Stop doubling down
What we can learn from risk habituation
One of the most consistent phrases I’ve heard in business since I was 21 is the idea of doubling down. It’s so frequent, it’s barely even registers as jargon anymore.
It exists in all strands of business, but it feels even more in vogue in startups. Why?
When someone asks me what ‘growth’ is, on my more concise days I’ll answer the following five things:
Know your customer. Understand the customer problem better than they do. #customerpyschology
Precise metrics and goals. Understand how your business will grow, define a narrow customer-centric north star metric and understand the levers that confidently will change those. #growthengine
Diligent attention. Unite the team behind that very narrow focus area, and say no to everything else. #focus
Rapid experimentation. Your rate of experimentation is directly correlated to chances of success. #experiment
Doubling down. Taking winning ideas and doubling down on them. #scale
That makes logical sense, right?
Why behavioural science reveals this might be a bad idea
I’ve recently read Tali Sharot and Cass R. Sunstein’s new book Look Again. Tali Sharot is the author of the excellent The Optimism Bias. And Cass R. Sunstein, along with Richard Thaler, wrote the behavioral science must-read Nudge. Sidebar, Look Again is already a contender for my book of the year.
The core thesis of the book is around habituation. This means that humans naturally get used to new situations very quickly. It explains why if you get a payrise, you’ll feel the benefit of it for a short period of time, but very quickly it will feel normal. It explains why humans naturally always want more.
It also provides another lens to the infamous Milgram Experiment. This classic of A-level psychology includes an authoritative figure instructing a participant to administer increasing electric shocks to another person. The victim eventually stops responding as the dial goes beyond ‘fatal’. Most people followed the orders. This was chalked up to how willing people are to follow authority. In fact habituation is important to consider. If someone was asked to go from 0 to fatal, we’d imagine far fewer people do it.
So why am I talking about habituation in this post?
Risk habituation means we’re losing future startup bets.
Risk habituation is why extreme sports people take bigger and bigger risks, which can often end up fatally. And it’s why those in positions of power in business or politics, get used to their level of risk taking working out, and then make worse decisions. David Cameron calling for Brexit, so posits Look Again, was a response to his risk habituation.
As I was reading, I kept thinking back to all the occasions in various businesses I’ve seen people ‘double down.’ And in retrospect, it caused more damage than good.
There was one story in particular I remembered that I’ve re-evaluated with the context of risk habituation.
How we habituated risk, doubled down, and made mistakes with PR stunts
Back in 2017, when I was head of growth at Thriva, we tested a PR stunt. Like all good startup experiments, we tried to ascertain what the minimum viable test would be. We set a budget of £10k and briefed a few creative PR agencies.
The winning pitch? We hired a guy called Ant Smith who had previously written a book called The Small Penis Bible named after his personal experiences. We got him to carry a placard around saying There’s Nothing Wrong With A Little Prick, because we were a finger-prick blood test. We turned that photo into a billboard, stuck it up in Wimbledon somewhere and then that became the PR hook.
That story went viral. The coverage across the MailOnline, Joe, The Sun, the Mirror, LadBible was the best you could ask for. The shares on Facebook and Twitter were in the tens of thousands, it was a real big moment for a week or so. And it won awards for the agency we hired, The Romans.
It was impactful too. The halo effect of that coverage meant that we could push far harder on our acquisition engine for two weeks than we’d ever pushed before. New customer contribution was its highest over any period for a long time.
The campaign was a success.
I should point out that before we did it, this was high risk. We were growing steadily through Facebook Ads, and starting to figure out some of our SEO strategy, but this was unpredictable. It was harder to attribute, the upfront costs were massive (comparatively), and it felt like it might be buzzy but non-incremental.
That it was such a success meant we were elated as a small team. And a month or so later as we started planning for our January 2018 push, we knew a creative PR stunt was a vital part of it.
What did we do? We doubled down. In fact we more than doubled down. We increased the budget for the next stunt by 2.5x.
We started listening to wilder ideas. After all the creativity of the first one was so out there – that must have really contributed to its success, we thought.
And we stopped judging and evaluating each idea as if it was the first one.
In short, we had habituated the risk of this marketing channel.
The lead up to the second campaign felt great. This time, we were launching a giant fatberg next to the River Thames. The whole team were going to come down and chat to people and flyer. We were bringing in a new experiential element to the challenge. We were hyped.
Launch day came around and suddenly… nothing.
On the ground, people didn’t get it. Lots of people looked at it puzzled, but when we engaged them, they just didn’t get it.
Press attention started to rack up. The Independent got wind, and we were happy with a broadsheet with the tabloids dominating so much of the Little Prick coverage. But there was no chatter. No shares.
And most importantly, there was no halo effect. Nothing changed in our acquisition.
This campaign was a fail by every measure.
Looking back there was lots we did in analysis afterwards as to why it failed.
There were lots of good hypotheses. One was that the link between the ‘why we’re doing this stunt’ and us as a product was too many steps. Another was that it was the virality that made the former work, and this one didn’t get any of it. Another that the monetary constraints of the first campaign allowed for more creativity.
But what we didn’t do at the time, was analyse our approach to how we did the work.
We used to make predictions as a team about what would happen with big experiments. With Little Prick almost everyone underestimated its impact. While with Fatberg almost everyone overestimated the impact.
We habituated the risk. As a result, we increased budget rapidly, we became less critical, and we got overly confident.
After that, we swore off PR stunts. I think in retrospect, that was a mistake.
It’s not always the big things
This stunt was an extreme example, but I’ve seen this thinking everywhere.
Very often in paid acquisition, we hear a call of “let’s just try doubling the budget,” which 90% of the time won’t deliver anything affordable in incremental acquisition.
In SEO I’ve seen a selection of blogposts or information pages drive affordable acquisition, and then the response was to ‘double down’ on SEO efforts. Lots of human time goes into copywriting, strategy, design, and then the new pages never land.
And in broader strategic decisions, I’ve seen small product areas hit it off with customers, and then as a result entire business OKR periods get shifted towards those. And the impact never as strong as they hope.
Winners in experimentation should not stop you to critically think
Why do we do this? We experiment in growth for many reasons: it creates momentum, it reinforces focuses, and it means we don’t create things that are bigger than they should be before validation. Those are the logical reasons why.
But the emotional reason for experimentation is it gives us confidence and a sense of control. It tells us that we’ve validated our ideas and therefore we can go do those things. They’ve been tested scientifically after all.
Experimentation is good. And it’s good for many reasons. But so too is understanding our human psychology in decision-making. When we habituate risk, we become more confident that we will succeed. And actually for many things in startups, that can cause more harm than good.
What have you been doubling down that might still be high risk?