and on conference stages.
It goes: "How do we add AI?"
It's the wrong question.
Not because AI doesn't matter. It matters a great deal. But "how do we add AI" assumes AI is a feature you bolt onto the operation you already have. Like adding a turbocharger to a car. Same car, more power.
The better question is: "What context have we built that AI can work from?"
Because AI without context is a party trick. Impressive in a demo, useless in production. The difference between an AI that changes how you work and an AI that produces plausible-sounding nonsense isn't the model. It's the context.
What context means
Context is not data. This distinction matters more than almost anything else in this book, so let's be precise. Here's the progression.
Data. Customer X bought Product Y on Date Z. Contact A opened Email B. Facts without meaning.
Information. Customer X has bought three products over two years. Contact A opens most of what you send. Data organized and added up, but still flat.
Knowledge. Customer X is expanding and may be ready for a bigger conversation. Contact A keeps reading your integration material. Information interpreted. Meaningful, but static.
Context. Customer X expanded because their new leader came from a company that used your largest plan, and the expansion landed right when their budget year turned over. Contact A's interest in integration started after your session on connecting systems, which suggests the kind of ERP complexity that's common for companies like theirs at their stage. Knowledge placed inside relationships, over time.
If that progression looks like the old data, information, knowledge, wisdom pyramid, it's a cousin of it, with context where wisdom used to sit.
Data tells you what happened. Information tells you what happened in total. Knowledge tells you what it might mean. Context tells you what it means, why it matters and what to do about it, because it connects the dots across time, across relationships and across everything going on around each data point.
An AI working on data can produce statistics. An AI working on information can produce summaries. An AI working on knowledge can produce reasonable recommendations. An AI working on context can produce intelligence: the specific, situated understanding that changes how you serve a customer.
Our methodology says the same thing in its own words. "Context is more than data," the reference says in describing the Unified Strategy View. "It comes in three layers: the data; the documentation and strategy; and the memory people never wrote down." And one of the five claims The Value Creation Protocol opens on is called Context as Substrate: "Context is not an input to operations — it is the substrate on which operations run." The handbook's first chapter teaches it.
The AI demonstration problem
If you've been to a technology conference in the last couple of years, you've seen the demo. An AI is connected to a company's CRM. The presenter says, "Show me everything about this customer." The AI produces a summary: contact details, purchase history, open tickets, last interaction. The room is impressed.
Look closely at what it produced. It summarized data that was already in the system. It organized facts a person could have found by clicking through a few screens. Faster, yes. Not different. It didn't produce context. It produced a prettier dashboard.
Now ask a different question: "What should I know about this customer before my call tomorrow?"
The demo AI gives you the same summary again.
An AI that works from context gives you something else. It tells you about the support ticket from last month and whether it's really closed. It tells you who on their side has gone quiet. It tells you what has changed in how they use what you sold them, and what happened the last time a relationship looked like this one. It hands you the conversation, not the record.
The difference isn't the AI. It's what the AI could see: the customer's whole story in one place, their patterns over time, and what has happened in relationships like theirs.
The context imperative
So here's the principle, and the rest of this book leans on it: your AI strategy is your context strategy. They're the same thing.
If you invest in better AI models but don't unify your context, you'll get prettier summaries of scattered data. If you unify your context, even with modest AI, you'll get intelligence that changes outcomes. Chapter 6 makes the case for why that advantage lasts, and why the companies that make the models have started acting as if they agree.
Where context lives today
So where is your context right now? Here are the four places it tends to hide, and none of them is safe.
In people's heads. Your best salesperson knows the CFO prefers email, is wary of long commitments and responds to a well-supported business case. Your veteran engineer knows the "routine" error message means something deeper. When people move on, what they knew moves with them.
Across systems. Your CRM has some of the context. Support has different context. Marketing has different context again. A salesperson preparing for a renewal would need five or six systems to see the whole picture, so it's tempting to skim the CRM and wing it. Putting the context together costs more time than anyone has.
In "tribal knowledge." "Talk to the person who handled their implementation." That's the sound of context failing to become the organization's. What that person knows is valuable and fragile. When they change roles or have a busy week, it's gone.
That last one is worth turning around. "Tribal knowledge" is the name for the risk: expertise stuck in someone's head, lost when they leave. The same expertise, written down in a form that people and AI can both use, is what our methodology calls Human Domain Expertise, an asset the organization keeps and builds on. The AI-Native Shift is built around making that turn.
In the reasoning behind decisions. Your CRM shows that you shifted toward manufacturing customers. The reasons, the market read, the board conversation, the customer feedback, live in a recording nobody rewatches, a slide deck in someone's drive and the memories of five people.
The context paradox
Chapter 3 ended on the paradox: organizations built just-in-case infrastructure that broke up the very context they need for just-in-time operations. Here's what that means in practice.
You can't fix it by adding another tool. An AI chatbot bolted onto a scattered system gives scattered answers. A business intelligence platform querying disconnected databases gives disconnected reports. A data lake that takes in everything from everywhere gives you a pile of information that's technically in one place and practically useless. The scattering is the problem. Another fragment doesn't solve it.
What solves it is unification. Not putting everything into one system for the sake of tidiness. Unifying the context, so that anyone, person or AI, who needs to understand a customer, a relationship or a situation can see the whole picture in the place they already work.
That's what the next chapter is about.