What is Growth?
All the questions you had about growth (maybe) but were too afraid to ask.
We're hiring our first grad role at the moment, and as part of that process, we've been trying to establish how to pitch the role. One question we came back to again and again was what is the understanding of 'growth'?
Growth has a lot of assumptions and connotations stuck to it.
Much like natural wine, growth doesn’t have a broadly accepted definition. In fact, if you ask 10 people in startups or marketing what they thought it was, I imagine you’d get 10 different answers.
I – and by that token us at Amphora – have a very specific view of what this is.
It’s built upon some historic, academic knowledge from those who created it. But it also brings in direct experience of actually having worked in UK growth teams since 2014.
This post includes:
Intro – and the tl;dr
The Gang of Four of Growth
Why there's so confusion around the term
How you can build growth into your business
"Growth – that's just a startup name for marketing, right?"
Let's kick off with the tl;dr. In my view, Growth comes down to these three core factors:
A scientific method applied to how you should work: rapid experimentation working towards a north star metric
A focus on user psychology and behaviour: deep understanding of why people do things, and then using that in your product
A repeatable and scalable engine that allows a business to grow (profitably).
Everything else that falls under the growth banner falls into one of those three things. We'll explore these concepts in depth and their origins in this post.
But first, a couple of the most frequently expected quesitons
Isn't it just a rebrand of marketing by startups?
In my experience no.
When I started dipping my toe into marketing between 2010-2012, and then probably in the years that succeeded that, marketing was a mixture of things.
Marketing was: campaign-focused, driven by budgets set a year in advance, focused on fluffy vanity metrics like reach. It was often the viewpoint of individuals and not of the customers. It was often very offline, and very manual. Workloads happened with long term planning, and little deviation from that. There was little that checked if things were working, if you even knew what ‘working’ meant.' And someone in a marketing team was more likely to have a English degree than a science one. In many businesses, marketing was seen as a cost.
As you’ll see throughout this post, these are almost opposites to what growth is.
You can argue that marketing could have been lots of these things. After all, it was marketers who invented the randomised controlled trial (RCT) 25 years before scientists did, and we called it the AB test. That was in the 1920s.
But while marketing had the opportunity to be these things. Growth brought together multiple aspects from the broad history of marketing, combined it with very modern traits that marketers weren’t doing, and then packaged it up for the 21st century.
Growth is just another name for performance marketing?
The other argument I hear is that this is just about ads, or spending money to drive action online. What many would call performance marketing. Performance marketing can be a part of growth, but you can do performance marketing while having no understanding of broader growth principles.
And likewise, I’ve worked with incredible growth people who don’t know the first thing about running paid ads.
The Gang of Four of Growth
Chamath Palihapitiya
Back in 2008, there was a 31 year old Facebook executive called Chamath Palihapitiya.
Chamath had a pretty exciting life before then. He spent his teenage years running blackjack games at his high school, took up gambling, and then competed in the World Series of Poker. He spent a very short time in banking before sacking it off to join WinAmp (sidenote: I used to love WinAmp). WinAmp got acquired by AOL, and then Chamath became the youngest VP in AOL history to lead AOL Instant Messenger (AIM). After a short stint as a VC, he took at job at a three-year-old Facebook.
Chamath didn’t get off to a good start at Facebook and by the end of 2007 was close to leaving/being fired. He and Cheryl Sandberg met and discussed options. He wanted to work on something new and believed he could have an impact.
Around that time Facebook’s growth had plateaued at around 90 million users. They had always been a small-g growth organisation. But a large number of their growth unlocks had come from product innovations, like introducing the Newsfeed.
Chamath wanted to lead the efforts to fix that.
There’s a brilliant chapter in Stephen Levy’s Facebook: The Inside Story, and I’m sharing here my notes from that chapter that I have found to be most useful.
Introduced the north star metric: monthly active users
The sole focus was to grow MAUs. Every action was focused on this.
Sheryl Sandberg said to Chamath: “Maybe you should just call [the team] growth”
Built on the data culture that had already begun at Facebook
Assembled a diverse team of engineers, marketers, and data scientists
Work was considered ‘unglamarous’ – most of the eng and PMs wanted to work on cool product stuff
Distinct culture within Facebook: Chamath purposefully tried to break the FB homogenous worldview of his broader teammates
Called it a ‘Growth Circle’ not growth team, so as to further create this separate culture.
There were lots of small change that had 1%, 5%, or 10% improvements – like SEOing the site to make people searchable on Google.
But also big ones: People You May Know (PYMK internally), borrowed from LinkedIn, but taken to a huge level.
PYMK had lots of dark sides to it. A sex worker once found recommendations of her clients, who did not know her real name. “A sperm donor got a recommendation for a biological child he never met.”1 And a psychiastrists’ clients got recommendations to friend each other.
They created ‘dark profiles’ – of people who hadn’t signed up, but Facebook had info on them – and targeted Google Ads to them to attract them to the site
Chamath identified launching internationally as the next growth frontier; they did this by using the community to help translate the service around the world: not just core markets
Recognised that internationalisation is usually an ops job, where ops would focus on making it cheaper.
When owned by growth, it’s refocused as an opportunity
Prioritised speed of execution and kept cost minimal
Recognised they’d run out of internet users, so created the proposal that went on to become internet.org – an organisation designed to give internet access to everyone in the world
Chamath left Growth in 2011 and was replaced by Javi Olivan, who the now-Meta-CMO called “very kind, considerate, just a great guy”
Growth started with a focus on registrations, invitations and new user experience, but Mark was the one responsible for expanding remit more broadly.
Growth would go on to own Internationalisation, Mobile, and then later, Trust.
There was a single-minded focus on growth in this team, and little worry for second order effects.
Sean Ellis
Also in 2008, a young Sean Ellis got a call from Drew Houston at Dropbox. Dropbox had been growing well but they were in a hyper competitive market.
Sean had previously had success at a few other startups like LogMeIn. There, he worked with the engineering team to find new opportunities to acquire users and get more value out of the existing ones.
This story is covered well in Ellis’ book Hacking Growth2, here are some of my notes from that:
Ellis in an early job at Uproar couldn’t compete with big advertisers. The company was running out of money, and so he and some engineers made a web widget that created low cost affiliates out of 40,000 websites
They AB tested copy, headlines and imagery in that widget relentlessly
Ellis went to LogMeIn, where they’d been running search ads which soon became unprofitable
At LogMeIn, they used customer insights to drive landing page changes, again, experimenting heavily in the process
The next insights showed that most drop-offs happened at another stage in the onboarding process. They ran experiments across install, sign-up and further down the funnel
Ellis recognised that experimentation was core to the process he coined as ‘growth hacking’
Ellis invented the “must have score”, what’s now more commonly seen as the product-market fit question
Ellis went to Dropbox and asked to solve the marketing problem without paid ads. He did a deep dive on data and realised that referrals was strong. They looked to PayPal who had spent $70m on their ‘give $10, get $10’ acquisition strategy. Dropbox didn’t have that cash and so gave aware storage instead. Invites went up 60%
The team started running “high-tempo tests” every week to see how to improve that referral rate. By 2010, 2.8m invite were being sent out per month.
Sean went on to found the website growthhackers.com
Andrew Chen
In April 2012, Andrew Chen published a now infamous essay called Growth Hacker is the new VP Marketing3. (Sidebar: this was my entry into the subject).
In it, Andrew reflects on Silicon Valley leading up to that point. Borrowing Ellis’ term, he expands on this. Here are my notes:
Coding / technical chops is required to be a great marketer now
At the core is the question: how do I get more users for my product?
Answers include AB testing, landing pages, direct marketing, data, virality
“Projects like email deliverability, page-load times, and Facebook sign-in are no longer technical or design decisions – instead they are offensive weapons to win in the market”
Technology has meant that going from 0 to 100m is a believable and fast path
The platforms that have 100m users (all social networks) can all be taken advantage of much more cost effectively than the old ones (TV, PR, etc)
Popularised the story of the Airbnb / Craigslist growth hack where Airbnb wrote code to scrape and post Craigslist listings to be able to brute force growth
Chen went on to lead growth at Uber: where their data understanding built the knowledge that if a wait time was more than seven minutes, the entire network fell apart. And so every action must be taken to make wait times under seven minutes.
Paul Graham
Finally, five months after Chen’s essay, Y Combinator founder Paul Graham wrote the article Startup = Growth4. Anyone who has ever worked with me will have heard me reference this dozens of times.
In short, Graham states that startups are businesses that are inherently and by their nature designed to grow big and fast. If you open a barbershop, it is not a startup, even though it is a small business. Startups have to have the ability to grow fast and big.
What do we mean by fast and big? Well in VC terms, that means a business that can reach $100m in annual revenue.
Why is there so much confusion around growth today?
Looking at the aspects of what those four talked about, they all fit into my summarised view of:
The scientific method applied to how you should work: rapid experimentation working towards a north star metric
A focus on user psychology and behaviour: deep understanding of why people do things, and then using that in your product
A repeatable and scalable engine that allows a business to grow.
The scientific method is something that exists everywhere. The data-focused, rapid experimentation was true across all reference points.
The focus on user behaviour exists across all of them too. At Facebook, their biggest lever was playing on psychology principles of being connected. At LogMeIn, the analysis started in understanding user behaviour.
As for the scaleable growth engine, if you take the Facebook example; you recognise that as a network effects business, users create more users. The Newsfeed creates more interactions. PYMK, and then internationalisation, and then internet.org all created more users.
The lack of tech businesses in the UK has skewed the view of Growth
Over the last decade, UK startup culture has borrowed from the US. However, what many people see as Growth today will likely look different from the Growth Circle at Facebook, or early growth hacker roles at Dropbox or Airbnb. Why?
We have far fewer tech businesses than there are in the US. We have a lot of tech-enabled businesses. Or a lot of deep tech businesses which are like research or deep b2b based. But we have far fewer true tech businesses either in social or b2b than in the US.
For reference, I’m used ‘tech businesses’ here to mean those where you are actually using a piece of technology. Microsoft Word is a tech product. A DTC crisps subscription is not.
This is important because when you look at those referenced businesses in respect of their growth engine, it changes how they grow. How does Facebook grow? It’s a social network, it must grow by bringing your friends onto the platform to make it more useful. Airbnb’s growth engine ultimately went on to become advertising, but that early growth hack got it a long way for free.
Conversely, if you’re a DTC subscription, the ideal user behaviour is likely one where the consumer never touches your tech. The value is physical, but tech just enables you getting it better.
Second, technology has now shifted a lot. In 2012, if you wanted to gather data on a funnel, and run experiments on it, you had to have engineers and data analysts. Now, there’s good enough software out there that marketers can do that without an engineer. Those marketers will still be tech-savvy, but they don’t need a CS degree to run an experiment.
This stuff is important when you consider the growth engine.
All DTCs grows big and fast and affordably predominantly from ads. You might be able to game ads for a little bit. You might be able to drive enough SEO to mitigate your rising CAC. But at the core, you serve an ad, sell a product, and reinvest that profit into another ad. If you’re a consumer brand, you follow the same path, but likely add in other distribution channels, and your marketing has do the double job of getting people to those other locations as well. The principle remains the same: you do some marketing, you sell a product, and the profit allows you to invest more into marketing.
Most e-coms and marketplaces will likely follow the same model. Depending on size of your catalogue, you may have bigger SEO opportunities. But ultimately you’re trying to get big enough affordably that you build up brand equity and become the go-to place for that thing.
And so within a DTC, your growth team will likely need an ads specialist, in the way a social network won’t do.
Remember: you don’t choose your growth engine, your business model does.
Building Growth into your team
I’ve seen first hand in-house in multiple startups, as a coach to dozens more, and as a consultant or agency partner to even more, just how impactful Growth can be.
Growth is not just your ads – though if you’re a consumer business, getting that right will be part of it.
Growth is that process and mindset that brings customer psychology, deep analytical understanding, and science together.
The frustrating part of Growth is that on paper it can seem so simple. And also so similar to lots of other things. “We test subject lines in emails” is not the same as doing Growth. Neither is “running a test on Facebook.”
On paper, Growth is those three things: (1) a scientific method (as run through a growth meeting and sprints), (2) a focus on psychology and behaviour, and (3) a profitable understanding of your growth engine.
Building that into any company, however, is incredibly hard. Growth was its own distinct culture inside Facebook – but one where it had the ability to seemingly do as it wanted. How many tech companies are there today that let their growth teams launch international – or build a free internet – so that you can continue to grow? Very few.
And outside of tech companies, when you think of that much wider majority of tech-enabled businesses, what does Growth look like there?
In the average tech-enabled consumer business, that’ll often include being the ones to grow existing products as much as possible (acquisition & retention), but then also be the ones to lead product development. Don’t give this to ‘NPD’, it needs to be thought of from a Growth perspective.
Growth is a business function, but it is also a mental model of how to run an organisation. That’s why it’s hard to implement, but when done right, the most powerful thing your business can do.
Bites of the week
We are currently renaming our agency and expect a new name by this time next week
We’re hiring a grad! If you’re a recent grad and want to learn about Growth and performance marketing (especially now you know the difference between the two), you should take a look here.
We recently announced some client wins: very happy to announce we’re running full service growth for both Rocketo, and Mother Root.
Facebook: The Inside Story. Stephen Levy
Hacking Growth. Sean Ellis.
Growth Hacker is the new VP Marketing. Andrew Chen
Startup = Growth. Paul Graham.