The Known-Unknowns Matrix In Ecommerce

If you’ve ever worked on a high-stakes product team, or in military operations, you’ve probably heard of the Known-Unknowns Matrix: a risk assessment model splitting knowledge into four phases. This matrix is appropriated from a psychological theory known as the Johari Window: a mental model addressing interpersonal relationships.

The DTC realm overlaps nicely with both of these constructs: brands are trying to build sustainable relationships based on knowledge of their consumer market, while internal and external influences create challenges along that path to sustainable success. Within Fairing, this Ecommerce Known-Unknowns Matrix is an idea we reference often when we discuss the importance of direct-from-consumer data.

Explaining The Ecommerce Known-Unknowns Matrix

The four phases of the matrix can be explained like so, using a few examples pulled directly from our conversations with DTC brands:

Known-Knowns: Knowledge – I know what I have

• We generated $2MM in revenues last month.
• 26% of purchases came from TikTok according to our post-purchase survey.
• GA reported a 31% increase in visits to this PDP week-over-week.

Known-Unknowns: Awareness – I know what I need

• Where can I get a second opinion on my Facebook-reported ROAS?
• Which of my customers are buyers and which are users?
• What generation does our customer identify with?

Unknown-Knowns: Bias – I don’t know what I have

• There’s no way to find the source of Direct traffic.
• Facebook’s reporting works well for us; we don’t need an additional tool.
• We’ve basically maxed out our conversion rate on this landing page.

Unknown-Unknowns: Ignorance – I don’t know what I need

• Will our MER dramatically increase next month?
• When will this customer repurchase?
• What will happen to us if we lose our supplier?

You might notice some interesting distinctions here, which we’ll get to in a moment. For now, it’s also worth noting that a brand could think of this structure as “who should I ask?” when pursuing knowledge:

• Known-Knowns Examples (KK): you don’t need to ask anyone, because you hold the knowledge. A good example here would be dashboards from last month you’ve already reviewed, like first-party data analytics or store revenue metrics.
• Known-Unknowns Examples (KU): what is Known Unknown? You ask others — your team, your customers, your tools — for the knowledge you’re aware that they have. Zero-party data and first-party data both live here, whether it’s asking a team member to pull data from a platform they manage, or surveying consumers for a simple data point only they would know. Knowing how to navigate from known to unknown is essential for gathering accurate data. To learn more about zero-party data vs first-party data, read our other article.
• Unknown-Knowns Examples (UK): you don’t know to ask others, because you assume to have knowledge. The bias here can only be avoided by asking people to challenge your own assumptions. Think: analytics experts who can reveal your misinterpretations of data, or consumers who can tell you something you wrongly believed about your market.
• Unknown-Unknowns Examples (UU): you can’t ask anyone, because nobody knows. In these situations, you’d have to dismantle the need for knowledge into more accessible pieces, attempting to gather enough information to mitigate or reframe the unknown.

Of course, we’d all love to sit squarely in the Known-Knowns corner all day long, running our businesses with total omniscience. But that urge is precisely what lures us into Unknown-Known blind spots, or stops us from critical thinking through Unknown-Unknowns, or causes us to deprioritize investigating Known-Unknowns.

Direct traffic serves up a perfect illustration for the Ecommerce Known-Unknowns Matrix. If we’re being honest as marketers, many of us simply assume Direct traffic is an unsolvable mystery, and operate under that assumption. 10%, 20%, 30%... it doesn’t really matter what % of your sales are attributed to “direct”, right? Because there’s nothing you could do with that information anyway. “What do I do with this Direct traffic?” is an unknown-unknown.

Now, what’s really going to bake your noodle is the relativity of this whole matrix. To you, Direct traffic may be an unknown-unknown; an ever-present anomaly on your books. But to a more experienced marketer who has awareness of the tools and processes that can address this gap in the data, Direct traffic is a known-unknown that can, at least in part, become a known-known just by implementing a few resourceful solutions.

And of course, the stinger here is that the more experienced marketer looks at what you’re doing and realizes something you don’t: that you’re actually stuck in the unknown-knowns quadrant. In other words, your assumption about Direct traffic is the blind spot, not the data.

Said another way, the same question can fall into different corners of the matrix depending on whom you ask. If the question is “how much of my revenue comes from podcast ads?”, a novice marketer might say the answer is an unknown-unknown and revert to measuring incremental sales lift whenever the ads run. But an Fairing partner like Right Side Up would tell you it’s a known-unknown — you just need to apply a measurement methodology using some key tools. Once you’ve done that, a brand like Füm would say it’s a known-known, sitting right there in the marketing analytics dashboard.

That’s a world of difference, and one that extrapolates to the entire business. A brand run by an overconfident self-proclaimed “expert”, who hires a bunch of yes-men and doesn’t talk to customers, is going to be dragged down by the gravity of their own Unknown-Knowns bias. These ecommerce brands think they’re winning, and it’s hard to convince them otherwise until something catastrophic shakes their thinking loose.

This “z-pattern” of progress towards knowledge is fairly natural, but it’s not at all optimal. Instead, you want to avoid Unknown-Knowns as much as possible… though that doesn’t mean your learning path will be linear.

So how do you know you’re moving in the right direction at any given moment? In large part, this comes from maintaining a team mindset of continuous learning, curiosity, humility, and experimentation. The Unknown-Knowns corner is where you get stuck when those traits atrophy — and whether novice or expert, we all have our blind spots which reside in that quadrant. Squashing them should be both a personal and professional goal.

We like to say data doesn’t lie until you start asking it questions... that’s when your team’s curiosity and resourcefulness will influence where you end up on the matrix. Knowing when, where, why and how to ask questions is your key, and perhaps that’s a bit of a brain-twister at first pass: the Known-Knowns quadrant expands when you obsess over every other quadrant.

More Real-World Matrix Examples

To get a more actionable feel for the Ecommerce Known-Unknowns Matrix, let’s touch on a few real-world paths to knowledge using direct-from-consumer data:

Personalization

1. UK: “Our core audience is college kids, so we’ll focus creative on that demo.”
2. KU: “But, maybe we should ask customers what generation or lifestyle they identify with?”
3. KK: “Turns out 71% of our customers identify as millennials, regardless of age.”

Channel Diversification

1. UU: “What media channels should we invest in next?”
2. KU: “Well, what ‘Other’ responses do customers write into our "How did you hear about us" survey?”
3. KK: “We’ve found customers are increasingly hearing about us from influencers on this new video platform that’s invite only.”

Revenue Analysis

1. KK: “Our AOV is reported as $55.”
2. UK: “So, our typical customer spends around $55 per order.”
3. KU: “I wonder, are gifters and event buyers skewing that AOV figure?”
4. KK: “We asked customers who the purchase was for, and apparently buyers & gifters are a totally different revenue tier, generating 2x the AOV of users.”

Consumer Insights

1. UU: “What will happen if we drop our fabric supplier who’s overcharging us?”
2. KU: “Where does ‘quality of fabric’ sit on our customers’ reasons for buying?”
3. KK: “The customers we surveyed had fabric quality ranked last, behind 4 other reasons for buying.”

Competitor Research

1. UK: “I’ve never heard of someone cross-shopping us against Competitor X.”
2. KU: “Hmm… how many of our customers replaced Competitor X with us?”
3. KK: “Actually, 20% of our buyers reported replacing that competitor with our product.”
4. UU: “So how many of our potential customers are buying Competitor X instead?”

In truth, what the Ecommerce Known-Unknowns Matrix reveals is that you’re never done learning… but each question broadens your perspective, and consequently, your potential for predictable and sustainable growth.

Taking Action On The Known-Unknowns Matrix

• Invest in always-on solutions for asking questions and testing assumptions at speed & scale.
• Keep your opinions loosely held, and give teammates a clear onramp to challenging status quo ideas.
• Build your data advantage on ground truth sources, instead of renting customer relationships and letting third parties grade their own homework.
• Challenge your team to break Unknown-Unknowns down into problems that can be tackled productively.
• Avoid stacking assumptions — if you have a metric that’s been daisy chained from a string of 3 other metrics, you might be looking at an Unknown-Known.
• Be critical of the questions you ask of your data. Too often, the question is where the limitation lies.
• Listen to the people and tools that have seen a world outside of your own.
• Don’t force data into models just because the models are rigid. The models exist to serve the data, not the other way around.

We hope that this article has helped you better understand the known and the unknown. Now that you’ve seen the Ecommerce Known-Unknowns Matrix up close, what examples come to mind in your own experience? Drop us a line; we’re always interested in hearing what brands learn from challenging their assumptions.