How I growth hacked my bot for 10x better user acquisition

Written by nofar.albarak | Published 2018/06/14
Tech Story Tags: growth-hacking | chatbots | bots | growth | messaging

TLDRvia the TL;DR App

What is one of the greatest myths startuppers tend to believe?“If you have a great product, it will sell itself!”

Right?! Well, not quite…If you’ll try to execute this belief as a marketing strategy, you will probably find yourself headed for a failure. And that’s not me saying that, but Peter Thiel, the author of “Zero to One” and one of PayPal co-founders, and part of the so-called “PayPal mafia”.

“No matter how strong your product- even if it easily fits into already established habits and anybody who tries it likes it immediately- you must still support it with a strong distribution plan” — Peter Thiel

So, what is a B2C startup biggest challenge you ask? It’s hacking your channel distributions. And you need to start thinking about it from day one!

In this article, I want to share with you what we, as “Me” did, to drive hyper-growth to our product. How we managed to bring our chatbot to be featured in Facebook Messenger’s Discover Tab, but more importantly — what we did to keep it there, and reach our first One Million users.

So, meet Me, our Ultimate Reminder Assistant, developed by Tiv.ai to balance our daily mental overload and help families become more organized, productive, and connected than ever before.

Discovery

Discovery is one of the main challenges for the evolution of bot technologies. The rapid proliferation of messaging bots and the increasing number of messaging platforms is translating into a discovery nightmare for the ecosystem. According to Expanded Rambling, there are over 300K bots just on Facebook Messenger right to Jan 2018.

Back to “Me” — At the beginning of our journey, we too were captivated by the myth that our great product will sell itself! And it did! All it took was a simple weekly approach and patience (you can read more about it here). However, our growth was still low, and even though people mentioned “Me” in their blogs, shared it with friends and we even got featured on BotList, we didn’t experience the exponential growth we were hoping for.

Screenshots of User’s reviews from Me’s Facebook page

User acquisition

Let’s skip to the bottom line for a second — Me is serving over one million users around the world. How did we do that? As a public service, and to save you some time (and money), here are some of the techniques we applied to acquire high and sustainable traction.

The first thing we tried was organic marketing. We believed that if we manage to hack this, we will enjoy good marketing. After we released some blog posts, we found that having a peak of dozens of new users once a week isn’t good for us. Since we were releasing versions on a daily basis and checking things all the time we needed a significant amount of new users to come every day and try our product.

Later we thought about trying Paid advertising, but when you choose this channel, you need to have a sustainable business model, so the Customer Lifetime Value (CLTV) should be three times than the Customer Acquisition Cost (CAC). Otherwise, you need to stop right now, or you might be doomed to failure. Since we developed a chatbot that sits on top of Facebook Messenger, using a paid acquisition strategies wasn’t the right way to go from one main reason:

“The market isn’t ready to pay for bots’ micro-services yet, so paying to acquire new users without monetizing, eventually will get you bankrupt.”

Our second try was to build an Inherently Viral Mechanism into our Product, so with every core-action, an existing user takes, new users found out about your product. When your product has a natural social competent, an internal growth could form. Let’s take NearGroup as an example.

NearGroup is a dating bot, which matches users (based on their preference) & allows them to start chatting without viewing images or profiles, pretty cool right?! Well, this chatbot is adding 800K new users every month.

Our core-action was setting a reminder, 10% of the total reminders that a user creates, were for other people. For example: ‘remind John to pick up the kids tomorrow’s afternoon.’ That sounds great right, but when looking on the average reminders created by users and crossing it with the number of people users interact within the bot, you found out that salvation will not come from here.

Last hope was the Discover Tab. Discover is where people can browse and find Messenger bots. Getting featured on any app or bot store is probably THE best thing that could happen to any product. That’s a free channel to acquire users but more importantly — getting their trust in your product.

“But to get featured not only should your product deliver a unique value proposition, but it should be better than 99% of the market.”

Old school, new neighborhood.

A quick look at the Discover Tab will show up to 20 bots in each list, 80% out of them represent enterprises or big tech companies. These companies, are also Facebook’s business partners, which means they are not going anywhere in the near future, and leaves us with only four places for bot developers.

That’s a challenge! But we’re not the ones who are afraid of a good challenge, and we decided to embrace it, and to do what it takes to get on the list.

For more than 2 months, I ate and drank the “Discover Tab” — studying how to reach it, exploring what KPIs we need to achieve, for how long bots stay in the list, and more importantly — what’s the expected outcome.

Lucky for you, here are some of important the things I have learned during the process:

  1. Getting featured is the ‘easy’ part, staying there is the real battle:Every week your bot will be tested against more than 300K bots on Facebook Messenger, you have to keep up your performances and keep getting better all the time.

  2. Being ranked at the top 3 will attract 5K new daily users; the top 10 will attract daily 1–3K, depends on the place, and at the top 20 will draw a couple of hundreds. (Of course the higher the penetration rate, the more users will encounter your bot).

During our exploration time, we tried to get our hands on numbers such as: good retention and stickiness rates, the recommended number and lengths of sessions, does it make any difference who start the conversation the user or the bot? Does the page or the bot reviews affect the rank, how does Blocked, Deleted or Marked as Spam conversations make any difference? And so on.

Facebook, as you can imagine, won’t provides the answers. But we had some luck. Magically, we found that “Me” is featuring under the Productivity Category (if you scroll down the Discover Tab, you will see more lists based on the topic). We use this as an advantage and figure out on which KPIs we should focus.

Discover Tab, Categories

Conquer the Productivity Category

Our first objective was looking for the productivity apps’ benchmarks. We released daily versions, tuned the product with incremental advance, until we improved on the competition.

Sources: SimiliarWeb, Adjust.com & Appsflyer

Since the list got updated on a weekly basis (Friday to be specific), we focused on one KPI every week and tested if our rank is improved. In other words — if that particular KPI influences the Discover Tab. After a while we found out that the relevant KPIs we need to focus on are the following:

  • Retention Day 2, Day 7 and Day 21
  • Stickiness (DaU/MaU)
  • Session length

There aren’t golden numbers for each KPI since every week you are measured against all the bots out there, based on the concept of “may the best Bot win”, and by “best” I mean the best performing.

Once “Me” dominated the productivity niche, we decided it’s the right time to scale up and look at the bigger goal, the Featured list on the Discover Tab.

Moving on to the Multinational League

After we discovered what we need to focus on, and which KPI makes the difference, it was easier to scale up. When you open Facebook Messenger and click the Discover Tab, the first thing you will see is the top three bots worldwide, this is the Holy Grail. To get there, you will not only need to get into the Featured list, but also to be at the top three.

That was our next mission — we didn’t know what good metrics are, so, we turned to the amazing bot community, and thanks to them, we managed to do this in just a couple of days. I was asked not to expose our findings, but I can share with you the way there:

  1. Read every blog post about successful bots and gather stats
  2. I reached bot developers and asked them to share some info.
  3. Paid for Analytics platforms that suggested ‘Competitive Intelligent’ features. These features can help you understand how good your product is relative to other products.

After two weeks, I got the sense of what we need to achieve and started to work on these metrics until we accomplished the KPIs.

One important thing, when we first improved our KPIs, we were already featured in the Productivity Category, but the second time was different. We couldn’t just ask to be featured. First, because we needed to go through the whole process by Facebook, and secondly, we didn’t want to waste our one chance of getting in. From my research, ‘regular’ bots didn’t hold more than two weeks on the Featured Tab, and we didn’t want to miss that.

Submitting and Getting Featured

To be featured, you need that a Facebook Partner will submit your bot for a review. You probably think what Facebook Partners are, and how do I reach them? Well, as an entrepreneur one of the most important lessons I learned is to get my foot in the door and say what I have to say.

After I send messages to every person that could potentially help me, a Facebook employee heard about us and introduced me to a Facebook Partner. We quickly send her an email with a background, explained why Me is essential to people and how it can affect their life positively, and some of our KPIs.

Let’s break down the process:

  1. Submit your bot for review in specific countries — They review all potential bots for Platform Featured Adoption, Policy compliance, User Experience, and use case. I’d highly recommend going over the Messenger Bots Best Practices before you submit your bot. This step usually takes one month, and you want to get the maximum out of it.

2. Once the review completed, you’ll get you a slide deck with feedback that you can work on, if relevant.

  1. Policy review — basically they are making sure that your bot complies with the messaging policies for the Messenger Platform. This step usually takes one week.

On February 16th, “Me” got featured in the US and the Philippines and ranked at the 22nd place in his first week.

Remember the fact that every week Facebook revise the list? The week after, we reached the 8th place, and in the 3rd week we conquer the top 3!!!

Since our launch, “Me” has processed over 200 million messages, assist over 1 million users worldwide, and has become the #1 Most popular productivity bot on Facebook’s Messenger in the U.S, Canada, and the Philippines.

User Activity from Jan 01 to May 01

I hope you enjoyed reading this article.

Please feel free to ask any questions in the Responses section below. Also, feel free to check out our bot Me.

👏👏Don’t forget to give a round of applause 👏👏

If you’ve enjoyed this piece, go ahead, give it a clap 👏🏻 (you can clap more than once)! You can also share it somewhere online so others can read it too.


Published by HackerNoon on 2018/06/14