But I’m not talking about elaborate experiments; I’m talking about crude ones. Not simple ones, crude ones.
We were taught the best experiments maximize learning, but that’s dead wrong. The best experiments are fast, and the best way to be fast is to minimize the investment.
In the name of speed, don’t maximize learning, minimize the investment.
Let’s get right to it. One of the best tricks to minimize investment is to minimize learning – learning per experiment, that is. Define learning narrowly, design the minimum experiment, and run the trial. Learning per trial is low, but learning per month skyrockets because the number of trials per month skyrockets. But it gets better. There’s an interesting learning exponential at work. The first trial informs the second which shapes the third. But instead of three units of learning, it’s cubic. And minimizing learning doesn’t just half the time to run a trial, it reduces it by 100 or more. It’s earning to the hundredth power.
Another way to minimize investment is to minimize resolution. Don’t think nanometers, think thumbs up, thumbs down. Design the trial so the coarsest measuring stick gives an immediate and unambiguous response. There’s no investment in expensive measurement gear and no time invested in interpretation of results. Think sledgehammer to the forehead.
A third way to minimize investment is to evaluate relative differences. The best example is the simple, yet powerful, A-B test . Run two configurations, decide which is better, and run quickly in the direction of goodness. No need to fret about how much better, just sprint toward it. The same goes for trial 1 versus trial 2 comparisons. Here’s the tricky algorithm: If trial 2 is better, do more of that. And the good news applies here too - the learning exponential is still in play. Better to the hundredth power, in record time.
I don’t care what norms you have to bend or what rules you have to break. If you do one thing, run more trials.
But don’t take my word for it. Dr. Seuss had it right:
And I would run them in a boat!
And I would run them with a goat…
And I will run them in the rain.
And in the dark. And on a train.
And in a car. And in a tree.
They are so good so good you see!
We want to be more innovative. That’s a good vision statement, but it’s not actionable. There are lots of ways to be innovative, and it’s vitally important to figure out the best flavor. Why do you want to be more innovative?
We want to be more innovative to grow sales. Okay, that’s a step closer, but not actionable. There are many ways to grow sales. For example, the best and fastest way to sell more units is to reduce the price by half. Is that what you want? Why do you want to grow sales?
We want to be more innovative to grow sales so we can grow profits. Closer than ever, but we’ve got to dig in and create a plan.
First, let’s begin with the end in mind. We’ve got to decide how we’ll judge success. How much do we want to grow profits? Double, you say? Good – that’s clear and measurable. I like it. When will we double profits? In four years, you say? Another good answer – clear and measureable. How much money can we spend to hit the goal? $5 million over four years. And does that incremental spending count against the profit target? Yes, year five must double this year’s profits plus $5 million.
Now that we know the what, let’s put together the how. Let’s start with geography. Will we focus on increasing profits in our existing first world markets? Will we build out our fledgling developing markets? Will we create new third world markets? Each market has different tastes, cultures, languages, infrastructure requirements, and ability to pay. And because of this, each requires markedly different innovations, skill sets, and working relationships. This decision must be made now if we’re to put together the right innovation team and organizational structure.
Now that we’ve decided on geography, will we do product innovation or business model innovation? If we do product innovation, do we want to extend existing product lines, supplement them with new product lines, or replace them altogether with new ones? Based on our geography decision, do we want to improve existing functionality, create new functionality, or reduce cost by 80% of while retaining 80% of existing functionality?
If we want to do business model innovation, that’s big medicine. It will require we throw away some of the stuff that has made us successful. And it will touch almost everyone. If we’re going to take that on, the CEO must take a heavy hand.
For simplicity, I described a straightforward, linear process where the whys are clearly defined and measurable and there’s sequential flow into a step-wise process to define the how. But it practice, there’s nothing simple or linear about the process. At best there’s overwhelming ambiguity around why, what, and when, and at worst, there’s visceral disagreement. And worse, with 0% clarity and an absent definition of success, there are several passionate factions with fully built-out plans that they know will work.
In truth, figuring out what innovation means and making it happen is a clustered-jumbled path where whats inform whys, whys transfigure hows, which, in turn, boomerang back to morph the whats. It’s circular, recursive, and difficult.
Innovation creates things that are novel, useful, and successful. Novel means different, different means change, and change is scary. Useful is contextual – useful to whom and how will they use it? – and requires judgment. (Innovations don’t yet exist, so innovation efforts must move forward on predicted usefulness.) And successful is toughest of all because on top of predicted usefulness sit many other facets of newness that must come together in a predicted way, all of which can be verified only after the fact.
Innovation is a different animal altogether, almost like it eats itself. Just think – the most successful innovations come at the expense of what’s been successful.
If I could choose my competitive advantage, it would be an unreasonably strong engineering team.
Ideas have no value unless they’re morphed into winning products, and that’s what engineering does. Technology has no value unless it’s twisted into killer products. Guess who does that?
We have fully built out methodologies for marketing, finance, and general management, each with all the necessary logic and matching toolsets, and manufacturing has lean. But there is no such thing for engineering. Stress analysis or thermal modeling? Built a prototype or do more thinking? Plastic or aluminum? Use an existing technology or invent a new one? What new technology should be invented? Launch the new product as it stands or improve product robustness? How is product robustness improved? Will the new product meet the specification? How will you know? Will it hit the cost target? Will it be manufacturable? Good luck scripting all that.
A comprehensive, step-by-step program for engineering is not possible.
Lean says process drives process, but that’s not right. The product dictates to the factory, and engineers dictate the product. The factory looks as it does because the product demands it, and the product looks as it does because engineers said so.
I’d rather have a product that is difficult to make but works great rather than one that jumps together but works poorly.
And what of innovation? The rhetoric says everyone innovates, but that’s just a nice story that helps everyone feel good. Some innovations are more equal than others. The most important innovations create the killer products, and the most important innovators are the ones that create them – the engineers.
Engineering as a cost center is a race to the bottom; engineering as a market creator will set you free.
The only question: How are you going to create a magical engineering team that changes the game?
We want it now; no time for new; it’s instant gratification for us, but only if it doesn’t take too long.
A short time horizon drives minimization. Minimize waste; reduce labor hours; eliminate features and functions; drop the labor rate; cut headcount; skim off the top. Short term minimizes what is.
Short term works in the short term, but in the long term it’s asymptotic. Short term hits the wall when the effort to minimize overwhelms the benefit. And at this cusp, all that’s left is an emaciated shadow of what was. Then what? The natural extrapolation of minimization is scary – plain and simple, it’s a race to the bottom.
Where short term creates minimization, long term creates maximization. But, today, long term has mostly negative connotations – expensive, lots of resources, high risk, and low probability of success. At the personal level long term, is defined as a timeframe longer than we’re measured or longer than we’ll be in the role.
But, thankfully, there comes a time in our lives when it’s important for personal reasons to inject long term antibodies into the short term disease. But what to inject?
Before what, you must figure out why you want to swim against the current of minimization. If it’s money, don’t bother. Your why must have staying power, and money’s is too short. Some examples of whys that can endure: you want a personal challenge; you want to help society; your ego; you want to teach; or you want to help the universe hold off entropy for a while. But the best why is the work itself – where the work is inherently important to you.
With your why freshly tattooed on your shoulder, choose your what. It will be difficult to choose, but that’s the way it is with yet-to-be whats. (Here’s a rule: with whats that don’t yet exist, you don’t know they’re the right one until after you build them.) So just choose, and build.
Here are some words to describe worthwhile yet-to-be whats: barely believable, almost heretical, borderline silly, and on-the-edge, but not over it. These are the ones worth building.
Building (prototyping) can be expensive, but that’s not the type of building I’m talking about. Building is expensive when we try to get the most out of a prototype. Instead, to quickly and efficiently investigate, the mantra is: minimize the cost of the build. (The irony is not lost on me.) You’ll get less from the prototype, but not much. And most importantly, resource consumption will be ultra small – think under the radar. Take small, inexpensive bites; cover lots of ground; and build yourself toward the right what.
Working prototypes, even crude ones, are priceless because they make it real. And it’s the series of low cost, zig-zagging, leap-frogging prototypes that make up the valuable war chest needed to finance the long campaign against minimization.
Short term versus long term is a balancing act. Your prototype must pull well forward into the long term so, when the ether of minimization pulls back, it all slides back to the middle term, where it belongs.
When engineers see a big opportunity, we want desperately to move the company in the direction of our thinking, but find it difficult to change the behavior of others. Our method of choice is usually a full frontal assault, explaining to anyone that will listen the opportunity as we understand it. Our approach is straightforward and ineffective. Our descriptions are long, convoluted, complicated, we use confusing technical language all our own, and omit much needed context that we expect others should know. The result – no one understands what we’re talking about and we don’t get the behavior we’re looking for (immediate company realignment with what we know to be true). Then, we get frustrated and shut down – opportunity lost.
To change the behavior of others, we must first change our own. As engineers we see problems which, when solved, result in opportunity. And if we’re to be successful, we must go back to the problem domain and set things straight. Here’s a sequence of new behaviors we as engineers can take to improve our chances of changing the behavior of others:
Step 1. Create a block diagram of the physical system using simple nouns (blocks) and verbs (arrows). Blue arrows are good (useful actions) and red arrows are bad (harmful actions). Here’s a link to a PowerPoint file with a live template to create your own.
Step 2. Reduce the system block diagram down to its essence to create a distilled block diagram of the problem, showing only the system elements (blocks) with the problem (red arrow).For a live template, see the second page of the linked file. [Note - if there are two red arrows in the system block diagram, there are two problems which must be solved separately. Break them into two and solve the first one first. For an example, see page three of the linked file.]
Step 3. Create a hand sketch, or cartoon, showing the two system elements (blocks) of the distilled block diagram from step 2. Zoom in so only the two elements are visible, and denote where they touch (where the problem is), in red. For an example, see page four of the linked file.
Step 4. Now that you understand the real problem, use Google to learn how others have solved it.
Step 5. Choose one of Google’s most promising solutions and prototype it. (Don’t ask anyone, just build it.)
Step 6. Show the results to your engineering friends. If the problem is solved, it’s now clear how the opportunity can be realized. (There’s a big difference between a crazy engineer with a radically new market opportunity and a crazy engineer with test results demonstrating a new technology that will create a whole new market.)
Step 7. If the problem is not solved, or you solved the wrong problem, go back to step 1 and refine the problem
With step 1 you’ll find you really don’t understand the physical system, you don’t know which elements of the system have the problem, and you can’t figure out what the problem is. (I’ve created complicated system block diagrams only to realize there was no problem.)
With step 2, you’ll continue to struggle to zoom in on the problem. And, likely, as you try to define the problem, you’ll go back to step 1 and refine the system block diagram. Then, you’ll struggle to distill the problem down to two blocks (system elements). You’ll want to retain the complexity (many blocks) because you still don’t understand the real problem.
If you’ve done step 2 correctly, step 3 is easy, though you’ll still want to complicate the cartoon (too many system elements) and you won’t zoom in close enough.
Step 4 is powerful. Google can quickly and inexpensively help you see how the world has already solved your problem.
Step 5 is more powerful still.
Step 6 shows Marketing what the future product will do so they can figure out how to create the new market.
Step 7 is how problems are really solved and opportunities actually realized.
When you solve the real problem, you create real opportunities.
Masquerading as maximizers, these minimizers-in-sheep’s-clothing have done deep harm. Though Six Sigma is almost dead (it’s been irrelevant for some time now), it has made a lasting mark. Billed as a profit maximizer, it categorically rejects maximization. In truth, it’s a variation minimizer and difference reducer. If it deviates, Six Sigma cuts its head off. Certainly this has a place in process control, but not in thinking control. But that’s exactly what’s happened. Six Sigma minimization has slithered off the manufacturing floor and created a culture of convergence. If your thinking is different, Six Sigma will clip it for you.
Lean is worse. All the buzz around lean is about maximizing throughput, but it doesn’t do that. It minimizes waste. But far worse is lean’s standard work. Minimize the difference among peoples’ work; make them do it the same; make the factory the same, regardless of the continent. All good on the factory floor, but lean’s minimization mania has spread like the plague and created a culture of convergence in its wake. And that’s the problem – lean’s minimization-standardization mantra has created a culture of convergence. If your thinking doesn’t fit in, lean will stomp it into place.
We need maximization at the expense of minimization, and divergence before convergence. We need creativity and innovation. But with Six Sigmaphiles and lean zealots running the show, maximization is little understood and divergence is a swear.
First we must educate on maximization. Maximization creates something that had not existed, while minimization reduces what is. Where Six Sigma minimization converges on the known right answer, creativity and innovation diverge to define a new question. The acid test: if you’re improving something you’re minimizing; if you’re inventing something you’re maximizing.
Like with He Who Shall Not Be Named, it’s not safe to say “diverge” out loud, because if you do, the lean Dementors will be called to suck out your soul. But, don’t despair – the talisman of guided divergence can save you.
With guided divergence, a team is given a creatively constructed set of constraints and very little time (hours) to come up with divergent ideas. The constraints guide the creativity (on target), and the tight timeline limits the risk – a small resource commitment. (Though counterintuitive, the tight timeline also creates remarkable innovation productivity.) Done in sets, several guided divergence sessions can cover a lot of ground in little time.
And the focused/constrained nature of guided divergence appeals to our minimization bias, and makes it okay to try a little divergence. We feel safe because we’re deviating only a little and only for a short time.
Lean and Six Sigma have served us well, and they still have their place. (Except for Six Sigma.) But they must be barred from creativity sessions and front end innovation, because here, divergence carries the day.
Mainstream follows mainstream – they don’t know why, they just do. In truth, mainstream likes to be lead by the nose because it’s easy, because they can get through the week without caring. But not caring is not right.
But when you care, when you really care, when you care so much it hurts, you’ve got it right. Loneliness hurts because you see habitual mistakes on the horizon; it hurts because you see bureaucracy trump thinking; it hurts because you see hierarchy squelch creativity. Put simply – it hurts because you care enough to look and you’re smart enough to see.
But take comfort in your loneliness. Though sometimes it feels they want to poke your eyes out, deep down companies want you to see. Sure, they’re afraid of the ruckus, but they want you to wrestle with the familiar. Yes, they won’t sanction your detectiveness, but they want you to investigate the crime scene. Absolutely – though with plausible deniability – they want your inner Nostradamus to conger the future.
Every-day-all-day loneliness is too much, but a low dose is good. 100% loneliness festers into anger, but now-and-again loneliness is healthy.
Take stock in your loneliness – it’s a sign your brain is turned on. And it’s a sign you care.
Know what’s new in the new design. To do that, ask for a reuse analysis and divvy up newness into three buckets – new to your company, new to your industry, new to world. If the buckets are too big, jettison some newness, and if there’s something in the new-to-world bucket, be careful.
Create test protocols (how you’ll test) and minimum acceptance criteria (specification limits) before doing design work. It’s a great way to create clarity.
Build first – build the crudest possible prototype to expose the unfamiliar, and use the learning to shape the next prototypes and to focus analyses. Do this until you run out of time.
Cost and function are joined at the hip, so measure engineering on both.
Have a healthy dissatisfaction for success. Recognize success, yes, but also recognize it’s fleeting. Someone will obsolete your success, and it should be you.
To get an engineering team to believe in themselves, you must believe in them. To believe in them, you must believe in yourself.
There are a lot of complex self-help programs out there to help us reach our potential. These programs are designed to take us to the next level by building on what we have. And they do help. But implicitly the programs assume we’ve got a foundation in place and it’s solid enough to stand on. But I think our foundation is rickety.
To be what we must be is our responsibility, and if we’re going to get any value from our self-help programs it’s our responsibility to establish the conditions for growth.
Here are three forgotten pillars that must be in place if we’re to grow:
Good sleep habits – most of us our sleep deprived, yet the data is clear that a rested brain thinks better. If we’re to grow, we’ve got to think better, and to think better we need more rest.
Good eating habits – most of us eat poorly. We eat too much which hurts our bodies and we eat too infrequently which hurts our brains. We all know our muscles need calories or they don’t work well, but what we forget is our brain also needs calories. (It’s 2% of our weight but consumes about 20% of our total calories.) And when we eat infrequently our blood sugar drops and so does our brains’ ability to think. If we’re to grow, we need to eat better.
Good fitness habits – most of us don’t exercise daily. Our brain is connected to our body, and they interact. When the muscles are exercised they cause the brain to make chemicals that sooth and calm. When we exercise we have a better attitude, have more stamina, and think better. If we’re to grow, we need more exercise.
The big three are far more powerful than the best self-help scheme, but they’re also far scarier. They’re so scary because they’re easy to measure and completely within our control.
After a one month trial you’ll know if you’ve slept more, worked out more, and ate better. And if you haven’t, you’ll also know who is responsible.
With the police state approach, detailed processes are created and enforced; rules are created and monitored; and training is dealt out and attendance taken. Yes, behavior is changed, but it’s fleeting. Take your eye off the process, old behavior slips through the fence; look the other way from the rules, old behavior clips the barbed wire and climbs over the wall. To squelch old behavior with the police state approach, gulag energy must be consistently applied.
To squelch is one thing, but to create lasting behavior change is another altogether. But as different as they are, there’s a blurry line of justice that flips innocent to guilty. And to walk the line you’ve got to know where it is:
- Apply force, yes, but only enough to prevent backsliding – like a human ratchet. Push much harder and heels dig in.
- The only thing slower than going slow is going too fast. (Remember, you’re asking people to change the why of their behavior.) Go slow to go fast.
- Set direction and stay the course, unless there’s good reason to change. And when the team comes to you with a reason, deem it a good one, and the cornerstone of trust is laid. (This is a game of trust, not control.)
But there are some mantras to maximize:
- Over emphasize the positive and overlook the negative.
- Praise in public.
- Don’t talk, do.
The first two stand on their own, but the third deserves reinforcement.
This isn’t about your words, it’s about your behavior. And that’s good because you have full authority over your behavior. Demonstrate the new behavior so everyone knows what it looks like. Lead the way with your actions. Show them how it’s done. For lasting change, change your behavior.
Even if changing your behavior influences only one person, you’re on your way. The best prison riots start with a single punch.