Chapter 10: Design for Memory, Not Just Explanation

Chapter 10 · Part Three: The Practice

Some of you have heard this one.

I’ve told it enough times that it’s in my own conference bio, and the bio openly says you’re probably rolling your eyes right now. So let me tell you the part I usually leave out, which is the part where I fail.


The setup

RSAC Innovation Sandbox. Ten finalists, three minutes each, live on stage, in front of a panel of judges who decide on the spot. For an early-stage security company it’s the single biggest thing that can happen to you. Past finalists have collectively been acquired and funded at a scale that makes the three minutes worth roughly what a Super Bowl spot costs per second.

On the Innovation Sandbox stage, 2019.
On the Innovation Sandbox stage, 2019.

We had been rejected before. Too early. That year we got in.

Dean, our CEO, was presenting. We had built the entire three minutes around him. Rehearsed for weeks. Every line tuned to sound like him, because a pitch that sounds like it was written for somebody else is a pitch a judge stops listening to.

The night before, his flight was delayed. Then cancelled.


The disaster

So I had less than twenty-four hours to rewrite three minutes of somebody else’s story into mine, and to convince the organizers to let a CMO do this at all, which as far as anybody could tell had no precedent. They agreed reluctantly, which is the correct way to agree to something like that.

I rehearsed it roughly a hundred times in a mirror.

Then at the tech rehearsal, my notes didn’t come up on the screen.

And I bombed. Badly. Badly enough that somebody in that room gave me a look I can still picture, and the look was sympathy.

So I left. I didn’t stand there grinding out repetitions of a version that wasn’t working, which I still think was the right call and which felt at the time like quitting.

And then on stage, in the real three minutes, I lost my place. Completely. Nothing there.

One second that felt like a month.


Back up

I want to back up, because before any of that went wrong, we had done one thing that made all of it survivable.

RSAC doesn’t publish a scoring rubric. It never has. What’s public is the format, three minutes plus questions, and the eligibility rules, privately held, under five million in revenue, a recently launched product you can demonstrate live. There’s no scoring sheet to download and there’s no list of weighted criteria.

So I watched dozens of past pitches instead.

And the format turns out to be brutally formulaic. Everybody answers the same five questions, in roughly the same order, with roughly the same emphasis:

  1. What’s the problem?
  2. Why is it big?
  3. What’s genuinely new about your approach?
  4. Why are you the team to do this?
  5. Can you survive if the floodgates open?

Everybody hits the rubric. Which means hitting the rubric wins you nothing.

That’s the observation the whole chapter rests on, and it generalizes far past a pitch contest. Any scored situation where the criteria are known and the candidates are competent converges on identical answers. The criteria stop being a differentiator the moment everybody can read them. They become a floor, and standing on the floor isn’t a position.

So the question stopped being “how do we cover the requirements” and became:

What will a judge still be able to repeat tomorrow?


The Camry

The honest problem was that asset management is the least exciting thing in security. Everything else in this industry gets to be a Lamborghini. We were selling the ability to know what you have and whether it’s actually protected, which is foundational and unglamorous and doesn’t photograph well.

So instead of arguing that we were exciting, I said this:

The Toyota Camry is the most sold car in America because it fulfills the promise. It gets you from point A to point B. Every time.

Nobody has a poster of a Camry on their wall. And every single thing you want to do that day depends on it starting.

The Camry, on the RSAC stage. Nobody has a poster of it.
The Camry, on the RSAC stage. Nobody has a poster of it.

What came back

So. Standing on that stage. Nothing in my head.

The thing that came back was the Camry.

I wrote that line so a judge would remember it the next day.

Turns out I was the one who needed it.

I finished at exactly three minutes. And we won.

RSAC Innovation Sandbox 2019: the Axonius pitch. The three minutes this chapter is about.

The proof, which isn’t the win

A Fortune 500 company called before I was off the floor. They had watched it, they wanted a meeting immediately, they became a customer, and a POC started shortly after. That’s the outcome, and it’s the least interesting part.

Here’s what I actually want you to notice.

Years later, people still walk up to me at industry events and say: you’re the Toyota Camry guy.

Nobody has ever walked up and repeated our value proposition.

We spent real money and real years on that value proposition. It was accurate. It was reviewed. It was approved by people whose judgment I respect. And its total lifetime recall, measured in strangers at conferences, is zero.

The line about the car cost nothing, took an afternoon, and is still in circulation.

Memory ran in both directions on that stage. Externally, strangers retained it for years. Internally, under maximum pressure with my mind completely blank, the only thing that held was the line I had built to be memorable. I’ve never had better evidence that this property is real and that it’s engineerable.

Why it stuck

There’s a reason it worked, and it’s one of the oldest findings in memory research. In 1971, Allan Paivio described what he called dual coding. Concrete words get stored twice, once as a word and once as an image. Abstract words mostly get stored once. So concrete things are remembered far better.

“Asset management” is abstract. A Camry is a picture.

Look at your own value proposition. I’d bet it’s abstract. Every one in this category is, mine included. Platform. Visibility. Outcomes. Nobody can picture any of it, so nobody repeats any of it. Find your Camry. It costs an afternoon.

The judge

One more piece, which I cut from the stage version because it teaches a different lesson.

The judge assigned to us came to the booth afterward for the demo. I didn’t know who he was. The demo went well. I found out later that he was Ashim Chandna of Greylock.

If I had known that walking in, I’d have been more nervous and I’d have given a worse demo. I’m certain of this.

I don’t have a clean method to extract from it. I include it because it’s true and because the version of this story where every good outcome was engineered is a lie. Some of it was that I didn’t know enough to be frightened.


How to derive a rubric nobody published

The transferable skill from the first half of this chapter, and it applies to any situation where you’re being evaluated against criteria that exist but aren’t written down. A pitch contest. An analyst evaluation. A vendor shortlist. A category review. An interview loop.

  1. Collect the artifacts. Past winners, past submissions, past reports, past decks. Ten is enough. Twenty is better. They exist more often than people assume, and the reason nobody looks at them is that watching twenty pitches is boring and looks like procrastination.
  2. Extract the questions, not the answers. For each artifact, write down what question each section was answering. You’re reverse-engineering the form from the responses.
  3. Count. The questions that appear in every artifact are the real rubric. The ones that appear in the winners and not the losers are the interesting ones and there are usually fewer than you hope.
  4. Establish the floor. Write the competent version of every answer. This is what everybody will do. You must do it too, because failing the rubric disqualifies you.
  5. Then budget the remainder for one thing. Satisfy the rubric as efficiently as possible, quietly, and spend everything left over on the single element designed to be repeated. In three minutes that was about twenty seconds. Twenty seconds is enough.

The mistake almost everybody makes is spending the entire allocation on the rubric, answering all five questions beautifully, and finishing with nothing left over. That produces the best possible version of the thing everybody else is also producing, which is Chapter 2 again, in a different costume.


How to build a Camry

Analogy construction as a discipline rather than a flourish. Five rules, in the order they matter.

1. Source it outside your category. Every analogy available inside cybersecurity is already in use inside cybersecurity, which means it arrives with fluency attached and can’t produce a new thought. Cars, kitchens, sports, construction, medicine, airlines, weather, plumbing. The further from security, the harder it is to build and the better it works.

2. Make it ordinary. The Camry works because it’s boring. An analogy to something glamorous flatters you and teaches nothing, and the audience notices the flattery before they notice the point. If your analogy makes your product sound heroic, you’ve written an advertisement and it won’t be repeated, because nobody repeats an advertisement.

3. It has to survive the second question. A good analogy holds when somebody pushes on it. A bad one is a simile that collapses the moment anybody asks what the equivalent of the engine is. Before you use it, have somebody hostile ask you three questions about it and see whether it bends or breaks.

4. Test whether it explains or decorates. Remove the analogy. Is the point still clear? If yes, the analogy is decoration, which is fine but isn’t this. A working analogy does load. The Camry is doing the actual argument: foundational, unglamorous, and everything depends on it.

5. Be slightly self-deprecating. The Camry line concedes that we aren’t the exciting one. That concession is what buys the claim underneath it. An analogy that claims a flattering comparison gets discounted on arrival by a skeptical audience, and yours is a professionally skeptical audience.

And then use it for years

The failure mode after you build one isn’t that it stops working. It’s that you get bored of it roughly nine months before your market has heard it once.

Internal boredom is the single most expensive emotion in marketing. Every time you change the line because the team is sick of it, you reset the clock on the only asset that compounds. The Camry survived because I said it constantly, in every context, past the point where it embarrassed me.


The test

The repeat-back test from Chapter 9 is the instrument for this chapter. Tell five people. Wait a week. Ask them what you do.

If what comes back is your analogy, you’ve built one.

If what comes back is a paraphrase of your category, you’ve written a description, and a description is a thing people read once and never transmit.

Design for memory, not just explanation.

Explanation is what you owe the person in front of you. Memory is what reaches the four people who weren’t in the room, and in a market where the buying window is ten minutes long and opens without warning, those four people are the entire business.


The Comfort Check

Chapter 10: the three-minute pitch

1. The comfortable assumption. You win a rubric by answering the rubric well.

2. The true objective. Be the one pitch a judge can still repeat tomorrow.

3. The redesigned approach. Satisfy the rubric quietly. Spend everything left on one analogy.