TL;DR: Here is a summary of how we found beta testers on Twitter
At Front, a collaborative email client for businesses, we are currently in private Beta. In the course of the next four months we want to onboard our first 1,000 users to test and iterate on the product. So we need to find these first 1,000 beta testers: through inbound marketing, word of mouth, beta users websites, social networks etc… You name it, we’ll try it. In this article we share with you how we are leveraging Twitter to find these first interested users.
The tool we use: mention
mention is a web and social media monitoring app that lets you monitor keywords and notifies you whenever a tweet, an article, a Facebook message (and more) containing your keywords is published.
How did we look for beta testers with it?
We created a list of keywords we thought people would use to talk / complain / recommend / ask about email clients on Twitter (ex tweets containing: “recommend email client”). So we could talk to them. And invite them to try Front.
For a couple of days we monitored these keywords and looked at the tweets to see if we could engage with these users or not.
If the tweets were relevant we kept the keywords and invited the users to try Front. Otherwise we dropped it or refined it. (by adding additional words for example)
Rinse and repeat.
The results
1 week to create (and refine) your first set of interesting keywords
average of 15 new subscribers to our beta list per week (qualified people)
Useful “side effects” including: - you find out who are the influencers in your industry
you detect blogs / websites which are talking about your industry (from small to big ones)
you see which articles resonate with your targeted users
efficient competition monitoring
all of these for around 45 min per day
The detailed version: How to find beta testers, step-by-step
In this article we want to share with you our experience using mention, a monitoring tool, to find beta testers. In order to onboard 1000 users on Front we first needed to find people who were interested in our collaborative email client. Twitter was definitely an acquisition channel we wanted to explore. We wanted to know if it was really possible to find potential leads and to engage with them directly on Twitter. And, if so, at which scale.
In order to do so we first needed a social monitoring tool. There are several of them on the market but we decided to use mention for two reasons: first, for its reputation as a great application, and second, because mention is also part of eFounders, the startup studio we also belong to.
mention is a monitoring app which enables you to monitor specific keywords on the web (blogs, forums etc…) and on social networks (Twitter, Facebook) . The concept is pretty straightforward: you create a set of keywords you want to monitor and as soon as mention detects a tweet or an article using it, you are notified through its slick interface.
mention is typically used by brands to monitor their own name (and their competitors). In our case we didn’t want to monitor what people were saying about us. We wanted to monitor social networks in the hope of finding beta testers to onboard. Our initial aims were simple:
find beta testers on Twitter
detect influencers (in our industry) and engage with those beta testers
no prerequisite in terms of scale (we don’t need to onboard 400 people per day at the moment, we just need to put our tool in the hands of real people)
In the rest of this article we’ll share with you our methodology to find beta testers, real examples of what we did to onboard them, and finally, our results (figures on how many people we successfully onboarded). This is basically everything you need to know if you want to find beta testers on Twitter too.
Get started find beta testers: setting up your first keywords
In my opinion this is the hardest part. The hardest part is not to use mention but to monitor the “relevant” keywords. If you manage to define the right set of keywords to track you’ve done 90% of the work. And there is no shortcuts there, it takes time and experiments to find them.
The first thing we did was to list the 5 keywords which appeared to be relevant for us. As we are a collaborative email client for group addresses (support@, hello@, contact@) we first listed “email client”, “company inbox”, “customer support”, “productivity tool” “email software”.
Our main goal was to find people to test our product so we made the assumption that people who were looking for an email client or a customer support tool would use these words in their tweets.
The same day we created an alert on mention for these keywords
Our advice here:
Quickly decide (as a team) which keywords seem to be relevant for your business. A.k.a the keywords you think people would use and for which your product brings a solution.
There is no general rule for these keywords. It really depends on the industry you are in and on the problem your are solving.
Don’t be scared to choose the wrong keywords. You will rapidly see if they are relevant for you or not. And you’ll be able to iterate later.
Don’t start with 15 keywords. Five keywords are really enough at the beginning to find beta testers. It’s better to start slowly and to add keywords later rather than overloading your mention interface right away.
Once you’ve done that, just create your alert and let mention’s magic happen.
You’re not done yet. You’ll probably need to refine your keywords in order to find the most efficient ones. But first let’s talk about how to engage with potential leads who used the keywords you were tracking.
Getting Into Conversations With Potential Beta Testers
To my surprise this part was really not the most complicated as mention offers a great interface to talk to users.
How to select conversations to engage:
There are three main factors you should take into account before starting a conversation with somebody you don’t know on Twitter:
the context
the mood
the person
The context. The context in which the keyword is used is obviously very important. It’s not because you see “email client” in a tweet that you should reply to it.
For example here we wouldn’t bring any value to this user:
The mood. Be aware of the “mood” of the person who is tweeting. On Twitter people like to complain a lot and it’s not necessarily the best time to start a conversation with them.
(Not so sure we want to answer him)
The person. What I really like on mention is that they include the Twitter bio of each user so you can see pretty fast if they fit in your potential users or not. For example a teenager tweeting about her email client is maybe not a good lead for us as we are targeting businesses.
This person tweeted about email client but our tool is tailored for businesses at the moment and not for student yet.
How to jump into conversations you were not part of
Again from a “practical” point of view mention interface makes it extremely easy to join a conversation on Twitter. You just plug your Twitter account and you can then tweet directly from the app.
The real challenge here is to jump intelligently into conversations with people you don’t know. The first thing to keep in my mind is that we are not selling a product. We are looking for beta testers. We didn’t oversell Front since our product is far from being finished.
The second important point is just to be relaxed and cool. The situation is the same as when you arrive in a party and you don’t know anybody. You have groups of people talking to each other in front of you, and when you want to join a discussion, you don’t stop everything. You first listen, and then at a good moment, you jump in and say something related. You don’t start by saying how great you are (even if you believe so).The same goes for Twitter conversations when trying to grab beta testers.
Basically our advice:
Don’t oversell (you don’t need a commercial speech when you are still beta testing) but quickly explain why you are here for (ex: you have a tool, still in beta, that might interest them)
Be natural, relaxed, and not robotic.
Accept that a lot of people will not answer you (and it’s fine, it’s part of the game).
Humor is a good ice breaker.
Don’t hesitate to be opportunistic from time to time. For example one day we saw a lot of people mocking a recruiting agency which did a mistake. They had sent an email to many developers telling them they had a lot of job opportunities. The problem is that they sent the email with everyone in cc instead of bcc… It was a perfect match for us as Front is designed to avoid this kind of mistake. Perfect time to jump in ? (we got a couple of beta testers after this one)
Should you hijack conversations mentioning other competitors?
This is a good question. I don’t like doing that so much, and I think you have more to lose than to win. At Front, we are not focused on competitors. We’re focused on building a great tool for our users.
Refining your set of keywords
Now that you know how to monitor your first keywords and how to jump into conversations, let’s go back to the hardest part: refining your alerts.
The first days after you’ve setup your initial keywords you will quickly see whether each alert is relevant for you or not. There are several reasons why a keyword might need to be refined:
The keyword is too general
When a keyword is too general, you end up having several hundreds of new tweets on your mention interface every morning. It’s too much to handle, and 80 percent of the time they are not relevant.
It was the case for our first keyword “customer support”. Our assumption was that people tweeting these words would do so to complain or to greet customer service teams. Maybe we could then start a conversation with the brand/company providing this customer support. The reality is that this keyword is too general, and we couldn’t exploit it just “out of the box”.
The solution in this case is to refine your keyword by adding more constraints. With mention you can add chains of keywords “customer support” “software” or “customer support” “solution” for examples. You can also exclude specific terms (great feature).
The keyword is irrelevant
Sometimes you think a keyword is relevant for you but in reality it makes absolutely no sense at all.
That was the case for “productivity tool”. To be fair we had doubts about the relevance of this keyword from the beginning but nevertheless we wanted to test it. Even after refining it we couldn’t really get something out of it so we just dropped it.
The keyword has another meaning which is more used
This is the story behind a keyword we thought would be fantastic for us: “bcc fail”. As I’ve explained earlier we detected a conversation in which people complained because a company added all recipient of an email in “cc” instead of “bcc”. It was perfect for us as Front makes it easy to avoid such mistakes. So after we saw that we happily created the “fail bcc” alert in order to communicate with the poor companies which did the mistake.
Oh boy, what naive French native people like us didn’t know is that apparently “bcc” also means something with school or college (I tried to look here but too many possibilities). So we had hundreds of teenagers speaking about ‘bcc #fail with no relation to email whatsoever.
In this case the solution is to add more constraints to your alert (extra keywords for example).
Our advice:
Just try your first alerts for a day or two and then analyse the results
Either refine them or drop them
A good strategy to find new keywords is to read the conversations containing your initial keywords and to look for more specific terms in it (it’s how we got the idea for “bcc fail” or for “inbox 0”)
Another good strategy is to monitor your competitors
Add words like “recommend”, “alternative”… to refine your keywords. Ex: “recommend email client”, “alternative email client”
Spend the first week refining your alerts by adding or removing keywords. Experimentation is crucial the first few days.
Long tail keywords versus hit keywords
This is more advice for organizations than a strategy to find relevant keywords, but we wanted to share it with you.
There are two kinds of keywords: the “hit” keywords and the “long tail” keywords
The “hit” keywords are keywords that bring you a lot of tweets but for which you have to look carefully because the majority of them are not relevant for you. A good example for us is “email client”. This keyword is positive for us as it brings us more than 100 mentions per day, but less than 20 percent of them are relevant to find our beta testers.
The “long tail” keywords are keywords which appear on Twitter only a couple of times per week, but which are highly targeted. Exemple: “recommend email client”.
Our advice is to separate in mention these two kinds of keywords in different alerts. Otherwise your long tail keywords will be lost in your “hit” ones. Also, like for an investment portfolio, find a good mix of “hit” and “long tail” alerts.
"Like for an investment portfolio, find a good mix of “hit” and “long tail” alerts."
Figures and results
What do you get from all this?
It takes 1 week to create (and refine) your first set of interesting keywords
Average of 15 new subscribers to our beta list per week (qualified people)
Average of 150 – 200 mentions per day
Useful “side effects” including: - you find out who are the influencers in your industry
You detect blogs / websites which are talking about your industry (from small to big ones)
You see which articles resonate with your targeted users
Efficient competition monitoring
All of these for 45 min – 1 hour per day
Written by Mathilde Collin
Originally Published: 17 April 2020