Growth Engineering is like Space Exploration: A Series on 10 Growth Experiments in 10 Weeks

The battle with growth engineering is deciding which strategies to prioritize. There is an overwhelming number of options for customer acquisition and each has a wealth of information across the web; however, the options are rarely compared against each other.

For example, if you want to learn about AdWords, there are videos, books, and courses dedicated to the topic. But when should you use AdWords rather than Display or Twitter marketing? Or should you used paid at all? Rather than fully investing in one or two acquisition options from the start, we’re running 10 experiments to see which specific growth strategies are the most promising for Kiip’s business model and current state.1 Growth Engineering is like Space Exploration. With finely-tuned machinery and the courage to explore the unknown, you’ll make remarkable discoveries.

Growth Hacking is Like Space Exploration

Before experimenting, setup your site to track customer acquisition sources

The most important part of growth hacking is tracking metrics. If you don’t track metrics, you won’t know which strategies to drop and which to fully invest in. We chose Google Analytics for its ease of integration and flexible feature set. The features we use are simple, but insightful. A conversion goal was set up to track advertiser registrations and all campaign URLs are properly built with utm tags so we can track attribution. We also use acquisition channels to track how much traffic each referral source brings. It’s an easy way to visualize the contributions of each referral source towards conversion goals and raw visitor volume.

Google Analytics Referral Sources
Google Analytics Referral Sources

Our analytics infrastructure is simple to start — tracking visit volume and goal conversions by referral source. Rather than building out analytics features that might go unused, we started with the essentials and will build out advanced features as required. We also put AdRoll into our pages to retarget all this traffic at a later time. (Foreshadowing…..)

Determine a clear, valuable performance metric to assess your experiments

All experiments will be evaluated on a cost per signup to ensure a comparable data set. Cost is a function of both time spent on the experiment and real dollars spent. For example, if we spent 5 hours and $1,000 experimenting with AdWords to drive 20 signups, the result would be:

(1000 + 5*40) / 20 = $60/signup

Why a cost per signup rather than cost per paying customer? Right now our sales process requires a fair amount of education, as mobile is still new. Additionally, our website and platform is not smooth enough to bring a customer from cold landing page to credit card signup. Our sales team is great at closing leads, so focusing on driving high-quality traffic and letting sales wrap it up is the most efficient option right now.

One thing is certain — at the end of these 10 weeks, we’ll have proven which customer acquisition strategies are unlikely to work for Kiip. And maybe we’ll have found one that has potential.

Where to Learn More

Although I dislike the term, “growth hacking” is the most common description for these experiments. I prefer the terms “growth engineering” or “growth science”, as the process requires technical ability, systematic experimentation, and careful observance of key metrics. However, if you are searching for more information, “Growth Hacking” is your best bet as a keyword. Here are our favorite Growth Hacking/Engineering resources:

Find a Growth Hacker for your Startup by Sean Ellis, arguably the first appearance of “growth hacker”.

Growth Hacker is the new VP of Marketing by Andrew Chen, the post that made growth hacking a recognizable term.

The 5 Phases of Growth Hacking by Ryan Holiday is a great intro to growth hacking, complete with all the famous examples.

21 Actionable Growth Hacking Tactics by Jon Yongfook gets you thinking about the world of growth hacking possibilities.


  1. To give context to these experiments, Kiip is a mobile rewards network. We reward moments of achievement in mobile apps (Android and iOS) with relevant brand offers. When you finish a 5k run in your fitness app, Propel is there with a free sample. When you finally beat level 15 in your favorite game, McDonald’s is there with a free MP3. By rewarding achievements, rather than spraying brand messaging to every mobile display, our model builds brand love and performance results. 

More from Kiip

  • I think that it’s a mistake to see something like AdWords as a “source of traffic” independent from the development, say, of your website. One of the biggest advantages of AdWords is that it is a source of market research data as much as it is a form of marketing.

    In the past, because clicks were so cheap, you could just “put $1,000 into AdWords”, get several thousand visitors and see how many of them signed up.

    As Google gradually implements changes to more accurately price traffic (ie. both organic and paid traffic through Google are getting more expensive) it’s becoming more and more important to execute your marketing and sales tactics integrated into a single, overarching strategy (could this be the definition of “Growth Hacking” we’ve all been looking for??)

    Referring specifically to your stated goals of testing “cost per sign-up”, what happens if you spend $1,000 on AdWords but you get $0 signups? Or your cost per sign-up is $500 because you only get 2 sign-ups? Or you spend $1,000 in the first 12 hours because you were bidding too high?

    More likely what you’ll need to do in order to make AdWords work for your business is:

    1) Start with some exploratory spending (betwen $500 – $1,000 is usually a good amount) to identify the best search term opportunities to target that are both plentiful enough in volume that you can learn something quickly but cheap enough that you won’t send yourselves to the poor house while you fail
    2) Develop lead generation strategies to make that traffic convert (eventually)

    In a lot of cases for tech startups the issue is that either people aren’t looking for what you do (in the case that you’re creating a new market or re-segmenting an existing one as a niche player) or that the traffic is so expensive you go bankrupt experimenting.

    In both cases it’s most likely that the solution lies in finding an affordable traffic niche in which people who are *in your target market* are searching, even if they’re not searching for your product specifically, figuring out a way to service their query cost effectively and then nurturing that lead through into a sale.

    To this end what you really need to do is setup a conversion tracking event that will indicate (with a high likelihood) that a visitor is *in your target market* without necessarily requiring the much higher commitment of signing up (for example viewing the pricing page, or clicking “learn more”, or even just not bouncing, which is easily tracked by setting up a conversion which fires on every page apart from the home page via a rule in Google Tag Manager).

    This is basically what I discuss in the Keyword Research video from my eBook (full 20 minute video linked below):

    http://iain.wistia.com/medias/foesgvs63j

    I don’t think that “taking a week to throw $1,000 at AdWords” will ever yield positive results (well, sometimes, but it’s the exception, not the rule), however taking a week to throw some money at AdWords can do a lot to inform your approach to almost every other form of online marketing, so it’s well worth doing.

    Also note re: AdRoll that you need a retargeting list of 1,000+ before you can start running your ads, so although it’s a good idea to put the snippet on your page right away, you may find you’re unable to start retargeting your AdWords traffic right away. It’s a great tactic, though, and AdRoll allows you to retroactively segment your visitors so you can change your ad depending on which pages people visited (this is especially useful when you’re targeting your AdWords spend in exact match keywords because you can “continue the journey” by creating landing pages just for specific keywords, and then segment those visitors in AdRoll so you can show ads to people based on the search term they originally used to find you).

    • KFishner

      Hey Iain!

      Thanks for the thoughtful comment. I completely agree with you about the benefits of AdWords are beyond a “traffic source.” The second post in this 10in10 series is on AdWords, and the post speaks to AdWords’ value as a tool to test various aspects of a business and website. I think you’ll like the message :)

      We’re also on the same page with AdRoll – that’s for a later post once we have enough visitor volume.

      Thanks!
      Kevin

  • Guest
  • Rich

    You’re inflating an already-inflated job description here. Before there were Ninjas, Rockstars and Growth Hackers, this was known as “Marketing”.

    Tell an engineer that you’re doing “Growth Engineering”. If they keep a straight face, you can keep the title 😛