Lessons from our pre-launch campaign
When we decided to start a pre-launch campaign for Taskable, we had two goals in mind. First, to connect with more of our potential users to interview them and understand their problems. Second, to build up a list of people who were interested in our product and would be our early adopters.
We are still very much in the middle of this pre-launch process. We'll be running this pre-launch campaign for the foreseeable future. We've 'launched' a couple of times: into alpha testing, into closed beta, into a quasi-open beta, back into private beta. Soon we'll launch a proper version 1. Shortly after, we'll launch version 2.
All this to say, this isn't a linear process. At least fit's not for us. We took inspiration from Kat Mañalac's (YC) How to Launch (Again and Again) Don't wait until something is pixel perfect. The sooner you have users actually using your product, the sooner they can tell you what it's missing and how to improve. So, just launch it already. And then launch again.
Below we share the timeline so far in running our pre-launch campaign. It's pretty long, so we start with the key learnings, and you can read on for more context if you are interested.
TL;DR Key Takeaways from our pre-launch campaign
- Ask for a user interview after someone has made a small investment in you already, such as taking a survey.
- Betalist is a good way to get subscribers - we would recommend posting, but not necessarily pay for the expedited launch day. Product Hunt's Ship/Upcoming page is meh. Most subscribers are spam/not that interested in you (sorry).
- Most people will forget who you are or why they signed up for your launch. Make sure you remind them when you invite them to the demo, especially if they pre-registered weeks or months ago.
- It's great to have collect email addresses and have pre-registered users, but don't expect all of them to take you up on signing up. Often they've moved on, found another product, or weren't that interested in the first place (again.. sorry).
- The best way we found to get early adopters was by spending time in communities and Slack groups where our target users lived. If they were at all interested in what we were working on, we'd try and book an onboard directly.
- Be targetted in whom you reach out to join. If you are doing a closed beta, focus on people whom you think need your product. Otherwise, you'll spend lots of time talking to, or trying to please, people who don't need your product. Respect their time and yours.
- Never stop launching - you'll get many cracks at it. Try stuff out early on, create buzz, get feedback, recalibrate, launch again.
Our pre-launch timeline
Or at least, how I remember it. We've tried to include tools we used, and even Zaps and processes. If you are interested in any of these workflows, we are happy to walk you through them in greater detail. Hit me up.
Tom and I began kicking around this problem we were both struggling with at the end of last year. After leaving our previous startup, we got together often over a beer to talk about work, ideas we had, etc. We had even launched and tested another idea we had under the Taskable brand previously (it wasn't a good idea as it turns out).
The latest problem we were kicking around centered around the disaggregation of information at work. We were both working with multiple clients as freelancers or with an agency, and that meant having various sources of tasks and communications. Searching around these tools killed our productivity and caused anxiety around planning and prioritizing our day.
The first indication we had that this was an idea worth exploring is that we both felt it very acutely. With that in mind, we set about connecting with other people, like us, who were potentially having the same problem. We took a Mom Test approach, trying to understand the problem. We started by talking with friends, and we found we were not alone with this problem.
Following speaking with a small group of friends, we wanted to expand out of our circle a bit. We did this by posting in various online communities we were part of asking for 15 minutes of people's time.
This approach fell flat. We got a few interviews out of it, but a lackluster response to our interview requests.
Since we weren't getting much traction straight up asking for an interview, we tried another tactic: creating a survey. We asked a few questions about people's productivity habits, and what tools they used at work, and what they thought about them. Then at the end, we asked for an email address so we could follow up with any further questions. Eight percent of people dropped off here. But of the 63 people who completed the survey, 48 shared their email address with us. The survey itself was helpful. More importantly, it was a great way to get people to make a slight investment in what we were working on. When we followed up with an interview request, they were likely to say yes.
From these initial interviews, we got a ton of useful information. Enough to keep working on it.
We also posted our idea to Betalist. We also signed up for Product Hunt's Ship on the Pro plan, which meant we were promoted on their upcoming page. When new Product Hunt subscribers would come in, we shared the survey with them and followed up for the interview as well. More on Ship later.
With our new data/validation in the problem space, we scoped out our MVP and got to work on it two days before Christmas 2019.
While Tom started coding, I built a landing page in Webflow, with a call to action to pre-register so we could collect emails of interested users.
I connected the Webflow form via Zapier to Airtable, which we used to keep track of our waitlist. We also sent these email addresses to Mailchimp to send a confirmation email and request that they take our now slightly amended survey.
We also had pre-registrations coming in from Ship. This was sort of annoying because we now had essentially two lists: one discovering us from our website directly, and one coming from Ship. We were on the Pro plan, so we didn't have access to webhooks, so we figured out a workaround by using the Slack integration and Zapier to achieve the same result.
Up until this point, we had several people subscribing to us via Ship, and a few here and there pre-registering through our website. We have a Slack notification that notifies us each time someone pre-registers.
One day just after Christmas the Slack notification started going nuts. We hadn't expected much from Betalist to be honest. I don't think I'd even realized it was happening. But in the end, it turned out to be a huge source of traffic and signups for our fledgling little project. In total Betalist led to 278 website visitors, 68 of whom pre-registered. It's our highest source of referral traffic after Indie Hackers.
We continued interviewing people as they pre-registered and did the survey. Also, we sent somewhat regular status updates on progress with Taskable to create buzz and help keep us top of mind ahead of the product launch.
Launch 1 - The MVishP and Alpha Testing
At the end of January, we were ready to release the MVP. We were pretty excited about our big launch.
Tom and I tested it. Everything seemed to work. Time to release to production and start notifying the legions (ok about 200) of hungry users. We planned to do video onboards with everyone because there was a bit of a setup process that wasn't straightforward to explain in the UI. Plus, we wanted to keep talking to as many users as possible.
We sent a newsletter to all subscribers saying we are live and ready to start onboarding, but we would do it slowly so we could provide excellent service, and get early feedback. We had a few people respond to that email saying they wanted in, so of course, we arranged a call with them right away.
With the rest, we started going down the list, starting with the people who registered first and inviting them to join an onboarding call.
At this stage, we learned a few things quickly, and one thing slowly.
First, the quick things.
Most people forgot who we were and why they registered. The first emails went out without any further info on who we were - just basically your invite to Taskable is here, schedule a video call. A few people got back to us asking basically who the hell we were. Some, kindly, informed us that we should include a reminder as to who we are in these emails, which we quickly added. And some even booked an onboard call. 🎉
The second quick thing we learned is that the subscribers we got from Product Hunt/Ship were of variable quality. A few were genuinely interested in the product, and have since become active users. One has even joined us as an advisor. However, the vast majority were spam, uninterested, kicking the tires, or were just using it as a way to connect with founders to sell them something.
The thing that took us longer to learn in this pre-launch phase was that we should have been more focused on whom we invited to join the beta early on. As I said, we initially just started going down the list in order of the date the prospective user signed up. This, in hindsight, was a mistake. We had a target user in mind, and we had information on many of these subscribers that gave us a good idea of who they were. What we should have done is picked out a few of those target users who fit our criteria and gone to them first. We keep on learning this lesson, as we'll get too soon.
At this point, we've invited probably 30 people, and have done about 5 onboards/user interviews. I won't get too much into the product side of things here. However, we learned a lot from these early onboards and users, which was we still had a bit of work to do to consider this a minimum viable product.
So we held off on invites for a bit and went back to work over the next couple of weeks making some fixes and preparing for our real proper launch.
Launch 2 - Closed Beta and the Real MVP
After continuing to get feedback from a few very early adopters, and of course dogfooding it ourselves, we were ready again to launch into beta in early March.
We again went back to our flawed strategy of going slowly down the list from the date of signup. We were a bit timid at first and did this slowly. I was sending emails manually, one by one down the list with a Calendly invite.
Then we were very fortunate to have a call with Joan, who's the community manager from YC's Startup School program. The end of our conversation went something like this:
Joan: So, what's your goal for next week?
Me and Tom: Well, we want to update Actions, and then do some UI changes on Signals, and blah blah blah
Joan: No, guys, give me a real goal - tell me a number. Something concrete. What is your goal next week?
Me: WE ARE GOING TO DO 10 USER ONBOARDS
Joan: Great, tweet at me when you do
Oops - now I told Joan from YC that we are doing to do ten user onboards over the next seven days when we had only managed to get 5 or 6 done over the prior two weeks. And something about having to tweet at her meant we couldn't fail this mission.
At this point we said fuck it, let's invite the rest of the list. We'd probably reached out to 100 people, leaving 150 people on our list we hadn't contacted to try and arrange an onboard for. So I decided finally to automate this. I set up an onboard invite email campaign in Intercom and uploaded our pre-launch list there. Then I connected up all new subscribers to this flow. No more faffing about (as the British would say). Time to start growing our user numbers. We had a product, let's get as many people using it early on as we can.
We did (just barely) hit the goal of ten onboards. It felt good to have a concrete goal. We kept up that pace over the next couple of weeks, upping the number of onboards from 10 to 12 to 15. However, we had tapped out our pre-registration list by now—time to open the beta up a bit.
Launch 3 - Openish Beta
We began inviting people right after they requested access to book an onboard. We also needed to find new ways to build up the top of the funnel or ask people directly to try us out.
We did this by spending more time in Slack communities and forums. Again, sometimes asking people if they were interested in trying out our beta. But most often just trying to be a good member of the community and sharing insights and content that would be useful. This is where the bulk of our signups and active users have come from as it happens.
All that time cultivating and worrying about our pre-launch registration list: as it turned, fewer people on there were as engaged as we'd hoped. Our strongest engagement came from just talking to people like us online who probably had the same pain points; the original reason we started working on this product. Who would have thought?!
At this stage, we finally realized the lesson about not just inviting people willy nilly. Better to be focused on your target users in this pre-launch phase. We were doing lots of video onboards. Usually, after learning a bit more about the user, I would have a sense if we were a good fit for them. Better to respect their time and ours, and be a bit more focused on who we should be onboarding.
Back to a closed beta/pre-launch mode.
Launch 4 - Close it!
Based on our experience doing the onboards, and seeing who our early active users were, we had a useful framework for whom we should start onboarding. And, whom we should hold off onboarding.
Now when users request access on our landing page, we take them through a Typeform. There we ask a few questions to determine if they meet our target user criteria. If they do, they can book an onboard via Calendly. If not, we tell them to kick rocks (kidding - we send an email saying we don't think we solve their headache and we'll be in touch when we think we do).
The rest we add to our waitlist. We combine their answers from the survey into the Airtable list. This allows us to identify why they weren't a good fit. In some cases, they will be a good fit in the future. For instance, many require a certain integration we don't have yet. So, when we do, we can then quickly go down that list and invite them to join.
This is the phase we find ourselves in now. It means fewer signups, but more potential users that are likely to become active (we hope anyway - still too early to tell). At the same time, we are working on features to respond to early user feedback, as we approach our next launch: Taskable v.1!