How to Design Growth: Growth Hacking 101 at Seedcamp in Rome

Introduction: this presentation was performed on the 5th of June 2014, at Seedcamp in Rome. I want to thank Seedcamp, the organisation, Nicola Mattina, Giuliano Iacobelli, Carlos Espinal and my company, Automattic, for giving me the opportunity to speak in front of such a great audience.

Growth Hacking 101

Good afternoon! My name is Luca Sartoni, I work at Automattic, the company behind and many other products and today I’m gonna introduce you to Growth Hacking.

How many of you are from Rome, the eternal city? A couple of hours from here there is a lovely town called Assisi, where many many years ago San Francis used to shock the system with his unorthodox approach to religion.

Many of you, know this prayer, attributed to him:

“Lord, grant me the strength to accept the things I cannot change,
the courage to change the things I can,
and the wisdom to know the difference.”

How many of you knew this prayer before?

So, it turned out that this prayer was mistakenly attributed to San Francis, while in reality it was written by Reinhold Niebuhr, an american theologist about 700 years after.

You know, Abraham Lincoln was right when he said:

“The problem with quotes found on the internet is you have no way of confirming their authenticity.”

If the Internet tricks us on simple attribution of quotes, how can we make good decisions as entrepreneurs?
How can we focus on what matters and make the right choices if we don’t know what is right and what is wrong?

How many of you are entrepreneurs? How many of you have a startup?

So I have a question for you: what is a startup?

Paul Graham, Co-founder of Y-combinator seed capital firm defined a startup as a company designed to grow fast.

How to design growth


There are things that we can’t control. We have to deal with it. I know that it’s difficult to accept it as a fact, but this is how reality works. We cannot change how people behave, how people nurture and cultivate their interests, how they interact with devices and interfaces. Of course we can come up with interesting elements of design, we can dramatically change people’s life. Just imagine how Apple impacted our world since the introduction of the iPhone in 2007. But they just didn’t show up one day out of the blue. There were years of research and development, a strong vision and hard work.

In a single sentence: “Growth is a matter of culture”.


So first element for a strong growth by design is education. Hire diversity and focus on all the abilities that as a technical team probably you lack of.

At WordPress we say: “code is poetry”. That wants to be an inclusive statement. Everybody should be included in your developmentm process. Every single diverse input is precious and you cannot even imagine to start a global project if you don’t have those elements of contamination in your company’s DNA.

Hire anthropologists, philosophers, artists. Give them a technical platform and education. Teach them code, let them contaminate your company and you’ll have much more chances to succeed. You cannot a afford to have an solid technology company if you just rely on pure technology, you need people to resonate with other people.

Make sure that everybody understands how growth is relevant to their sigle contribution to the company and how they can deliver a better product.

A common misconception is that growth is all about user acquisition or all about revenues. This is dead wrong. It depends and it’s a case to case deal. There are projects that make sense to focus on acquisition only, other project are focused on revenues, other on more strategical indicators. This is something you need to define and then be honest to yourself.

If you jump from one metric to the other depending on what is good this week, you are not going very far. It can work for a few months but it’s not gonna work on the long run. Be honest with yourself and define your key metrics. They can change, but only as a result of a pivoting, not as a consequence of the weekly performance report.

In order to design your growth you need a scientific approach, contaminated with all the possible diversity you can. But the core of your approach needs to be mainly scientific.

Which tools we can borrow from science and apply to growth?


Let’s have a look at three must-have tools for growth: A/B tests, usability tests and sliding windows.

How many of you are familiar with A/B tests? Very simple: given a population, we randomly assign them to groups (bucketing) and we try different design elements or flows according to the groups they belong to. We analyse the results and we determine which element or flow worked best.

There is an entire universe behind A/B testing, and there are many cognitive traps you can fall into. So please study and improve your A/B testing processes. But let me give you three quick tips on how to avoid the classic mistakes: seasonality, meaningful measurement of impact, micro optimisation fallacy.

Weekly seasonality: it does not matter how much traffic you have, people on monday afternoon behave differently than on thursday morning. I always run tests for a week, at least. This helps me to have more complete sample of the population and helps to reach the statistical significance.

Meaningful measurement of impact: let’s say you have a complete marketing funnel like this, from awareness to revenues. Let’s say you run a test on the signup interface in order to increase activation. If your test involves a specific point of the funnel, use a relevant metric within the same segment of the funnel. If you measure revenues, for instance, there is a high chance to have meaningless results because many other variables are able to influence that metric.

Micro optimisation fallacy: this is a tricky one because it is a direct consequence of the previous step. If you do it right measuring within the same step of the funnel, there is a good chance that one day or the other you fall into this fallacy. This happens when you reach points of relative optimisation and you don’t realise that there are better opportunities just because you have a short sight. How to overcome this? Strong vision, bold decisions and experience.

Usability tests are big helpers on this. Are you familiar with usability tests?

They are easy to explain, very difficult to perform correctly.

You simply take real people and you ask them to try out your product. You ask them to perform key tasks and you study how they behave. Is sounds simple, right? Well, it’s not THAT simple but with a few tips you can get good results from the very first test.

Usability tests help you to find roadblocks and major fails. So, when it’s the right time to run usability tests?

When your site is ready for going live? NO
When your producs is almost ready for shipping? NO
When you MVP is almost ready? NO

The sooner the better! Ask people to try out stuff as soon as you have the idea. Ask them what they think of it. Gather costructive criticism as soon as you can.

As soon as you have a piece of interface, let them go through it, and keep looking where they get stuck. And then ask them why they got stuck.

The big advice I want to give you here is: don’t listen to them when they tell you how it should be done. I know, it sounds awkward but it’s true. If they were able to make it better, they probably would have made it better before you. They are not your engineers, they are your testers, so listen to them carefully about the problems they had, but most of the time their solutions are simply not good enough and they are not the right people for that.

Judge them by facts, not by their words. So if they say that something was easy to perform, don’t trust them. Measure how much time they spent on it and compare it with other people’s performances. When they find a major roadblock or say that something is broken, ask them questions to understand why they find it broken, not on how to fix it.

The third and last tool it’s important to know it’s very trivial. But very tricky at the same time: sliding windows.
Always compare metrics on a 28 days basis. This way you always include 4 weeks of data, you don’t have longer and shorter months and the comparison is always normalised.

Just to wrap up, let’s have a look at the serenity prayer, with a modern growth twist:

“Grant me the ability to adapt to the things I cannot change,
the culture to change the things I can,
and the right tools to know the difference.”

Thank you very much, if you have any question, drop a comment.


Traffic vs Distribution

We all hate being stuck in our car, jammed for hours on our way back home. We are not actually stuck in the traffic, we are the traffic. Nevertheless we keep on considering it a necessary pain if we want to go somewhere.

There is another form of traffic that we love, instead: online traffic.

Every single online project owner I met in my career wanted to have more traffic on his websites. And they are ready to do whatever it takes to increase that traffic. They invest in SEO consultants, SEA agencies, they buy traffic on Adwords or FB advertising.

But does it really make sense to consider the traffic as the primary goal for an online project? My answer is no. Not even if the project lives of ADs. Let me explain why.

I have a tech background and in the past I used to design networks and firewalls. When you design a network you consider the traffic a cost because if you expect a large amoount of trafic you need bigger infrastructures and more expensive devices. Exactly like the road traffic, the more the cars the more the hassle.

So why people tend to consider traffic as a positive indicator for their online projects? Because today traffic is very cheap for them. Traffic does not introduce a large cost and modern infrastructures are able to cheaply handle an amount of traffic that is much bigger the size of their business.

But in reality they don’t want traffic, they want business. That is declined in sales for ecommerce sites, clicks for ADs sites, views for news sites, contacts for SMEs, etc.

Because traffic can increment all of them indistinctly, people tend to consider traffic as the holy grail to acchieve success. In reality they are not focusing on what matter the most: business.

My suggestion is to change perspective and start thinking about distribution instead of mere traffic.

What is distribution? Distribution is a way to take your product to your customer or your customer to your product.

Distribution makes you sell more. Even better, distribution makes you get your business objectives. Just by definition. It’s not about sending as much people as possible to your funnel, it’s about having different funnels on different channels and adapt every single funnel to that specific segment of customers. It involves multiple experiences and multiple platforms where only consistency is the limit.

Distribution makes you sell more because you stop thinking how to send people to a landing page and you start thinking where people have needs and how I can reach them to provide my solutions. You stop wasting your time on Google Analytics and you start deeper analysis on your sales funnel and integrated strategy.

Traffic is a fad, distribution is solid. Traffic can stop when Google changes its algorithm, your SEA budget is over, your website is down. Distribution always works because it’s a mindset, a process and have multiple implementations.

Can traffic be considered a form of distribution? Yes sometimes. And the hard job is to make sure that traffic is used as a form of distribution and it’s not as the primary metric to take decision upon.

News work

Asking is better that guessing: Qualaroo

How many times you got stuck in a meeting with designers, developers, project managers and sometimes people from the MGMT in the middle of a religious debate?

No I’m not referring to real Religion. I’m talking about something that usually goes like:

– “I think our users wants to get $this_product as soon as they get on our website”
– “No way! It’s clear that they want to see $other_product first!”
– “Most users don’t like $other_product because it’s too advanced”


Steve Krug has a very similar example in his book “Don’t Make me think” so I borrowed his definition of “Religious debate” for my example as well. Ask Krug says:

They consist largely of people expressing strongly held personal beliefs about things that can’t be proven-supposedly in the interest of agreeing on the best way to do something important.

But we are scientists, right? So let’s do something scientific: let’s ask our users what they want!

Here comes Qualaroo, a simple yet powerful survey tool. With a simple snippet of code on your pages you can ask any kind of question to your visitors.

The tool is very powerful and I’m going to get in details on how to use it in the future, by now just remember that asking is much better than guessing even if users tend to lie, forget and change (cit. Luca Mascaro).


Recap on my online presence: blog, portfolio, twitter, linkedin, flickr

Main Blog: general purpose aggregator

On my main blog I post stuff that I like and that is somehow realted to my work/life/dreams/expectations/career. I don’t want to focus on a specific topic and I have to admit that lately I didn’t post that often. This does not mean that I’m not interested in keeping it open. I post from time to time so subscribe to it if you like. You can subscribe via email here!

Twitter: news channel

Everything that happens to me or that I like, goes straight to my Twitter stream. Follow me if you want to keep up to date!
Link: @lucasartoni

Portfolio blog: photography experience

Where I display my photo portfolio and I regularly blog about photography. Most of the times are photo sessions but I also like to reblog other people’s inspiring stuff. I’m trying to keep a regular activity there, have a look and leave feedback!

Flickr: photo repository

All of my public pictures go here. I don’t have a theme or a topic. Every single image that I publish goes here. I use Flickr as the main repository for my final images and as a distribution channel for my photos. All of my images go out under CC-BY and it’s so easy to send people here when they want to download a picture I took. I organize the pictures into sets in order to keep everithing together.

LinkedIn: professional life

On LinkedIn you can find all my professional activities. I don’t produce specific content for LinkedIn yet, but I’m seriously thinking about it. It looks like a good idea. At the moment is my curated friends’ repository. I strictly select people to be connected with. So if you want to connect with me don’t use a standard request. Tell me where we met or why it should be appropriate to be connected. I tend to refuse standard requests from unknown people.


Twelve Truths I learned (& never want to have to say again)

  1. The Internet does not exist: it perfectly matches the physical world, we go there as we go to the office, to a bar or as we enter our bedroom. The only real difference with respect to physical environments is that it allows us to be ubiquitous.
  2. On the Internet there aren’t different conversations, it’s that you listen to conversations of people very different from you.
  3. Online relationships are just like offline relationships: a few of them are deep, many are superficial, as many are opportunistic, mannered or false.
  4. Technology enables the change, it does not generate it: a person who has nothing to say or write does not become active and generous only because they may do so. Explaining the change to him again and again is like explaining a joke if one did not laugh the first time you told it.
  5. Technology enables the talent where it exists, it does not create it.
  6. Digital natives are accustomed to technology, not aware of all its potential: being amazed or disappointed that they use Facebook to comment on X-Factor rather than to make a revolution is like giving me a pole and wondering why I’m not jumping from one rooftop the other.
  7. The awareness of the significance of a medium (any medium) belongs to a minority of professionals. Bridging the digital divide does not mean making everyone professionals.
  8. In being an enabler and not the cause of the change, digital media cannot be classified as good or bad, right or wrong, useful or dangerous.
    Labeling them with those attributes is just a cheap shot to win the stage.
  9. If someone – even competent – explains in detail the issues of the Internet, they’re are just telling you about their problems with the Internet.
  10. The Internet is a medium in which the spoken language prevails, or even better, the transcribed thought. We need a new syntax.
  11. Most of the exchanges that occur online are phatic nature, no transmission of information.
  12. It’s about the story, not the book.

This list was posted yesterday by Mafe De Baggis in italian. With her consent (and Sara Rosso‘s help) I republished it in english.


Tonight at Online Tuesday #18 in Amsterdam


Tonight at 19.30 I will be speaking about vertical search engines and people search at Online Tuesday #18 in Amsterdam.

I will be in Amsterdam until friday to attend the Picnic Festival 2011. If you are around send me a tweet at @lucasartoni and we can go for a coffee.