Just after raising a round, it looks like your startup should have it all together: there’s plenty of money, a product that more or less meets customer needs, and a coherent strategy that convinced the investors—and maybe you too—that you have a billion-dollar future.
But over and over again, at over 30 startups in the last three years, I’ve seen chaos descend at just this moment: salespeople race off to close insane, undeliverable deals, programmers disappear for weeks or months and return with half-built, totally wrong features, or marketing and product teams get into a fight to the death over what the company vision is. It seems there's something uniquely difficult about taking your organisation from small and experimental to medium-sized and disciplined.
And what's particularly surprising is that although I'm often brought in to "sort out tech" or "train the CTO", I wind up having a much bigger effect by addressing a different class of problems that are leading to this chaos: failings in the four key areas of leadership, goal-setting, feedback and relationships that block progress across the board. For instance, at BookingBug I helped sales, tech, and operations improve in those four areas, leading directly to a significant change in product strategy that put over £1m on the bottom line.
The Squirrel Test is my attempt to help you discover what you can do to improve your scale-up company's performance, with a series of 12 yes-or-no questions that should take under five minutes to answer. It’s totally non-scientific, being based entirely on my direct experience consulting and turning around multiple Series A startups. And it doesn’t say anything about tech choices, business model, market size, or product strategy—if you get a high score, it means your team is running smoothly, and once you get that sorted, you should be able to figure the rest out with their help.
Here’s the test. Give yourself one point for every “Yes” answer. (Don’t worry if you don’t know what everything means—there’s more explanation for each question below.)
How did you do? A score of 11-12 is awesome and 9-10 is pretty good; below 9 means you’re at risk of implosion. But don’t panic if you’re not where you want to be—read on, there’s lots you can do to improve your score.
Experience matters a lot more after Series A—you’re growing fast and don’t have time to make mistakes, and the early pioneers who do everything have to make way for the incoming settlers and their greater specialisation and outside knowledge. A crew of total novices is almost certain to crash the ship.
But the good news is that it’s the average level of experience among senior leaders that matters most, not where that experience sits. A 15-year veteran CFO can bring in financial and organisational discipline where there was chaos in marketing and sales; a Head of Product with 10 years of hard knocks under her belt can help the whole business line up behind a common vision. If you have—or can find—even a single veteran to sit at the top table, that can make all the difference.
The company started almost by accident, as a side project that blew up with no warning. Two years later, that first product outsold all the competitors combined, but attempts to follow it up were tanking and no one could figure out why. Projects started willy-nilly with no clear goal; developers used wildly inappropriate and overcomplicated tech for simple tasks, and marketers were advertising in countries where the company didn't even trade. The leadership took two months to settle on an annual strategy, then the founders changed it a month later.
When I polled the senior leadership team on my first day, every single one was completely new to the job. Sure, some had led teams or even companies, but none had done his or her current role in another startup.
Well, I should say that everyone was new except for the grey-haired guy working quietly with a small team on fulfillment, who’d been doing it for decades at multiple companies. Unsurprisingly, his was the only part of the business that ran like clockwork.
In the early days, everybody could fit into a phone booth and you didn't need any structure to get things done. But suddenly there are teams trying to collaborate, not just individuals, and growing from n people to n+1 means that new person needs to talk to all n of the others. Because this growth rate is exponential, you develop a need for more structured interaction and co-operation very quickly.
Just getting everyone together is a great start, but to do well you're going to need to see meaningful output from each session. Anti-patterns here include everyone giving status with no discussion, or endless discussion that goes nowhere. Try focussing on the goals that you've agreed and talk through what everyone is doing to achieve those—you'll be surprised how much alignment and co-operation you can get quickly.
A client once had—I kid you not—a one-week senior leadership meeting to set strategy. It was one in a long series of directionless sessions that sapped energy and didn't achieve alignment. Only when the founders brought discipline and focus to their leadership meetings (including prominent use of an alarm clock!) did they start to get value from them.
Management by exception is one of my favourite chaos-tamers. The idea is simple in theory:
The problem for most people is step 3. They try to fix everything and wind up with loads of activity and no output. (When computers try to do this we call it thrashing, an apt description.) The whole point is to help yourself (and your team!) focus tightly on high-leverage actions and ignore less-valuable problems, painful though it may be.
The other very important principle is that you must share what you are doing with as many collaborators as you can (as a founder, that means the whole company). If you don't, others may expect you to fix a lower-priority issue and be disappointed, or worse, they may conclude you don't care, which kills morale.
A great book on management by exception (even though I think it doesn't actually contain that phrase) is High-Output Management by Andy Grove, former chairman of Intel.
As a very junior programmer in my first job, I overheard a senior leader in the kitchen saying casually, "I'm focussed on acquiring this competitor for the next two months, so I've put George in charge of my team. He's keeping the lights on, but not really doing very well; I think we'll miss some if not all of our targets this quarter."
I was astonished that anyone would admit that her team wasn't performing. Wasn't she responsible? Didn't she care about missing those targets? It was only years later that I understood she was using management by exception, and clearly had decided that the acquisition was far higher in value than whatever her team was trying to deliver.
Of course everyone will tell you that you need a vision, and most companies can drag one out of a closet or dredge up a neglected About page that describes where they are going. Startups are usually better off; they often have at least one messianic founder who can paint an inspiring picture of the world-begirdling future of the company, no matter how humble its current circumstances may be.
But the danger is not that you won't have a vision, but that it will either be too diffuse ("revolutionise pet care" rather than "the world's best bird food delivered to your door") or widely ignored (“Sure, we’re going to build robot florists, but right now I’m too busy sorting petunias to think about it”). In both cases, everyone dashes off in a different direction according to his or her own understanding, and product, marketing, and sales decisions rapidly diverge, with dire consequences.
One of my clients managed to overdeliver on the vision thing - they had two clearly defined, well-thought-out visions, which naturally were completely incompatible with each other and championed by different senior leaders. Unsurprisingly, the poor Head of Product was left high and dry by this, with no chance to make meaningful plans and roadmaps. The company was paralysed and unable to make meaningful progress for months. Only when the founders managed to pick a single focus did matters improve.
Explanation to be added here...
Still collecting stories to fit here!
I'm often asked by founders to explain how to give direction without micromanaging. Even more often, I observe leaders bouncing back and forth between hands-off hope and hands-dirty detail management, never realising that they are missing a happy medium in between.
Luckily there is an easy way out, if you're willing to learn from the 19th-century Prussian army and their "directed opportunism". Briefly, this means briefing someone in your team by telling her the situation, your intent, your direction to them, what she is free to choose, and what she must not do (her constraints). Then (this is crucial!) book a time for a back briefing where she describes her plan and how she will be accountable for delivering it.
For more, read The Art of Action by Bungay, who manages somehow to be both a management consultant and a military historian!
A Prussian general had just defeated the French and wanted to surround their fleeing army. Inconveniently, it was three in the morning and, lacking conveniences like radar and radio, he had only the vaguest idea where they were or even where his own two divisions had wound up. Undaunted, he wrote up a briefing telling his commanders the situation (the French are running), his intent (to make them surrender), their direction (move north to trap them) and constraints (don't be drawn into skirmishes). He handed the copies of the briefing to two messengers and told each to ride in the direction of the armies, returning with an answer. The answers arrived by morning and these back briefings allowed the general to move his own army to close the circle and force Napoleon III to surrender.
Explanation to be added here...
Still collecting stories to fit here!
Explanation to be added here...
Still collecting stories to fit here!
An amazing thing that happens when you start to scale is that you actually, finally, have a meaningful volume of real users to talk to. I'm constantly astonished that so many startups fail to exploit this to the hilt, relying on founder intuition and a (probably false) "understanding of the market" instead of actual feedback (whether quantitative or qualitative).
Frank Boyd crystallised the value of rapid feedback in the OODA loop - a cycle of orienting, observing, deciding, and acting. He found that American fighter planes with bubble canopies allowed pilots to see much more around them, so they could go through more OODA loops than Russian pilots in faster planes with more limited vision. As a result the Americans won more dogfights in the Korean war than anyone would have predicted based on the planes' relative maneouverability.
To measure your OODA loop speed, consider how long it takes you to get from an idea ("users would like coffee delivered by birds") to a test ("try our new pigeon delivery option") to learning ("pigeons spill the drinks, let's try owls instead"). Hint: this should be measured in days, or you're way behind the curve.
Many of my clients release new software every day, but one takes it to the extreme. They build in tiny increments, so tiny that sometimes what they release doesn't even work. For instance, they might start with just a button that says "Colour picker", but nothing happens when you click it. They count the number of clicks and only if there are enough to be interesting do they actually build the colour picker, proceeding one tiny piece at a time: first a static image of a colour picker that can't be clicked, then a picker that doesn't store its results, and so on. And as they release these changes day after day, they measure usage, talk to users, and adjust what they are building to match the feedback.
Can you get round the OODA loop this fast? If not, why not?
Explanation to be added here...
Still collecting stories to fit here!
Explanation to be added here...
Still collecting stories to fit here!
Explanation to be added here...
Still collecting stories to fit here!