How We Got 1,000 Beta Signups in 2 Weeks : The Blueprint

This is the story of how HappyFox Chat, got our first 1,000 beta signups in two weeks. How we did it, how we failed, what we learned along the way and how you can make use of it for your beta launch.

The idea behind HappyFox Chat is to democratize live chat.

On March 3rd, we represented at the Launch festival, where we had the opportunity to speak to startups and get one on one feedback with founders, face to face. Our biggest validation came from the conversations we had at the startup demo pit.

For example, we gave a sneak peek of HappyFox Chat’s Shopify integration feature to the @Shopify team at the demo pit. They were very excited about it and suggested a feature that we hadn’t thought of initially — the Shopping cart tracker integration. We went back to the team and got it implemented the very next day. This was one example of how we could make the product better from actual user feedback.

Via @PulkitSankhla

Most startups with no dedicated support/sales agent used shared logins within the team, so that they can remain on the products free tier. Some did use multi-agent paid licenses but still not being available to their visitors as much they wanted to. We decided to break that model and make live chat available to everyone along with offering a superior customer support experience.

“Wow!” was the common response we got from those who got a sneak peek.

One even said “This is like the Tesla of live chats”. ☺

Following is the story of how we achieved the first 1K signups.

Our first 1,000 Signups: How we did it

10,000 Beta signups — That is our target. Ambitious? Perhaps. But goals always should be. Don’t you think? As they say ‘Go big or go home!’

We believe that communication between customers and businesses shouldn’t be a privilege of just a few or limited by any means.

What we had in hand at the time of private beta launch to make this ambitious target possible, was a good product offering a great experience that we had validation for.

Focus on smaller achievable goals first. Learn. Unlearn. Repeat.

We knew 10K was a long shot. That 1K was the first milestone.

First things first. Who are we talking to?

We had a clear idea of who our target audience was from day one — there were three clear segments we needed to engage with — companies passionate about customer support, retail companies running e-com stores and lastly folks who were passionate about marketing.

We didn’t look at them as who we could target but rather asked the question, “Who all could we think of who would benefit from the product where it would serve to amplify whatever they were doing?”, “Whose eco-systems could we add value to?” and the answers helped give us the clarity on the who we should talk to.

We used Ash Maurya’s Product Market fit principles, mentioned here, to get this done. Go check out his amazing presentation on YouTube or subscribe to his blog here — it’s invaluable.

The Goal & The Objective

We wanted to get the word out to the early adopter crowd and get the initial traction before the public launch. If 10K signups was our goal, this initial 1K, was our near objective.

The Product

Before joining HappyFox, I when I met Shalin, he told me that he had a vision for reimagining live chat and the first question that popped in my mind was… “Why?”

Me: Yet another live chat product, huh?

Shalin: Um, wrong question!

Do we need to re-conceptualize live chat in 2015?

Hell Yes!

And then he gave me a demo of HappyFox Chat and I knew what he meant.

The whole experience of HappyFox Chat got me excited and something in me, told me that this is bigger than what I was seeing.

But here are the three main aspects of HappyFox Chat focuses on:

  1. Agent Experience:
    This is not about just the chat app’s interface but how does it interface with other tools that carry your customer data. We bring context and interactivity to deliver great experience to customers by focusing on hard on agents experience and workflows.
  2. Ultra-Fast Chats:
    Every second has an impact on chat continuity and abandonment rate. HappyFox Chat is not built on third-party messaging APIs or engineered with technology stack that is not designed for real-time communication. We have engineered chat ground up to deliver every message in fraction of a second.
  3. Democratize Live Chat:
    Don’t limit Live Chat tools to live chat agents. Well, at HappyFox there are tons of people who participate in support & sales and are not support or sales rep. Our pricing model helps bring all of your team members onboard and get closer to customer communication and feedback.

The Strategy to getting first 1K Signups

Our strategy was to introduce the product to a slice of our target audience, in the shortest possible time, efficiently. Entice them with a sneak peek of the product and get them on board. Leverage their immediate network, to get more referrals.

We realized however, that blindly reaching out to everyone wasn’t a good idea, it wasn’t efficient. Starting with low hanging fruits, made sense.

And there were two segments, that we could reach out to first.

a) People who had experience with other live chat solutions (The assumption here being — They would understand the value in our product easily by comparison) and..

b) Early adopters who would want to try out a brand new, promising product.

How Twitter Outreach Helped

In our research we found that, a lot of people using existing live chat solutions had problems with those solutions and were seeking help on Twitter. Obviously, they were our first and easy target. So, our outreach was mostly on Twitter.

We created private Twitter lists of people who were following/engaging with competition and tweeting about certain interest topics (like hashtags). And we engaged with them whenever possible, like when they were having an issue with their live chat software or were seeking help with customer support. Or just, being friends with them by engaging casually.

Re Tweets, favorites, sometimes just engaging into real meaningful off-topic conversations, helped us get their attention.

To our favor, there were several issues and limitations with the leading live chat software, and people were tweeting about them most of the times.

All we could see was opportunity.

And whenever we reached out to people who were asking for such solutions (like in the screenshot above), people would gladly switch, because it was so easy to.

Here’s what Twitter analytics looked like for the last 28 days.

Those are some pretty neat numbers to achieve, and all the more so for a brand starting off from scratch.

Learnings from Twitter

  • Imagine Twitter is like a party and you’re just there mingling with people. You don’t want to go ahead and hand over flyers to people at a party. People can smell that, from a mile away. Be cool.
  • Try to add value to conversations otherwise people find it very annoying that you’re trying to sell them something.
  • Don’t focus too much on the numbers. Focus on the intangible benefits of being there, joining the conversation with others.
  • Answer promptly. We made lot of good friends by just replying soon to tweets.
  • Not all profiles views end up becoming signups and that’s okay too.
  • There are so many metrics other than profile views and follower count that matter, but unfortunately, many of these cannot be clearly monitored or defined in most cases just yet.

Tools that helped with Twitter engagement

Buffer : For their amazing scheduling options, content suggestions and analytics.
Tip: Check out their Pablo tool, it’s crazy amazing. (Lets you create images from text with pre-built templates etc.) We used this extensively, and images performed a lot better than text in engagement.

Tips for using Buffer for maximum Engagement

  • Get on their paid plan. It’s definitely worth the value.
  • Use all the available time slots. We had 100 content pieces going out each day, at one point.
  • Their time scheduling can be a bit messy, so make sure you don’t schedule tweets too close, or it could be annoying to your audience. (We made that mistake once, but fixed it soon enough)
  • Check out their “Content Suggestions” and don’t forget to check out various topics that will fit your audience. Whenever we ran dry of content, we used it, and worked like a charm.
  • When you tweet content from major publications, make sure you quote the author.

Typically Buffer does not let you do it, but we manually added the author with a “via @author” option and this helped in getting their attention on Twitter. Most of these people have a huge following and perhaps you could get a RT or favorite from them, and it could make your day!

Reaching Out To Early Adopters

Our second target was early adopters from the technology scene. Sounds cliche, right? Everybody out there on social media is an early adopter in one way or another. Well, we wanted to reach out and introduce the product to folks who had already tried a similar product before, and wouldn’t mind trying someone new.

We could think of several options including BetaList, Product Hunters, Hacker News, Reddit etc. We decided to take one step at a time, as we realized that there were several nuances in each community that made them unique. And a blanket strategy for all of them would literally mean spamming(Yikes! We wouldn’t want to do that, not even unintentionally!)

We decided to focus on BetaList first, as it was the apt community for a private beta product like ours. (Product Hunt wouldn’t accept private betas, so we kept it for later.)

BetaList is a vibrant community where the audience is hungry to try new stuff everyday. Getting on BetaList however, is close to impossible if you go the organic way — like Product Hunt. We realized that their wait periods can go indefinitely long after a submission. So we decided to go the paid route and spent $199 on a one day approval process.

Folks at Betalist were really good and fun to work with. They were very responsive on emails, and managed to get us on BetaList at short notice.

In one day of submission, after being on their homepage for 2 days straight, in the “Trending Startups” section and having been included in their email blast, we got about 616 visits from BetaList (excluding the Twitter traffic from their followers)

Tips For Making The Most of BetaList

  • Choose a Monday or Tuesday (recommended) so you get all the attention you can early on and squeeze the best off it for the rest of the week.
  • Contact their support via Email, well in advance (at least 2 weeks) and discuss your plans.
  • Go check the share statistics of other companies already trending, and get an idea of how many shares you’d need to trend. (I’m assuming you will get buried by top performers, if you don’t meet the threshold.)
  • Keep a watch on ReTweeters from BetaList and connect with them on Twitter — This is additional audience for you.
  • Make sure you submit with the right tags on BetaList, because we made the mistake of adding a wrong tag (probably not significant but still) and it stuck to the posting.
  • Make sure your Title and Description is Twitter optimized so Tweets don’t look awkward with truncated text.

How Referral Program Helped

Inspired by RobinHood, we tried several referral programs out there, but unfortunately, nothing worked the way we wanted.

The idea was to have a referral program in place that would reward users for actions they take and give them clarity on position updates, rewards and referral status at any time.

We wanted a referral program that would create a rewarding experience.

Here’s what our referral program looked like in a nutshell.

  1. Interested people sign up for private beta.
  2. Gets added to a waitlist/queue.
  3. Is allocated a position in the queue based on time of entry.
  4. Is given unique Referral ID.
  5. Is prompted for making referrals and promised early access.
  6. Moves up the queue based on referrals made.
  7. Email updates are sent for position updates/ referrals required.
  8. Gets on list of top referrers if actions met.
  9. Is given access to app.

It’s a little more complicated than that to be honest (We’ll be writing a separate post about it later.) But from a user’s perspective, it is simple.

The more friends you refer, the earlier you get access to the app.

At first we tried KickOff Labs. It could do a lot of things from the above list, but not quite enough. For example, once the user signed up and was “in”, we wanted to send email updates to the user at certain specific points, based on his behaviour.

  1. Welcome email with referral id/url & position.
  2. Whenever user made a referral and moved up the queue.
  3. When user wasn’t making any referrals at all. (At various times..)
  4. After X days, with details on actions to take to get access sooner.

KickOff could show stats on it’s Thank You page and send emails at specific points (when user makes X number of referrals), but their email functionalities were limited. At the time of testing, there was no option available to tell the user what his current position in the queue was or how many more positions he/she needed to move up to get access.

So, we built our own referral engine..

KickOff Labs were pretty supportive in trying to get us what we wanted, but they soon ran out of options.

Now that we were clear on what we wanted, we tried couple more open source programs, including Untorch and UnFlare, both of which were used by other brands for earlier private beta launches. Unfortunately, even they were limited in options and their makers had stopped development and moved on.

So, rather than re-building other tools out there, we built our own little referral engine and backed it up using Customer.io’s behavior targeted email functionalities.

Customer.io offered good segmenting options which we could use to reach out to users based on their behaviors via email.

Email updates were sent whenever the user moved up the queue and specific calls to action were given at crucial points to tell them just how many points away they were from getting access.

We set up two sets of email episodes on Customer.io that would prompt the user with position updates and call to actions at specific behavior triggered points.

But not everything went well with Customer.io.

Via Their status blog.

Just as we got our referral engine working and ready to go, Customer.io went down for more than 24 hours and we had to postpone our Betalist launch because of this.

The Numbers & Performance

Top Referring Sources

Obviously, top referrals were from Betalist, Twitter, HappyFox.com and Facebook.

(There are some discrepancies in how the total numbers were shown in Google Analytics v/s Customer.io.)

Betalist alone gave us 616 visitors and 217 signups in total with a conversion ratio of ~35%!

The most traffic was on the day of submission itself (April 28th) with up to 840 shares on Twitter and 45 likes on Facebook.

Here’s the original submission — http://betalist.com/startups/happyfox-chat

Betalist traffic was more like a social media stumble, where you get a lot of traffic in the first spike and then it dwindles on.

Twitter Traffic

Twitter on the other hand is still giving us that steady but unpredictable traffic. The first 840 shares via Betalist was a good push, as can be seen from the spike.

But, interestingly enough there was equally strong traffic from Twitter even later (ex: May 10th). These came from influential people sharing their referral URL via Twitter. (Thanks to e-mail episodes, which we will be talking about later in this article.)

Most traffic from Twitter came via Social Referrals

All these referral URLs are unique urls generated by our referral program.

Referral Traffic from HappyFox

Since, we put the HappyFox chat widget on our primary website happyfox.com, we got a considerable amount of traffic from there as well. This was more qualified traffic compared to Twitter or Betalist and got up to 3.28% in conversion ratio.

Referral Traffic from Facebook

Interestingly, Facebook also referred a decent amount of traffic and subsequent signups. The most significant being one on May 10th by Vijay Anand, an influential mentor in the startup ecosystem. His one share attributed to 103 in referral traffic and subsequent signups. Rest of the Facebook shares were from people sharing their unique referral URLs.

Best converting traffic

If you look at the number of referrals, Betalist is the top performer with 616 visits. But if you look at quality of traffic, HappyFox.com, Facebook and Twitter were the biggest contributors.

Where did users come from?

Geo split of traffic.

From everywhere, mostly USA and India. Check out the map.

Geo split of Signups

Most conversions happened from India with 386 unique signups, followed by US (223 signups) and UK.

How Email Campaigns Helped

We ran email campaigns inviting existing HappyFox customers to try HappyFox and there were about 160 people who signed up from various geos. This was a no-brainer as most of our customers recognized the brand and in fact, had requested such a live chat tool earlier on, so this was not surprising.

User Behavior on the site

Engagement : Average Time Spent on Website

The average time spent on the site by a user was less than 2 minutes — which is a pretty decent number, considering this is a private beta site.

Bounce rate was up to 50% which is very high. But considering, this was a landing page with private beta invites, this is understandable. We hope to bring it down with new updates.

How Email Campaigns Helped

So, soon after someone signed up for the beta, they would go into a queue system, wherein they get a chance to move up the front of queue with the help of referrals. The more friends you refer, the higher you move up the queue, increasing your chances to get access to the private beta.

We also set up SOS style email episodes that would go out to people in the queue, reminding them of their positions and how much they could move ahead of the queue.

Here is an example of what their performance looked like. Below is the Welcome Email’s performance for the last 30 days.

Interestingly, the open rates were as high as 68% and click rates 4.8%, which were both amazing stats, way above average.
  • Referral percentage : How many people referred at least one person.
  • Average number of shares/referrals made by a person.
  • How many people signed up via referrals in emails.

What we failed at

We met with couple of failures along the way as well. Goes without saying.

• Reddit
I’m a long time Redditor and I know how things work there. Guess I knew too much. I was very apprehensive of posting something on Reddit as I knew, catering to the Redditor is tricky and unless you can provide solid value to the community, whatever you post is going to bomb. Nevertheless, we posted a feedback thread on Reddit, r/Startups. We got 44 hits, but no signups!
And rightly so, the community gave some nice feedback on the site, but at the back of my mind, I was hoping some of them would sign up. But no.

• Quora
We realized that there were several threads on Quora where there was scope for product specific engagements. Although we found ways to spot them, we couldn’t figure out scalable ways to address them and make use of it. However, we plan to spend some quality time there.

The Road to 10K

So, now that we’ve made it to 1K in two weeks, what’s next?

We have couple of things lined up. Like I said earlier, rather than doing everything at one go and rushing through things, focusing on one channel and excelling at it, looks more efficient and so, we’re taking baby steps.

Community Building

One of the early learnings from our interactions on Twitter is that, if you reach out to the right people with the right message, everything falls into place. And if you are human, making sense, rather than being all “authoritative” and “salesy”, people tend to like you. We’d like to take this learning and scale it up by building a community around us, much like the awesome folks at @SEMrush and @Buffer. We’ve done some early test and have validation to this hypothesis, so this one’s definitely happening.

Content Marketing

How could we not do this? In my previous roles, content marketing is what has worked for scale. Personally, I think it’s a little over rated and saturated now, with almost everyone having a blog and throwing 10 Tips for this and that. It still works for sure, but I’d want to do something differently, this time. Perhaps, at a different scale. If you are a SaaS company, wanting to do amazing things with content marketing, please ping me.

SEO

For sure. Now that Mobilegeddon is gone past, this probably is a good time to invest in some solid SEO. It’s a long fight, but hey, this is up next.

Ads & ReTargeting

We also want to test out some advertising opportunities, PPC and Social Ads to be precise. Our natural inclination is to go for inbound efforts, and spend less money on paid marketing, but we’ll be adding a few channels in there as well.

Streamlining Sales & Marketing

Ideally, what we’re trying to do with the whole Marketing process is to streamline things. There are various touch points to finding customers, and each of them are different in their own ways. Inbound marketing has it’s own way of segmenting processes, and sales has it’s own way to processing leads.

Our goal is to simplify things and streamline processes, so that the buttons are few to turn while the signal is amplified.

Considering there are several tools out there that do this effortlessly, I’m in the pursuit to find the right tool that will help HappyFox do this, efficiently.

Summary

Overall, it was a great learning experience getting the first 1,000 signups. The biggest push was via Betalist, social media outreach and engagement with our existing client base, amplified by the referral program which is continually growing.

Our next milestone is 10K. Watch this space for a follow up post.

Also check out http://www.happyfoxchat.com and sign up if you haven’t yet.

TL;DR

  1. We built a kickass live chat software & opened it for private beta signups — http://www.happyfoxchat.com.
  2. Got 1,000+ signups in under 2 weeks.
    What helped — BetaList submission, Custom Referral Program, Social Media Outreach, Twitter Outreach.
  3. What didn’t work — Reddit, Twitter Hashtag Chats
  4. Next stop — 10K Signups

Looking forward to your comments. Thanks.