If you are a professional, these are some alternative terms to use instead of Growth Hacking: User Acq, Growth Optimization, or just Growth.
The hacker term is a buzzword that can be safely ignored. It comes from the belief that people who are good at user acq somehow know a trick. They don’t. They just use data instead of gut feel to make decisions. They also work on marketing and product optimization at the same time instead of viewing those as separate disciplines.
“ Half of my advertising works, I just don’t know which half”
This quote pardons marketers for spending lots of money on things that can’t be measured. Growth Optimization is the opposite of this quote.
The growth approach is to only spend on channels where a profitable cost per acquisition can be calculated.
This is the equation:
When a growth optimizer is thinking about marketing, everything they do is in pursuit of balancing this equation.
The popularity of the growth hacker term comes from the belief that not only do they know how to acquire users, they know how to do it for free.
This This means that they have built a product in such a way that it achieves a viral coefficient.
This is the equation:
When thinking about product decisions, a growth optimizer is primarily concerned with ways to increase the two variables in this equation.
A growth optimizer uses data to fill in the two equations listed above. They make product and marketing changes at the same time to optimize the variables in the two equations. Take a look at an example from the first equation.
This equation doesn’t balance. We are spending $100 and the user is only making us $50. A growth optimizer would look at ways to change the two variables, cost per click and conversion rate. He could lower the cost per click in Adwords while increasing the conversion rate with product optimization. Either one is a lever in reducing the cost per acquisition.
That’s pretty much it. This may seem obvious in retrospect, but it’s a complete 180 from how most startups operate.
I’ve been thinking a lot about the future lately, both near and long-term. I wanted to get my predictions down in writing.
There is a war that is about to explode this Christmas over the living room. A computer will replace the cable box, and there isn’t a clear winner in the space. The winner will be as ubiquitous as the smartphone, and we’re not even close to that yet. Just as the iPhone showed us what a mobile touch operating system looks like, we are still waiting on someone to show us what an operating system looks like that we control from a couch. It’s not the same as a computer, and it’s not like anything we’ve seen yet.
Batteries limit most of the technology we use today. The reason you have annoying cords for your ear buds is due to the fact that we don’t have light, thin, long last batteries to power wireless versions. Batteries dictate the weight shape, size, and useable applications of almost all of our technology.
The problem is physics. They can’t pack electrons any tighter into lithium ion batteries. We’ve hit the ceiling and now we need something new. One promising option is a nobel prize winning substance called graphene. It’s a superconductor with power storage properties. It might not extend battery life for us, but it may reduce re-charge times to a second. In 3-5 years, graphene will be at the core of our most amazing advancements.
There is a bit of movement in this field with Up Bands, and a rumored Apple Watch. This is a just a bit of noise before the biggest paradigm shift since the Internet.
Ray Kurzweil said it best when he pointed out that looking at the Internet through a monitor is the equivalent of looking at the world through a keyhole.
When we were young we thought of the Internet in terms of “going online”. Now the Internet is just on. Ask a child the difference between online and offline and they won’t know what you’re talking about. The same is going to take place with computers. There won’t be a separation between being awake and being connected.
Google Glass is the first prototype in this field. It’s not a product that’s ready for mass consumption, but it is a necessary first step to figure out how this new paradigm is going to work. Very few people are going to wear something that looks so absurd, but soon it will look like any other pair of glasses. Eventually it will be a contact lens and at some point in the future it will likely be an implant.
I can already hear the cries of people saying that they don’t want to always be online. Take a look at the employment rate amongst people who can’t use computers. You won’t have a choice, and to be honest, you won’t want one.
Thinking about a Twitter stream projected onto your retina probably isn’t that appealing, but that’s not what this is going to be. A great deal of how we experience computers is limited by the constraint of having to displaying information in 2D on a flat screen.
Eventually information will just be overlaid onto reality. Augmented Reality is the term if you want to look it up. When we look at something online we expect there to be information and context. Prices, ratings, specs, explanations, and comparisons are standard in the digital world. Now imagine all that information overlaid onto the real world. Look at a person and their name and background will pop up.
We’ve spent a lot of effort trying to make computers more brain like. It’s a lot easier to just marry our brain with a computer.
Right now our interface with computers happens tactically by typing and touching. This is incredibly inefficient way to interact with a computer.
The first challenge is to allow a machine to read the information stored in our mind. This is a lot closer to a reality than you might imagine. Watch this for a demonstration. Once a computer can interpret our thoughts, we are freed from the ball and chain of our keyboard. The visual world won’t just be augmented with extra data; our brain will literally be augmented with a computer.
If you want to remember something, you won’t have to repeat it to yourself. Think to your internal computer that you want to save it, and it will be done. Retrieval of information will be just as easy. You won’t forget anything, ever. You will able to do calculations at lightning speed. All of the daily mental power that goes to low-level mundane tasks will cease. The computer will take care of it, and you will be left to think. A million productivity apps have made this promise and failed, but that’s because it’s painful to tell the computer want you want.
Right now, when we want to create something with a computer, we need to write technical instructions for it to understand. It requires thousands of lines of code, or hours of clicks in software like Photoshop and CAD. We are translating what’s in our mind into a language the computer can read.
If the computer can read our thoughts, all of those technical instructions become unnecessary. We will think of something, and then the computer will make it so.
Here’s a rundown of my work setup. I’ve tested all of these products extensively and consider them indispensable.
Computer: Macbook Air-13 inch 256GB
Monitor: Thunderbolt. For a long time I thought that this was an overpriced monitor. I was very wrong. It makes every other monitor I’ve used look broken in comparison. It definitely reduces eye strain at the end of the day.
Sizeup: SizeUp allows you to quickly resize and position your windows with keyboard shortcuts. They call it the missing window manager and I have to agree. I mainly use it to resize windows to the left and right half of the screen. This allows me to take advantage of all the screen real-estate on a cinema display monitor. For keyboard mappings I have option-q = Left, option-w = Fullscreen, option-E = Right.
Jumpcut: Shows the last 10 things I copied to the clipboard.
Quicksilver: It’s similar to spotlight, but I find that it does a much better job at launching applications. I also use it to control iTunes from the keyboard while I’m working in other apps. I have keyboard shortcuts setup to rate tracks and display the currently playing song on the screen.
Dropbox: I save all my created documents in Dropbox so it’s always backend up and available on other computers.
Notational Velocity Alt: I use it to take all my notes. I find Evernote to be really heavy and it’s footprint is much too big. Notational velocity takes up a tiny amount of screen real-estate so I can use it in conjunction with other apps while I’m working. The Alt version adds the ability for theming. I use a black background with white text, which reduces eyestrain.
Simplenote: I use this on my iPhone for note taking. It works perfectly with Notational Velocity so text is always synced between my desktop and phone.
Afloat: I use afloat on Notational Velocity to keep that window on top of other windows. I often have an app open full screen, but I need to take notes or refer to them. I hate having to switch windows, so Afloat just keeps that windows on top of all the others.
F.lux: Looking at a screen late at night when the surrounding environment is dark, increases eyestrain. Flux gradually adds a tint to the screen at night to make it easier on the eyes.
Streambox: Having a music service like Pandora in the browser just doesn’t work. I accidentally close it all the time, and it’s hard to find the right tab to see what’s playing. Streambox is a Pandora app that sits in the menu bar.
Airfoil: I love this app. It allows you to beam music from your desktop to another desktop, iPhone, or Airport.
Adobe Creative Suite: Pretty standard. I haven’t tried Cloud yet, but I really skeptical. I’m still on CS6 desktop version.
JMP: I use this for statistical data analysis. It’s similar to R but with a usable graphic interface. It’s ability to quickly clean data sets is unparalleled. I use it to quickly run distributions, get rid of outliers, and look for correlations in data. There is way more in there than I know how to use. If you’re reading this you’ll probably be interested in it’s ability to do logistic regressions, and partition analysis.
Tableau: I just started using Tableau but so far I’m really impressed. I would describe it as Excel PivotTables on crack. It makes manipulating and visualizing data very easy. I can almost always do the work in Excel, but Tableau does it in minutes instead of hours. Also, when I’m dealing a very large data set, Excel chokes and Tableau handles it like a pro. Tableau is Windows only so I run it in Windows with Parallels.
Rubymine: I’ve just started to get better at programming. Rubymine is a full featured editor and it makes it a lot easier for a beginner.
Chrome Canary: I recently discovered Google’s beta version of Chrome, which they call Canary. It’s a completely separate browser from Chrome, so I can be signed into my work account. I use regular Chrome to sign into my personal gmail.
Skitch: I use Skitch for screen-capture and annotation. It was a lot better before Evernote acquired it, but it’s still decent.
Pinboard: I bookmark all my web findings in Pinboard. It’s a Delicious clone which I signed up for when it looked like Delicous was going to be shut down.
Mailbox: I use Mailbox for work email on my phone. Their ability to snooze an email until later is perfect for work email.
37 Signals Backpack: I used Backpack to write all my technical specs. They unfortunately stopped supporting it and new users can’t sign up. I am in the process of building a new version for writing specs.
When I’m working on a user acquisition project, I use Optmizely to A/B test the email capture landing page. I always create a custom event that triggers when the user successfully submits an email. Unfortunately, Optmizely doesn’t have a proper API which makes doing this difficult without redirecting to a new page.
First create a regex function to validate an email
Next, create a function that fires when the user clicks the submit button
Put the entered email into a variable
Write an if statement that uses our ValidEmail function to test if the email is valid
If the email is valid, send the event to Optimizely
In the last line we are calling our event ‘email_submit’. You can change this to any event name that you want.
All together it looks like this:
In Optimizely create a new custom event goal. Set “Custom Event to track” to email_submit.
Next week we’ll be releasing an app called Matchbook. Signup to be notified when it’s out. We’re a proponent of the lean startup methodology, so we wanted to share the process we used to get this app out the door.
We like to build software that mimics real life. The goal of software should be to make already occurring behavior easier, not to create new behavior. So, if you’ve ever taken a matchbook from a restaurant to remember it later, then you have an understanding of what this app does. Matchbook is a dead simple bookmarking application for places. When someone gives you a recommendation about a bar, restaurant, or shop you can bookmark it. The app will organize those places so you can make a fast decision about where to go out. We’ve heard it described as Delicious or Instapaper for places.
I called up a buddy I often discuss tech with and said, “Something is nagging me about the location based space. It doesn’t feel like mainstream America is quite ready for the check-in.” The question became, “What type of location based activities are normal people ready for?”
Mobile location research should be preformed in real locations, outside of the office. To answer our question we sought out feedback from normal people instead of from the tech industry.
To achieve this we planted ourselves at a bar, approached groups of people, told them we were about to build an app, and asked some questions. We also used the dating site HowAboutWe.com to go on dates so we had the undivided attention of a female for market research. No judgment; we paid for dinner. This turned out to be a great place to do market research because:
This is what we found:
We started wireframing the app in Omnigraffle. We spent most of our time removing features until we had what we thought might be the minimum viable product.We went back out to the bars and tested them. We rigged up a clickable prototype with a great app called Interface that allowed us to do our user testing. We would get a nights worth of feedback, re-do our wireframes, and then go back out. We iterated through this process about 30 times.
We kept going until:
When we began, we thought that Matchbook would be a social app. We envisioned it helping people make plans, share tips, or share bookmarked places. As we talked to more women, we found th
at they were a little burned out on social and a more then a little concerned about sharing their location. The number of women that perfectly articulated the
social circles problem was amazing. As a result, our wireframes pivoted away from social and became a personal app. We will probably add in social in the future, but we need to rethink exactly how that should work for this market.
The MVP is a bookmarking application for places. The user can:
Once we had our MVP, we moved onto the development phase. We outsourced the entire thing, which involved a good chunk of time spent iterating through developers instead of code. That will be the subject of another post, but in the end we found a great team. My co-founder and I developed the entire thing for about $10,000, paid for out of our savings.
A key problem with building an iPhone app is that Apple only allows 100 slots for beta testers. This was rough as we tried to test our assumptions. We needed to ASK all of our users to download it, which skews the data.
After some brainstorming we came up with an alternative. We are going to launch in the Canadian app store first. Since we can’t do a private beta, this will be our beta test. People in the US can’t see the Canadian app store so we will localize things there. We’ll use our Canadian launch to get feedback and gather metrics.
Once we’ve iterated based on that feedback we’ll launch a more polished product in the US app store. The idea is to couple the download traffic from launch PR, with the iTunes Recently Released app list. This concentration of downloads will hopefully bump us onto a Top Downloads list in our category.
These are the assumptions our lean process has yielded. We will be testing these in Canada next week:
We started with this step at the same time as Step 3. We decided that offering local deals is the best bet for monetizing a location based startup. Since we don’t have the money for a sales force we began our customer development process by speaking with group buying sites. We found out that they:
To better understand the group buying market, we offered to help out a NY based group buying site with their metrics. This gave us enormous insight into the types of challenges our customers face, and we learned great tactics for optimizing daily deal sales.
That’s it for now. The app will be out in Canada in a week, and out in the US shortly after.
Thank for reading,
For the last two years I have been building a company called Matchbook. It’s an iPhone app for saving those must-remember restaurants that are probably cluttering up the notepad of your phone.
It’s been fascinating to build an iPhone app, talk to excited users, and learn about mobile user acquisition. I built it into a popular app with a very engaged user base, but now it’s time for me to move on to new projects. At the start of the year I sold Matchbook to Quotidian Ventures. There is a new team of great people who will be caring for it.
This is not one of those situations where I say it was acquired, only to have it shut down the next month. The new team is focused on Matchbook full-time, and they will continue to improve it. For those of you who are users, thank you for all your feedback and support. I’m leaving you in good hands.
Cross-posted from Business Insider
“Love ‘em or hate ‘em, you have to admit GoDaddy’s Super Bowl ads are effective.”
That’s how Mashable opened their recent blog post on GoDaddy’s biggest sales day in the company’s history. What’s perplexing is that the numbers don’t add up to a profitable campaign. In fact, they seem to suggest that GoDaddy lost up to $7 million dollars. Despite this, I applaud GoDaddy for releasing these numbers. This type of transparency keeps the cost of advertising in check by allowing us to calculate its value.
Here are the stats from Mashable:
- Hosting sales jumped 45%.
- Dot-com domain sales rose 40%.
- New mobile customers increased by 35%.
- The company added 10,000 customers in total.
We need a few more stats to calculate GoDaddy’s ROI:
That makes their customer conversation rate .009% (10,000 customers/108.41 million viewers).
The Cost Per Acquisition for GoDaddy is $750 ($7.5 million / 10,000 customers).
A campaign is only considered successful if it acquires a customer for a cost less than the customer’s lifetime value. In this case the lifetime value of a GoDaddy customer would need to be greater than $750 for this campaign to be considered successful.
Let’s gather a few more data points from GoDaddy’s website to calculate LTV (lifetime value):
Let’s assume that the average hosting customer stays with GoDaddy for 1.5 years, making them worth $89.82 ($4.99 hosting cost x 18 months). Some customers will cancel after a month or two, and some will stick around for a long time; 1.5 years is an educated guess on the average.
Let’s also assume that the average domain renews at least once since that happens automatically, making it worth $18.78 ($9.39 x 2).
Assuming that there was a 50/50 split in the products purchased, the lifetime value of an acquired GoDaddy user is $54.30 ($89.92 hosting + $18.78 domain / 2). That means that GoDaddy lost $660 per user ($750 CPA – $54.30 LTV) or $7 million dollars ($7.5 million – $54.30 LTV X 10,000 customers).
To be fair, GoDaddy was only reporting the number of customers they received on Monday, the day after the Super Bowl. They will likely continue to gain new customers for a longer period of time. Let’s assume that each day they acquire 20% fewer customers, as a result of the Super Bowl ad, than they did the day before. This should account for the ripple effect of media attention. That would mean that the Super Bowl ad would have a one-month effect, and GoDaddy would end up with 50,000 new customers. If we re-do our calculations:
Cost per acquisition: $150
Lifetime Value is still: $54.39
That means they lost $95.61 per user, or $4.7 million total.
I hear a lot of talk from marketers about intangible benefits. Maybe the intangible benefits of running a Super Bowl ad is equal to $4.7 million, but I have a hard time wrapping my head around that. An often-cited explanation is that the media buzz from running a Super Bowl ad exceeds the cost of the ad. According to the Wall Street Journal, companies value the media coverage at an additional $10-20 million dollars in equivalent advertising. Measuring the return on investment by calculating the value of the additional advertising they receive seems circular. Why aren’t they measuring their return on investment in sales?
This isn’t a condemnation of GoDaddy. It seems that the ad industry is significantly overcharging for their Super Bowl slots, and not delivering on their promise of customers. Companies are willing to pay, so you could argue that they are valued appropriately for demand. However, as we move away from traditional marketing and towards measurable user acquisition, I imagine that will change.
The success of LivingSocial and GroupOn has garnered a lot of interest in their business model. Almost every content site is evaluating whether they should add a daily deals component to their business. This is a guide to decide whether or not you should start a daily deals business, and if so how to go about doing it. I take a very metrics driven approach. If you are evaluating this from a marketing and strategy perspective, I would refer to Yipit’s guide here.
As a metrics consultant, I have helped a number of companies in the daily deals space through this process. As a word of warning, this is EXTREMELY hard. The simplicity of the user-facing software makes this business look deceptively easy. All of the complexity is hidden behind the scenes. That said, here we go.
Part 1: Should I Start a Daily Deals Business?
We won’t build anything like a Grand Central Station, cathedral, monument, or pyramid. Everything we build is temporary and will shortly disappear.
Our culture will be gone too. Written documents will go out of style with printed books to make way for cheaper and more efficient digital files. The day we make the full transition to the digital world will be the day that the cultural trail goes cold.
In the irony of being the most documented, over sharing generation ever to live, that digital information will be lost. Services will shut down, technology will change, and the electricity may one day go out. A hiccup in our civilization caused by war or worse will cause the network to crash. It wasn't arch
itected for the relentless erosion of time. It's built on a flimsy patchwork of services that barely work when given the constant attention of engineers.
If the system goes down, even for a brief time, all that data will be gone. It will be locked on magnetic disks and memory chips in ones and zeros, and we will lack the precise technology to decipher them. Our focus will be on getting the system and our civilization back up and running. The task of figuring out how to retrieve all of that data will be insurmountable. It will be stored on machines that no longer work, in software that no longer runs, written in languages that are no longer used.
It's already happening. The data and high quality footage from our first moon landing, a seminal event for our civilization, is locked on giant reels of magnetic film. There are no machines left to read their contents; it might as well not exist.
The digital representation of our culture will become dead bits, and instead of reviving them we'll start over. My blog seems like the perfect place to lament this eventuality. In a short time it will disappear, along with everything else posted here.
Cross posted from my guest contribution on Business Insider.
Blippy failed because they didn't understand social. They built a product to broadcast credit card purchases. They based their social features on the model, “Friend Everyone, Share Everything.”
This has been the mantra of social software since the rise of Facebook and that era is reaching its end. Going forward, sites that follow this model will fail, just like Blippy, Apple's Ping, and Google's Buzz.
I'm not saying that people will stop sharing their every thought, photo, and link. People love this freedom of expression. Facebook, Twitter, and LinkedIn will be fine because they form the foundation of the social web. They are the proverbial phone book for our society, and that's a necessary function. The last five years has been the rise of the Social Network as a utility platform. Now that it’s created, the next five years will be about building on that foundation. New social apps will focus on sharing with select groups of people. In that sense, The Era of The Social Network has ended, and we are now entering The Era of the Social Circle.
To illustrate how I see this flux happening, I turn to birthday parties.
When a pre-schooler has a birthday party they invite the entire class, which constitutes everyone they've ever met. It's not until years later that they develop friendships and social circles that form the basis of inviting one person over the other.
Social Networks are like that pre-schooler’s party. Rather then being made up of “friends”, they are repositories of people whose existence you acknowledge. You connect with everyone you've ever met because the tools have not been developed to create more complex social circles that make up mature adult relationships. It makes sense that the evolution of digital social will follow the same developmental path we take as humans.
Social Networks are characterized by, “Friend Everyone, Share Everything.” Social Circles will be characterized by “Group Dynamically, Share Selectively”.
Social Circles will focus on groups, automatically created based on a person’s real life social circles. They will dynamically shift to include new close friends and remove those that become distant. “Friends” will come to mea
n the same thing as it does in the real world, a group of people whom you share a close connection with. Content will be shared selectively with those that it’s most relevant to, mirroring the intimate sharing of real-life friendships.
I can already hear the arguments that Social Networks allow for this type of interaction. They don't, and let me prove it you.
I've had this conversation again and again in the last year.
Me: Do you use a check-in service
Them: I don't want EVERYONE to know where I am
Me: But you don't have to let everyone know where you are, just add your closest contacts as friends
Them: That won't work. I use Facbeook and I get friend requests from everyone I've ever met, I feel bad, and then I accept them.
When people talk about privacy concerns around check-ins, it's not that they don't want ANYONE to know where they are, they don't want EVERYONE to know where they are. The problem is the lack of tools to create the dynamic, intimate groups of people they would be comfortable sharing with. Once that problem is solved, this space will see a huge uptick from mainstream users.
Facebook has quite a few tools for managing groups of friends. These fail because they rely on the user to manually curate these groups. Users won’t do the manual work necessary to make a Social Circle work, just like they won't be selective with whom they friend on a check-in service.
Solving this problem is going to be very tricky. If you want to know why Path and Color were able to raise so much money, it’s because they are tackling this problem, not because they solved it. The investors were placing a bet that if anyone can solve something as difficult as this, it will be those teams. As for me, I’ll place a bet on anyone that’s willing to put “Friend Everyone, Share Everything” aside to build software that “Groups Dynamically, and Shares Selectively”. That’s what we’re doing with our app Matchbook, which is for bookmarking place recommendations. If we take our cue from human development, it’s going to be hard, but it’s time to grow up.