How to launch (again and again)

As part of StartUp School, this week we watched Kat's video on How to launch again and again (slides).

TL;DR Our key take-away from this was that we should not be thinking about a single "launch" where fireworks go off and a choir sings. We should be finding customers, signing them up (or getting feedback) and iterating. Every time we try something new we should think of it and treat it as a launch - measuring the outcomes and looking for ways to iterate.

Launching

I've launched a lot of things and each time it's come with a sense of apprehension - what if it's not enough? what if people hate it? what if ...?

And every time the results are almost exactly the same - nobody cares!

Kat's main focus in this talk is that there is no such thing as a single launch; each product has many launches and each launch is an important step in the journey. She stresses that you should be launching all the time, as early as possible. It definitely fits in with YC's general ethos of getting something in the hands of users as quickly as possible. But it's more than that. I think the takeaway here is that what we think of as a "launch" is actually many small things we are mentally grouping together - but the time at which these things occur can be very different.

We need to stop thinking of the single launch and think of how we can be getting something out to users (new features, new messaging, new ideas etc) and gathering data on it. Kat notes early in the video how terrible it is to see founders take ages to launch anything - don't spend 6 months planning for something, you will probably be dead by then!

A common phrase in the startup community these days is "always be shipping" and this is echo-ed by Kat in the talk as "ideally you are never NOT launching". It makes perfect sense in the context of always talking to customers and getting their feedback.

To help us understand, Kat breaks down what founders commonly think of as "launching" into these types:

Silent launch

Just whip up a landing page - we are very firm believers in this and it's a huge reason we are building everypage.

You want to show people what you are building and you should have a call to action, usually one that gathers some data. This could be as simple as an email sign up (so you have an audience for every iterative launch) or you could throw up a google form to gather information about the problem your solving or any other way of getting something from the visitor.

Friends and family launch

As soon as you have an MVP get it out to your friends and family.

The example Kat gave of reddit was great - they got their first users by sharing it in the YC founder community which allowed them to get loads of feedback and start building their own community. The caveat here is to not spend too much time in this "phase", mostly because your friends and family are a) possibly not your target audience, and generally not early adopters and b) going to be nice to you.

Even if they aren't really target users though, you can always learn something by watching peoples reactions to your product, so get in front of them and show it off. I had the harrowing experience once of trying to demo my product to a non-technical friend and whilst demo-ing I forgot how to use the product because I didn't have my own computer where all the shortcuts were saved. It's great to get these hiccups out the way and also highlight areas that you might have overlooked since you are so deep in the product everyday.

Stranger launch

Find people that have the problem you are solving and get your product in front of them.

Again, the example Kat gave here was amazing - the founders of a furniture delivery company called Lugg waited outside IKEA, looked out for people struggling with their furniture and offered their service as a solution.

It's always nerve-wrecking to talk to strangers and even more so when you are putting your MVP (which you can probably find lots of holes in) in front of them. But it's so important because it really gets you to work hard at a) finding people that actually have the problem you are solving and b) succinctly describe your solution in their moment of need.

Online communities launch

Find communities of people who would like your product and share it with them.

This seems like one of the most common approaches now, but as Kat says in the video it's very unlikely to result in the "amazing" outcomes we so often hear about. The advice from Kat there was to understand what the community actually wants and not just go into random groups and spam them.

Looking at the results, I think this works best when either a) you have a product that is truly "viral" and everyone wants to talk about (but, let's be honest, you probably don't) or b) you have solved a problem for yourself and you share your product in the communities you hang out in.

This isn't always easy though. Earlier in my career I built a product for location based message (Kites). I built it for myself as an easy way to get recommendations for friends, but I couldn't think of who else would want it. I didn't share it to hacker news or the sub-reddits I was a part of because I thought "why would programmers want this?". So I shared it with travellers. It initially took off very well, but as a non-traveller myself, as the product iterated, it started to loose touch with travellers too. I think the moral from the episode was to not iterate too far away from the initial problem i.e. fall in love with the problem, not the solution.

Kat also highlights "Don't use jargon" or just "Speak like a human". I think this is super important, but again super hard. I've always tried to use the "friends and family launch" to try and get rid of all the jargon I can. One way I found to do this was to ask my friends to explain the product back to me or ask "what would you use it for?".

Request for access launch

Building viral loops into the product can get more users to your door quickly.

I've seen this pattern a lot recently and it definitely seems to work for some products. Kat gives the example of superhuman, who add a "sent with superhuman" line to your signature resulting in all your contacts seeing it. We have something similar in everypage, where sites made with the free account have "everypage" branding added to the bottom to lead visitors. In a previous product I asked the user for permission to post to twitter whenever they create content in the product and that worked relatively well too.

In general this launch works well for use-facing (mostly consumer) products, but it's difficult with larger b2b products. At my work we used blogging and "thought leadership"-type content to drive traffic to our products but the results were mixed. When you are pitching to large companies for a relatively expensive product, tactics like this come across as gimmicky.

I've seen many products that try the "wait-list and share to bump your place" approach. It's important to remember that these only really work when the demand is already there - either you have influencers talking about your product, or the problem is felt so widely that everyone just really wants one. Both of these were true for superhuman and also for hey.com, where the strategy worked very well.

In general this tactic is best avoided unless you find a natural fit within the product. Don't look for ways to add it on, like adding a wait-list to your new note-taking app - people won't care enough.

Social / media launch

Getting bloggers and influencers to talk about your product in a relevant context can drive lots of traffic.

Getting your product into TechCrunch is still one of the most "launch"-y things you can do (although somehow this doesn't feel as big as it did 5 years ago). When I used to talk about launching this was the perfect scenario I imagined.

Kat's example was a product that build wedding websites. They got into a few wedding related blogs and it drove a significant amount of their traffic. This really changed my thinking - she barely even mentioned TechCrunch! It seems obvious now - for getting customers it's much better to get featured in a blog (or something similar) where the domain is actually related to your field. The TechCrunch articles seem great for hiring and getting your brand know in the tech world, but I can't imagine many of their readers are looking for wedding website building advice.

On the the key takeaways here was something I fought with a lot on previous projects: "Don't pay for it". Kat suggested to find creative routes to get your product out there, don't use up what little capital your startup has.

Pre-order launch

If your product is physical, getting pre-orders is a great way to find customers.

The stories of how Pebble, Exploding Kittens and Sheertex (Kat's example) did well on pre-order sites like Kickstarter are very inspiring. To me the real lesson here is in how well these guys have validated their products before building them.

It's a bit tougher for software, especially when your first instinct as a technical founder is to write code. Even for me, having worked on products for close to 10 years now throwing away so much beautiful code, it's difficult to resist the temptation. It's definitely possible though and as many of you would probably know, writing code once you have users is so much more rewarding than just writing it for yourself!

New feature launch

When you add something new to your product, "launch" again!

Kat's example here was Stripe. Whenever they launch, they get involved in the community really well. As a developer I've defintiely noticed this and am always impressed with how well the founders and employees of Stripe address everyone.

I think the unsaid thing here is very important - you can't be spamming your community with every small feature you implement. With everypage, we are trying to build in a micro-products architecture (ye, I made this up!). Our plan is to re-launch small products as supporting pieces for everypage.

For small updates, it's best to launch within your own community. This could be as simple as having a weekly newsletter with the features you added this week. Courtland Allen does this very well with indie-hackers - they announce new features within into the community on indiehackers itself. It works very well because everyone there is already invested in the product (they wouldn't see it if they weren't).

Build a community

Finally, Kat briefly mentions that it's always great to build a community before hand. This has been talked about a lot.

Honestly, I struggle with this one as I'm not a generally social person. Whilst I would love to have a huge twitter following and a huge mailing list, I don't. But it's not going to stop from trying to build one out - from what I can tell the key ingredients seem to be clarity and consistency. So people should know what community they are joining when they sign up, and they should hear from you at some frequency.

Conclusion

Launching is often thought of as a daunting process. After hearing Kat's talk I'm convinced that we should be launching something every week - launching our product by talking about it out loud, launching a new pitch to potential users, launching a new pitch to a new category of users, launching a new feature to existing users, and so many more.

One of the questions you are asked very week as part of StartUp School is "Have you launched?". Now that I've watched this video it seems like a trick question - I launched the moment I spoke to anybody about my product. We will, of course, be filling in "yes" from next week on.

For everypage, our key takeaway is to continually be trying new things and measuring the impacts. It's easy to see how many new people visited your site from TechCrunch when they write about you and yes, it's lovely to see the analytics graph. It's harder to measure the impact of how well users are responding to your new messaging (and changes like this), but we will be analyzing these a lot more going forward - as if each one was its own launch.