8 min read
Next in trending

Growth hacking: A Black Art?

People tend to underestimate just how hard it can be to acquire new customers.

Growth hacking: A Black Art?


People tend to underestimate just how hard it can be to acquire new customers. When you check online articles for user acquisition tips (or ‘growth hacks’), you often end up with the examples of site such as Airbnb, Facebook, and Dropbox. Let’s dig deeper and see what works and what doesn’t work for a signup page based on a case study…

During the past few months, I have been working on a tool for PhD students as one of my side projects. Before I started, I looked at what worked for companies in the same scene. One of them grew quite fast by aggressively mailing people when a colleague signed up (by using Facebook login, they had an overview of the connections, however users were not always aware that their colleagues would receive mail from them). This worked out for them and they now have a few million users.

Another company grew by having an ambassador program and letting people give presentations about the topic (such as the below comment on Quora indicates). They also commented on every single (blog)post made about them, were present at conferences, and had a good story to tell about open science. It is clear that this strategy takes quite some time.

To create my tool, I focused on two things: twitter followers (currently around 3,120) and signups for the mailing list (>1,500 in two months and increasing). This may not sound a lot, but academics are very busy and hate spam so they are reluctant to provide their mail addresses — some even mailed me saying they wanted to wait until the product was finished before signing up. Furthermore, I did not want to use any aggressive tactics that could have resulted in more signups. The ultimate goal was to find what sources could drive traffic and what sources could not.

You should always start with an assumption about what could drive growth. Note that you should measure what works and what does not. The easiest way is to set up different mailing lists in MailChimp, which also gives you a dashboard overview of your activities.

  • What Worked:

Landing Page:

Approach: Created a basic landing page with information about the product. I later posted it on betali.st (you have many alternatives, but betali.st is my personal pick), and on a few research related websites in order to ask for feedback.

Result: More than 400 users signed up

Price: 60 USD (design); 3 days of work

Lessons Learned: Driving traffic to a landing page is a lot of work. You need to find out what sites users frequent.

Tool:

Approach: By following conversations on Facebook groups, I learned that a lot of PhD students find it hard to complete their tasks. Thus, I created a tool that lets you state your goal at the beginning of the day, and at the end, you will receive an email asking if you were able to finish the task. If not, you will receive the same email the next day in order to create accountability.

Result: More than 600 people used it; posted on Reddit, Hacker News, etc.

Price: 50 USD (code, hosting, and domain name); 2 days of work

Lessons Learned: People seemed to like this tool. However, I should have added an incentive to share the link. Do not forget to add a disclosure that they signed up for your mailing list. Be sure to never spam people.

Quiz:

Approach: Used typeform.com to create a small quiz titled ‘what type of researcher are you?’. The result was one of five types that were delivered to your mailbox. This was posted on several Facebook groups. Since PhD students like to laugh, comics and cartoons are very popular (see PhD comics), so I added some cartoons to the quiz.

Result: More than 500 people participated

Price: 50 USD (cartoons), 20 USD for Facebook app; 2 days of work

Lessons Learned: Of the 50+ reactions I got, only 2 were negative. Some users did not want to share their email address at the end. In general, people seemed to like this. Do not forget the disclosure! During a second phase, I tried to integrate this into a Facebook app but, but received much lower results (maybe the permissions scared people away).

The ‘Gadget Freak’ scientist — one of the quiz results

Twitter:

Approach: Help people, interact with them (make sure to use the hashtags they use!), and contact them when they say negative things about the competition.

Result: >100 views on signup page

Price: Free, but takes a relatively large amount of time and must be a daily activity

Lessons Learned: Reaching out to potential customers this way requires a lot of time and effort.

  • What Did Not Work:

Mini Cards from Moo.com:

Approach: I created cards and paid someone to place them at strategic locations during academic conferences.

Result: Did not work at all

Price: 50 USD; 1 day of work

Lessons Learned: In retrospect, I believe the design was unappealing and contributed to failure. Moreover, scientists are very busy at conferences so this approach may have been a bad idea in general.

Content (1):

Approach: Write an e-book, post different places (e.g., Slideshare)

Result: Some positive feedback, >50 signups

Price: 0 USD, reused the cartoons from the quiz; 3 days of work

Lessons Learned: If there is a lot of content, and therefore competition (such as the large amount of PhD student), it may be hard to differentiate yourself.

Content (2):

Approach: Create a blog. After a while, I began contacting people to ask if I could reblog their posts since there was a lot of available content. This lead to less work and exposure on twitter (i.e., retweeted links).

Result: >50 signups

Price: 0 USD, 4 days of work

Lessons Learned: Content drives traffic, but can take quite some effort and time to generate. Other approaches work faster.

StumbleUpon.com Ads:

Approach: Purchased some StumbleUpon ads (cheap!)

Result: Did not work at all

Price: 10 USD; 1 hour of work

Lessons Learned: Ads and landing pages are not always a good fit — your landing page needs to be very clear if you choose to try this. I do not like to use ads for landing pages since I have to pay for every click and they might not like testing a beta version of a product. It can be used to test your value proposition, though. I bought these ads after several PhD students told me they used StumbleUpon actively, and I had some money left on my account.

Flyers:

Approach: Payed a user of fiverr.com to post flyers at large Universities, close to the rooms where PhD students work.

Result: <5 signups

Price: 5 USD; 10 min of work

Lessons Learned: Does not work, but was worth a try.

YouTube Video:

Approach: Created a video based on a professor imitator (Title: sh*t supervisors say).

Result: >5,000 views, but very few signups

Price: 15 USD; 30 min of work

Lessons learned: Keep the focus on your product. If people only know you from an unrelated video, they might not be very interested in your product after all.

Infographic:

Approach: Create an infographic based on relevant data (using infogr.am).

Result: A few sign ups, but limited (<10)

Price: 0 USD; 1 day of work

Lessons Learned: Too much work for the end result. PhD students care more about actionable content (e.g., learning how to write an article).

Meet-ups:

Approach: Post on meetup.com.

Result: Some people signed up, but less than 10 per meet-up.

Price: 5 USD/month; 1 hour of work

Lessons Learned: Takes quite some effort to set up and grow the meet-up.

Twitter bot:

Approach: Retweet open PhD positions.

Result: <50 sign ups, tried two bot services. Both of them broke.

Price: 0 USD; 2 hours of work

Lessons Learned: Try to add value for your (would be) users. Twitter bots tend to break (did not create one myself).

Other Techniques:

I tried a lot of other things that failed miserably with <50 sign ups (e.g., contacting mailing lists and podcasts, giving away free cups of coffee if they took a picture and tweeted with a hashtag, google ads based on conference names, translating the landing page into Spanish, automatic posts to Google+, using badges in signup mail to incentivize tweets). One must simply move on and try other things. Once the product is finished, it will be possible to try other techniques and measure actual usage.

Conclusion:

Looking back, I should have focused more on the things that worked by sharing the link on more websites. More iterations (for example, improving the design) could have made the unsuccessful approaches into successful ones.

As you can see, it can take quite some time and effort to find beta users. A small percentage will sign up, and an even smaller number will ever use your product. If I had used only successful approaches, I would have paid +- $0.10 for each beta user (I have now paid +- $0.15 for each). If they provide valuable feedback, this may be worth it.

Experiment and see what works for your target audience. Document your progress and, no matter what, keep trying!

A visual summary can be found here


You can follow me on twitter @seysconstantijn or contact me at seysconstantijn[at]gmail.com

Resources:

Neil Patel

Noah Kagan

Andrew Chen

Growth Hacker TV

Trak.io

Customer dev labs

The Sparkline