Umso Blog

How we're hosting over 20,000 ultra-fast websites on a global network

Tue, 20 Feb 2024 17:12:21 +0000

As a website building company we wouldn't get very far if we didn't also host websites. When our customers click "Publish" in the editor we distribute their website files across our own global server network. While many website builders use 3rd-party CDN providers for this task, we've built our own system. If you work in tech, you might be rolling your eyes already.

I've never read about another company doing this so I thought I'd share our experience with those who might find themselves in a similar situation or are simply curious about this type of thing.

This is the story of how and why we built our own specialized, globally distributed, high-speed, SSL cetificate issuing, auto-scaling, auto-failover, hosting network with a 100% cache hit-rate on all sites.

Let's start at the beginning

Our website rendering engine is running on React. When a website is published on our platform, we turn the React component into static HTML and CSS. All published Umso sites are (mostly) static sites. When we first started we simply hosted all those files on a single AWS EC2 instance in the Oregon region. There was no load balancing, no CDN, or anything else beyond a server returning some html, css and media files.

Why even run a server if we just have static files? Why not use one of the many static site hosting services? There's one thing that makes us different from most people hosting static sites: we have thousands of them. And with those thousands of websites come thousands of domains and SSL certificates.

Manually setting up domains or managing certificates was never an option, and other hosting solutions weren't built for so many sites. Therefore our very first website server ran a custom Golang webserver with a plugin that would automatically issue LetsEncrypt certificates.

Where to go from there

As our company matured we wanted to provide our customers with faster websites and higher reliability than a single server in a single location could offer. However, we didn't want to deal with the effort of building and maintaining our own hosting network so we started looking for existing solutions.

Because we wanted to host static files with quick world-wide access, CDNs seemed to fit the bill. That's until we started trying them out. There's a lot of cool edge-computing, global network, cloud companies now, but the landscape was a little less exciting 6 years ago.

We looked into almost every option on the market and even tried some out in production for a while. However we'd always run into limitations sooner or later. And that's because we wanted a lot of things: pricing to support thousands of sites, quick invalidations, robust API, high cache hit-rates, willingness to support our use-case and flexible configuration options.

I could write an entire blog post about our CDN adventures but now is not the time. However I do want to bring up one point which I believe is an issue across the entire software industry, and that's approachability.

We're a small company with no outside funding or fancy connections in silicon valley and that means for a lot of large companies, we're simply not a worthwhile customer. However, in some industries (CDNs included) the large companies have the better product. Or simply the better track record.

If you're looking for a profitable startup idea, maybe have a look at where a successful company has grown so much that it's no longer catering to the smaller business. But then maybe don't, who am I to say?

Either way, we built our own CDN. Sort of.

The system we built

We like to keep things as simple as possible so you might even find our setup a little underwhelming. Do we run our own hardware? omg no. Serverless? no. Edge computing? kind of but not really. Docker? never tried it. I don't mean to be dismissive. My point is: you can get very far with a few simple tools.

Our entire system runs on AWS. We use Global Accelerator to provide our customers with an Anycast IP address which routes visitors to their nearest datacenters. We have four identical server setups in Canada, the USA, Europe and India.

They all consist of a Network Load Balancer distributing traffic to an autoscaling fleet of EC2 instances running our own custom Golang server. Our Go binaries (some 15mb) are built with GitHub actions and deployed via CodeDeploy. On average it takes less than two minutes from merging a PR to all of our instances running the latest code. We never SSH into our instances.

Each location also has a handful of DynamoDB tables and S3 buckets. Together with some security groups and IAM rules, those are all the resources we have in our hosting locations. Of course our website building product depends on a lot more infrastructure. The setup I'm talking about here only concerns the site hosting aspect.

There is no central server in our content delivery network. Each of the four server locations has all the data to function independently. When a site is published we distribute a copy of it to each datacenter. If our API and central Database go offline, our websites will still be available. If three of our network regions go offline, all of our websites will still be available.

Because we're a small team and don't like complexity, we didn't want to deal with database management in four different regions and we also didn't want to handle the data distribution ourselves. And we found the perfect tool to do it all for us: DynamoDB Global Tables.

While there are services that offer many databases with global replication, we liked DynamoDB for two reasons: it's serverless and it's cheap. While it wouldn't usually be regarded as a cheap service, it is for our use-case. We can cache files on the EC2 servers to keep our requests down and the table size is almost neligible.

As far as local caching goes, we run a distributed caching service as part of our main Golang program. All the servers in the same region share an in-memory cache which means that the cache size also scales linerary with server load in any region.

Speed & Reliability

Our initial goal was to improve the speed and reliability of our customers sites. Of course there were hiccups along the way while we figured things out but for a few years now our system has been running really solid.

We're monitoring our uptime and server speed with a third-party tool called Hyperping which makes requests to our servers and measures their speed and response. Our uptime of 99.998% means that over the period of 30 days we would have 52s of downtime. That's a really high availability and better than both Squarespace and Wix. We also make our uptime and response speed public on our status page.

Other Benefits of our own network

While our speed and uptime numbers could also be achieved with a third-party CDN or hosting solution, our own system has a lot of additional benefits.

100% Cache HIT-Rate: Because the edge servers get all necessary site files whenever a site is updated, they don't need to make any requests to an origin server. If a file is not in memory, it'll be loaded from the local DB, which takes no longer than 10ms on average.

Big CDNs are good at having high cache hit rates when few files are accessed by many people, but don't do very well with thousands of small files that are less frequently accessed. For our customers just starting out, every pageview counts so it's great that all of them are fast as well.

Instant site publishes: When there's new content, we don't have to invalidate the existing content, we just overwrite it. That means new content is available as soon as it has been globally replicated which takes less than two seconds on average.

Flexibility, Transferability: We have full control over the code of our server which means we can add new functionality quite easily, such as implementing site & page passwords or server-side analytics. While it certainly is easiest to run all of our infrastructure within AWS, we could also host our edge servers with another provider if we wanted to.

We own it: We don't own the infrastructure, but that's been mostly commoditized. We do own the system. That means we won't have to deal with sales teams, worry about deprecation, our provider pivoting, or anything that comes with buying a service from a third-party provider.

EU-endpoint: Because we already have a server setup in the EU, we added another endpoint which European customers can use to direct all of their traffic only to our EU servers. This negates the performance benefit that comes with a global network but it gives some customers peace of mind when it comes to EU privacy laws.

Cost: You might think that running all this infrastructure ourselves would be very expensive, especially with AWS egress pricing. However most of this is offset by the fact that we don't pay anything extra for additional sites. We essentially just pay for egress traffic and server load. With our code being highly optimized and our servers scaling down during slow periods, we end up paying not a ton more than we would with a third-party provider.

The downside?

It took a lot of time to figure out how to build all of this and how to make it work together. While we love what we have now, we'd definitely look for a third-party provider before building anything if we hadn't already. All we'd need would be: unlimited sites, flexible pricing, high cache hit rates, instant invalidations, flexible configuration, a solid and well documented API, root domain support, and a long history of reliability, without having to drag ourselves through a long sales process.

There's also the issue of having to maintain it all. However, because we have kept the number of moving parts as low as possible, there's not a lot to do. It would be great to just pay someone else to deal with it entirely. However, we would still need to make our origin server available at all times if we used a CDN. If there was a service that would handle all of it for us, we'd still have other infrastructure to maintain. We're an always-online company either way.

Conclusion

We did the not-very-startup thing and built something ourselves which is already being provided by a giant industry. It turned out to be a lot less complicated than one might think. Somehow this worked out well for us and I hope this article might inspire you to go down an unusual path and learn something new as well.

If you find yourself in a similarlish situation and want to run your own global hosting network, you might get very far with just GlobalAccelerator, DynamoDB Global tables and Caddy. If any of this has been helpful or if you have any questions, feel free to reach out to me on Twitter. Thanks for reading!

Quick Q&A:

Why don't we have more server locations? Diminishing returns. Big difference if a round-trip takes 200ms or 20ms but form 20ms to 15ms not so much.
Why don't you use XYZ technology? What we use works well for us. There's lots of a ways to solve a problem but we only need one.
I know of a service that does all of this! Please let me know.
Do you have any plans to productize this? While that has been a daydream for a while, we already have our hands full.
Can you open-source it? No sorry, this is too ingrained into our business logic.
How do you handle DDoS? Very fast scaling and caching

12 AI powered tools to grow your startup. (No, they’re not all ChatGPT.)

Wed, 05 Apr 2023 04:00:00 +0000

When used as what it is–a tool–AI can be a powerful member of your business strategy. It’s not a replacement or a magic wand (wouldn't that be cool?), but it’s definitely an ally that can help you save precious money and time, something those of us in small teams particularly appreciate.

As with any tool, it helps to know how to wield AI to make the most out of it. Although there is nothing like practice, a pointer or two from an experienced sufferer (that’ll be me for today) can’t hurt. So here is a list of the best AI tools out there, designed to help you squeeze lots of value from the beautiful technology available to us right now.

AI for content writing

Blank page syndrome haunts the best of us, and there is nothing worse thank knowing exactly what you’d like to say, yet not be able to just write it; so you close the untitled draft, and promise to come back again tomorrow.

AI can help. And as far as writing tools go, the list does go on.

Grammarly.com

Often the first one to come to mind, and a recurrent fan favorite due to its ease of use, Grammarly helps users write cleaner content through writing suggestions, spelling, and punctuation corrections on your text. Is great at spotting typos on an email–or a blog–and keeping a consistent tone through your writing.

It’s also introduced a number of AI powered tools recently, including a desktop app designed for writing, and an upcoming AI writing generator under the name of GrammarlyGO.

Creaitor.ai

But as far as AI writing generators are concerned, the internet has its fill of those. Creaitor.ai is one of those copy generators that truly tries to cover all bases. From website copy to blogs, song lyrics, and even love letters, all the way to marketing content such as social media copy and Google Ads—the list goes on.

Creaitor's interface is easy to use and intuitive enough, and as with most AIs available, the output you get is only as good as the input you give. Specificity yields the best results, and though there’s truly many shiny things you can do, I see it best as a tool to use with a specific goal in mind. Say, writing an ad with certain keywords, or laying out a few copy ideas to pick and choose a final one.

Here is me dorking around and asking Creaitor to come up with a recipe based on some basic ingredients. I was hoping for pancakes, and that’s exactly what I got, so 10/10 on that one. (Though it's a rather simple request, so definitely amp up the expectations for anything relating your business' content.)

Sudowrite

Designed primarily as a tool for writers, Sudowrite takes out the fad and gives you a perfectly neat space to focus on what you need to do: write. The AI is comprehensive and responds well to instructions (e.g. a short script for a movie based on four characters looking for x, y, and z), plus, it allows you to edit your drafts and pick from the best parts of each version, which is nicer than going back and forth on different drafts and different docs (been there, done that).

Sudowrite is ideal for bloggers. It’s perfect for bouncing ideas, outlining articles, and brainstorming or rewriting content. It’s impossibly simple to use, and offers very basic–and therefore very effective–options for the writers in your team.

None of the tools above will write groundbreaking copy in your stead (phew!), but they will help you get through blank-page syndrome, and can overall help you become a more effective writer; Especially if you are juggling content creation alongside the many other responsibilities of running a business, or if you are just getting started with website copywriting, or blogging, and need a little push to feel more confident.

AI for design

To my chagrin, I'll admit I am not the most skilled designer, and often rely on my teammates skills for visual assets, which isn't always ideal when you have limited capacity. Though definitely a hard-skill I need to work on, it's relieving to know there are currently tools out there that can help me now.

Colormind.io

I’m a fan of tools that do one thing, and do that one thing well. Colormind is one of them. It gives you a neat color palette in seconds using colors taken from a picture, or a hex-code. It prompts out user cases and several templates to use, and it's overall a solid tool if you are looking to harmonize (or create) brand elements.

Runway

A hub for everything visual, Runway offers AI generated video and imagery, as well as advanced editing tools for creators. It's packed with tools, including background noise reduction, subtitle generation, and green screen backgrounds for videos; plus color changes, blur, and filters for your images.

You can also “train” your AI model and customize designs to look exactly how you want them; as it goes with AI, the more you use it, the more it learns.

I’m not running to have this one framed or anything. But for a vague prompt of <<cats eating tomatoes>> I’ll admit it’s got its charm (maybe don't look at it too closely, though).

Canva

With its latest round of updates comes Canva Create, a new branding hub with text to image, advanced picture editing, and AI writing and design, where–in a step further from their traditional approach–you can now simply tell Canva what you’re looking to create, and let the engine build the templates for you. Or work from an image, as I did below.

Quite neat!

Canva has paved the way as an accessible design tool for the every-day user; bulk creation features save tons of time, adding your branding elements simplifies the content creation process, and being able to post (or schedule) straight from the app makes social media management a breeze. Though its AI features are still in beta, they are already a valuable addition to the Canva toolshed, and one certainly worth keeping an eye on.

As far as design is concerned, AI is making strides every day, and in the meantime, for the designers out there, there are plenty of tools and extensions available to make their current workload a tad lighter.

AI for marketing

We are not yet at the stage where an AI engine can create and run a media strategy for us—but we are getting close.

Albert.ai

What a time to be alive, when there is a little AI engine that can seamlessly help us optimize advertising dollars across the web.

That’s exactly what Albert does! It works across all major social media and search platforms (Meta, Youtube Google Ads, the lot), and helps test and tweak campaigns to optimize your campaign money; It scouts for advertising blind spots and automatically redirects the target to make the most out of the content, time, and money you’ve predetermined.

Reporting is comprehensive and data-rich, with top insights prioritized to help you and your team make better decisions, faster.

Surfer SEO

Maybe I have a thing for writers, maybe I recognize how important good content is, and how critical it is to have it be SEO optimized if you want to see its performance soar. Maybe I’ve just been using the Surfer SEO tools for a while, and happen to enjoy them quite a bit. Who knows.

The fact is that few tools offer such an optimized experience as Surfer does. From SEO friendly outlines, the handy keyword finder, all the way to the content editor, it really is the home for SEO optimization.

The sweetest part? Most of it happens automatically!

Though its text recommendations can at times be wonky-looking, they’re not meant to be a final product, but a guide to help your copy be as search engine friendly as it can be, bringing out relevant keywords by region, industry, and even demographic.

Semrush

The <<all-in-one>> of content marketing, ready to run campaigns, optimize your SEO, run your social media calendar, and track everything with its analytics solution. The bulk of its value lives in the SEO solutions, going beyond keywords search and ranks, the tool-shed includes backlink building solutions, domain audits, and in-depth SEO writing support powered by AI.

It’s also got competent advertising and marketing features, especially for things like PPC advertising, and competitive research.

AI for sales

Sales enhanced AI is one of those game changers that gives you back the one thing you can’t have more of: time.

More time to invest in the good leads, less time lost in the unqualified ones, and less time spent adding contacts, and following up, and adding meeting notes for the next person to find, and updating that mailing address, and... you get it.

Coming from sales myself, I can’t tell you how much I wish we’d had at least a few of these technologies available way back when.

HubSpot

Best known as a one-pit-stop CRM for small businesses, HubSpot has recently introduced a couple of new AI enhanced features, and our eyes are particularly pointed to its ChatSpot.ai, designed to make the HubSpot experience a seamless one.

ChatSpot can pull HubSpot data for quick reference, draft follow-up emails, add contacts to your leads lists, and even prospect new sales opportunities for you. Combined with other HubSpot AI resources (such as data cleansing and import), it truly takes the power of a good CRM to a new level.

People.ai

Packed with integrations (Marketo, LinkedIn, Outreach, Salesloft, Demandbase…) People.ai is an optimizer to the sales cycle, needling its AI superpowers into every step of the pipeline, guaranteeing salespeople make the most out of every qualified lead, and, perhaps more importantly, leaving out the noise of unqualified prospects.

Drift

One of the most competent chatbot solutions out there, Drift’s AI is able to qualify leads, answer customer questions, and automatically merge to the sales pipeline, all in multiple languages!

It integrates with Salesforce, Marketo, Salesloft, Fullstory, HubSpot, MS Dynamics and GSuite, to name a few.

Drift also helps you accelerate the buyer journey and connect qualified buyers to your sales team, monitoring every step of the way to optimize the sales process and increase productivity.

AI is only going to get smarter from here, and the plethora of tools flooding the market today are only a small fraction of what we can expect to see in 2023 (and beyond).

Personally, I’m excited to continue trying the technologies that simplify every-day tasks, and give us more time for big-picture stuff (exciting product updates! writing thought-leadership pieces! stamping Umso’s marketing all over TimesSquare!), and though most of these tools are still in early stages, there is plenty of value to be reaped from them as they are, and only more to look forward to.

What AI developments are you looking forward to? Are you using ChatGPT4, or any of its variants into your strategy? Let us know!

One price, All Features. The simplest pricing in the industry.

Thu, 05 Jan 2023 17:19:18 +0000

Why did we change our pricing?

Glad you asked! After several team meetings and an in-depth analysis or two, we've realized our old pricing wasn't great. So we're changing it.

The industry's leading model of "the more features you want, the more you'll have to pay" is admittedly limiting, and startups often find themselves constricted within the bounds of heavily tiered plans, where your business' growth becomes conditioned to how much you can pay to keep growing.

But it's hard to grow if you haven't got the right tools, isn't it?

Even the more "affordable" options include odd categories like "seat-based-pricing" (AKA team members, a variable that will inevitably change as you grow), and important storage limits. Talk about a set-up.

Running—let alone starting—a business has only gotten more expensive, and these tier-based models are the easiest, least questioned way to turn a profit for the organizations offering them. At the end of the day (and the fiscal year), most of these companies have investors to respond to, and face consistent pressure to show incremental growth.

Fortunately, it doesn't work that way with us. At Umso we aim to remain a sustainable business (for you, and for ourselves, of course) and continue improving to reach even more people. We've got no investors to share returns with, and no boardroom of executives awaiting quarterly results.

Which is why this pricing change has been a lot simpler to take from an idea to a reality. It's a decision made to benefit all of our important "stakeholders". That means you, our user; and us, the small, bootstrapped team behind Umso.

Nice sentiment! Now, how did pricing change, exactly?

We like things simple and transparent, and this pricing redesign keeps those values at heart. From now on, there are two plans you can work with: Free, and Paid. And everyone loves free stuff, so let's start with that one.

What do Free sites get? Pretty much everything!

While custom domains and badge removal options are still only available on the paid tier, essentially every other feature can be found in the free model.

Really, all of them.

Multilingual pages
Mailing lists
Customized code
Analytics dashboards
Password protection
Cookie banners
Custom forms
Blogs (like this one!)

The tools your website needs to grow, at no cost. Of course, you get unlimited team members for your sites, and you are free to create as many pages as you need, which comes in handy as your business evolves from a landing page to a fully-featured site.

When you are ready to take it even further (that is, adding your custom domain and removing our "Made with Umso" badge) switching to a paid plan is a breeze.

What's on the Pro plan? Everything.

Our former pricing tiers have been compacted into a single one (goodbye “basic”, “startup”, and “business” plans, and thank you for your service), where you can find all of the Umso features for a lean $25/month—or $20/month, if you go for the yearly subscription.

As we mentioned above, on top of all features, the paid plan also includes the ability to add custom domains, and allows you to remove the “made with umso” badge from your pages.

The paid tier is great for businesses ready to make their brand shine on their website, and if you are coming from a free plan, there is nothing new to learn, just the features you already know and love, now with your name all over it.

What to expect from Umso, and this pricing change:

We’ve been through several price changes over the years—with varying degrees of success—and learned from each of those iterations, searching for a formula that rightfully balances “value for the money” and business sustainability.

Although this pricing change isn’t the biggest one in terms of money (it’s certainly not an “overnight millionaire” plan, and we’re fans of modest success), it’s impactful where it matters, giving more value to our current and future users.

And now, this is the part where I kindly invite you to try Umso out (if you haven’t already). Go ahead, it’s—literally—free!

P.S. If you are currently on a paid plan and wondering what these changes mean to you, there should be a mint email in your inbox with all the details. You can also review our Canny entry on the topic here.

P.S.S. You can always reach us with questions about the new pricing (or anything else Umso!) at help@umso.com. This should be a smooth transition for all users, and we are here to help.

What we've been working on and why you can't see it (yet)

Tue, 13 Dec 2022 09:45:01 +0000

We haven't shared many product updates over the past year, but that's not because we've been quarantining on the couch (I promise 😇)! Instead, we took a step back and worked on some big changes to our internal systems. First we'll explain what they are and then why we've prioritized them.

New Engine: Our "engine" is what renders your websites based on the design and content you created in the editor. We have learned a lot since we started and now we've used that knowledge to build a new system. In 2020 we rewrote the engine to become a standardized set of components which we can be more easily extended. This new version has been live since the end of last year and you probably haven't noticed. Sounds boring? Not at all, just read to the end!

New Editor: Our new editor has been a work in progress since May 2020 and will enter open beta very soon. As with the new engine, our goal was to standardize everything to make new functionality easier to add. We've also used the opportunity to improve the design and performance.

New Hosting: We have moved around between CDNs to host your websites but none of them really suited our situation. That's why we built our own CDN which has been in production since February 2020 and provides some of the fastest hosting in the industry.

Why we did this now

We did none of those big changes for fun. We would've much rather worked on beautiful new things you can see right away, but there's a good reason. When I started Umso nearly three years ago, it was a minimum viable product. That means it was just meant to help me figure out whether or not it could become a valuable business.

Many websites, paying customers (thank you so much 🙌🏻), and two full-time employees later we have decided that YES! it is valuable business, hooray! Now we had to figure out how to turn the successful MVP into the amazing product we actually want to build.

We have a great vision for what Umso can be. However, those things add a lot of complexity which our existing Engine, Editor and Hosting couldn't handle. That's why we worked so much on improving them.

The things we will build next

You can already find most of this on our Roadmap, but here's a quick summary of some big updates that wouldn't have been possible without our systems overhaul.

Multi-Language Support: one of the most requested features would've been an impossible mess without our new Engine and Editor. With the new system however, it'll be a breeze (we've already tried it out!).

Engine design overhaul: some of our sections have been designed three years ago and haven't been updated much since. Now that the Engine has been completely overhauled, we can use it to create much more beautiful and responsive designs.

Dynamic data: Because we have our own hosting system with 100% customizability, we will be able to soon support dynamic data sources on our edge servers. With the new Editor you will be able to create the templates for that data.

We have a lot more planned, just head to our Roadmap to find out and subscribe to stuff you're interested in.

Will Umso turn into another super-duper complicated website builder that I need a PhD for?

No. We believe that simplicity is key to our product and we will continuously work on keeping it that way. In fact I think we can make it much more user-friendly while also adding new features.

Our goal is to let you create a beautiful website in record-time and edit it with ease. However, you should also run into very few roadblocks when you need more advanced functionality.

We're still a long way from realizing our vision, but we're getting closer every day and we're excited to share this journey with you!

Is your website multilingual yet? It definitely should be!

Tue, 20 Sep 2022 17:45:54 +0000

Why, you may ask?

There is the obvious: Serving a broader, richer market. A larger chunk of potential customers, plus a greater regional reach equals increased sales.

Yet this success isn't only reliant on translating your website's content. Google can do that, and for free. It might change pronouns here and there ... and turn carefully crafted copy on best-selling products into an awkward concoction of nonsense ... and leave visitors with more questions than answers.

But hey, at least it's in their language now, right?

Right.

Or you can take a different approach. One that recognizes the importance of not only translating, but localizing content in a different language. You are doing more than translating words on your website; You are opening your business to a new market.

Language is a comfort-zone, and a deeply-rooted part of someone's identity. Wouldn't you want to put your best face forward, and greet the people like a local? To feel invited, rather than invasive?

Why, we've got just the guide for you! Let's explore some less-known perks of having a multilingual website, such as a heightened brand perception, and climbing the international ranks (SEO ranks, that is).

Different language, same rules apply: it's still all about the customer.

Language builds trust. By saying things "in their language" your audience's relationship with your brand will transform from "foreign" to "familiar". A relationship that translates (pun-not-intended) into literal sales:

Out of 8,700 participants of a study (across 29 countries), 76% stated their preference for buying products with information in their native language.

40% said they "wouldn't even consider buying" from websites in a different language; and an outstanding 75% declared they are "more likely to buy again from a brand if customer care is in their native language", even when they had a confident grasp of the English language.

People will appreciate marketing and content in their home-language. It's personal, and inviting. It will be welcomed by engagement with the content and, more importantly, (repeated) purchases of your product.

Is worth mentioning that this approach will also do wonders to position your business in a new market. But we'll get to that later.

Climb the (SEO) ranks.

For organic growth nothing does the trick like SEO. Consider that 72% of web-users spend most (if not all) their time on the web on websites with content in their native language, and a well translated website will support overall online presence.

More than merely translated, you want content to be localized. This means looking out for spelling details (is it color, or colour? well, it depends!); region specific mannerisms and phrasings (Spanish from Spain ≠ Spanish from Ecuador); and even country specific jargon.

An online search can go from the very broad to the nitty-gritty specific. Your goal is to organically blend in keywords that best-match your industry, whilst artfully applying cultural and language differences to your benefit.

Optimizing for SEO also looks like translating the meta-data of your site, so it will properly rank in regional search engines. Adding unique URLs with language specific subdomains will also make it easier for the SE to find your page and rank it.

These are simple-to-implement changes that have a significant impact in your business' organic ranking, especially as you breach your way into new territories.

⚠️ Not all SEO words are made equal. ⚠️

A ranking keyword in English might not be an exact translation to its Portuguese or French counterpart. Do your research and, if possible, work with a marketing specialist to devise the best target-language SEO strategy.

Increased accessibility. AKA increased sales.

The premise of a multilingual site increasing sales only stands true if the contents of said site are valuable. The same way it works with the content in its original language.

Making your content accessible means widening your marketing channels in a target-language. Translating video subtitles and creating social media accounts for audiences in different regions is a great way to build a connection.

It's important, too. 68% of online customers like to interact with brands in their own language. You know the sales are there, first take the time to build a relationship and become someone people want to buy from.

A visible language switcher, localized-copy, and well-translated versions of high-value assets are all examples of accessibility. Consider also "tropicalizing" your brand beyond the product. Invite visitors to immerse themselves in your exclusively crafted-content, and let potential customers truly understand the product.

Be more successful than your competitors. In every language.

The internet isn't exclusive to English speakers, in fact, only 25% of the internet speaks English. There are goldmines of untapped customers awaiting beyond the language barrier... you just gotta meet them halfway.

As the proverb says,

"A well translated website is one you can't tell has been translated."

Overlooked wisdom.

A study showed that more effort is given to the English version of websites than to their non-English counterparts, and that page responsiveness and usability was higher in the English versions. For bilingual visitors, this creates a negative impression towards the brand, as there is a clear lack of consideration for their "home-language".

Beating your competitors is not necessarily about getting there first, but getting there better. With the time and dedication it takes to make of your website a great place to visit in any language.

And the best part? It happens naturally. As long as you keep the customer at the heart of it (which you are already doing anyways, right?).

Keeping the customer at heart encompasses everything we've covered in terms of localizing content and connecting with users in their native language. It might mean adjustments to marketing strategy, pricing, and even business model—you will follow the necessary steps to show your brand's commitment to doing business within a new group of people, in their language, and on their terms.

Wrapping-up.

The large corporate conglomerates aren' the only ones who get to enjoy international tenure. A website translated into multiple languages is possible for every business.

Translate high-value content first, and localize your copy, marketing, and SEO efforts to math that of your new target market. Connect with people in their own words.

You don't need grandiose budgets, either. Begin translating top-ranking pages and build up from there. If you are bilingual in your target-language, take the time to translate and revise your site's content. It's also possible (in fact, recommended!) to hire a professional translator to do it for you, a service you can often get for a couple hundred-bucks. Additionally, make sure to assess your meta-data and subdomains to mint search engine rankings and guarantee a successful multilingual debut.

Remember that multilingual site goes beyond increasing sales and showing-off your expansions maps (though hooray for that!). It also shows compromise, builds customer loyalty, and transforms a brand into a higher version of itself.

Oh! One more thing, while I still have you here...

if you’re looking to build a multilingual website that allows for all the important items we've just talked about—look no further! Umso is the website builder for you.

Our multilingual features make it a breeze to translate and section content for different languages. Plus, you can add language specific subdomains, and edit page meta-data to optimize for search-engines in other regions.

Did I mention our websites are beautiful and simple to work with? Because yes, they are. Of course they are.

What the best No-code products do right, and why it’s not about code.

Mon, 20 Sep 2021 14:59:49 +0000

I've learned a lot about product design over the last three years of working on our website builder product. When starting out, most of my product decisions were based on intuition. I wasn't sure what the product should look like, what features to add, and which ones to cut.

With practice and experience however, some common themes started to emerge and things got easier. I developed a mental framework for product design decisions. For a long time these ideas have only existed in my head. I've finally been able to put these ideas into words and wanted to share them with you.

These ideas apply especially to no-code tools such as Umso, but could also be applied to most other software products. If you're at all interested in product design, user experience or no-code tools, then you will hopefully enjoy this read.

What's No-Code?

No-code tools are getting more and more popular. They allow anyone to create websites, applications, and automations, which could previously only be created by experienced programmers.

Not having to know code is a major advantage of these tools, but a good no-code tool doesn't just replace code with a pretty user interface. Code is very flexible and complex. If we tried to match that power with a no-code tool, we'd quickly run into trouble.

Designing a no-code tool

Let's have a look at what happens when we try to replace code with a user interface, the problems we will run into, and how to fix them. To do this, we will design a simple no-code tool for building websites.

So far, so good. The graphic above shows the code and the no-code approach to displaying a simple text. Instead of having to know the HTML syntax for a title, the no-code UI lets you type in your text, and select a text type from a dropdown menu.

Now let's talk about something you can't see in the example above: flexibility. With the provided UI, you can not change the color, position, background, or any other aspect of the text. How boring!

In this next example we will add some more settings to our no-code UI to make our website title a bit prettier. The code column has also been updated to reflect those same changes.

Things are starting to look a little more complicated. We’ve got 8 lines of code and 6 UI inputs, all for a little box with text. The settings now allow you some more creative freedom but are still very limiting.

Before we make things even more complicated, let's talk about optimizing the UI.

You could argue that my design isn't great, and you'd be right! Settings could be grouped differently or hidden by default. There could be a better visual hierarchy.

However, no matter how you organize the UI, there is one thing you can't change: you can not remove any of the options without making them unavailable to the user. Even worse, if we want to change more parameters, we need more input fields. That can quickly get out of hand and it wouldn’t be any fun if I didn’t show you.

When our No-code falls apart

People who use our no-code tools expect useful and beautiful results. We need to give them the option to create more complex layouts, apply more styling and use different design elements. Let's give the people what they want!

Oh my, that's a bit though to look at. Our UI has quite a lot of options now for such a little box. What would this look like if we had an entire website to design? What about all the design options that aren't even available in this UI?

As you can imagine this wouldn't end well if we kept going with our current approach. Our customers want better results, which require more design options. More design options will cause a more complicated UI. We simply can't match the flexibility of code with our UI. So where to we stop, what do we do?

How no-code succeeds.

The easiest way to create a better UI is by limiting the parameters a user can control. However, the more limits we impose, the harder it will be to create high-quality results.

Since we don't want to jeopardize the quality of outcome with our limitations, we can apply a common programming approach: abstraction. We can pre-create common elements which can be modified through limited parameters.

When we create these abstracted elements, we need to make some decisions on behalf of the user. For example, we might decide what technology is used to achieve a certain effect. This will be something the customer can not control. In the programming world, we call this opinionation.

It can be hard to know which things to limit and which to allow, what to abstract and how to implement it. That's where our super-power comes in: knowing the customers goal.

Instead of trying to find out what the customers goal is, we can design our no-code product with a specific goal in mind and then let the customers come to us. Umso, for example, is marketed as "the website builder for startups". The goal of our customers is to create a typical startup website.

The more specific your goal is, the easier it will be to build a good no-code solution. If you have a no-code product with no specific goal in mind, then this approach won’t work for you.

In conclusion, a good no-code product uses Goal-oriented Opinionation, Abstraction and Limitation: GOAL. Let’s see what that looks like in action.

GOAL UI

This is what no-code looks like at its peak. We turned 22 input fields into 4 and got the same result. With this UI, you can create the same design as above, but without the hassle.

You can't create anything else with it and that's ok. For each new common element on a website, we can create a new GOAL UI. Because we know what our customers want to achieve, we won't need too many.

GOAL, the No-code Code

Yes, that's an oxymoron, and it works well to prove the point. We can apply the GOAL principles to code as well. Even if you're not a programmer, you can see the difference.

Our code on the right side is short and sweet and helps programmers in the same way it helps no-code users. In fact, this approach is very common and programmers use their own and third-party GOAL code all the time.

Summary

The example below shows the GOAL approach with both code and a UI. Both approaches look really similar here, and very different from their previous versions. That’s why it’s not about code. It’s about meaningful goal-oriented, opinionated, abstraction and limitation.

This concludes our little journey. Please keep in mind that neither the code nor the UI are perfect and only serve as visual guides. This approach has worked for us but certainly isn't the only one. If you have any feedback, let me know on twitter @felixguertler or @umsohq. No code was harmed in the making of this blog post. Thanks for reading.

Custom Forms Update

Thu, 04 Feb 2021 23:04:40 +0000

Umso is built on the concept that powerful tools don't need to be overly complicated. We are constantly thinking on how we can build the simplest feature that can deliver the most value to our customers.

One of the most used Umso features is custom forms, which we released about a year ago. We have heard your feedback, worked on bug fixes as soon as we took notice of them and constantly iterated on it until we delivered a very flexible, yet simple form builder.

Form Design

You should be able to use your custom form on many different sites and pages, and they can have completely different themes. The form structure is detached from its final design so you are able to add the same form in different places and freely design each one of them.

Using our brand new editor, you can change the background color, the color of the labels, placeholders, fields background, and so on. Play with the size, change the divider, and much more.

Managing Responses

Once your custom form is ready to rock, you can see all of your form responses in your form responses dashboard, where you can filter and sort any column. You can also export form responses as CSV file, get email notifications for new submissions and send data to your own webhooks.

Want to delete a response? No worries, we've got you!

Zapier & Custom Webhook Integration

"Hey, but I don't have time to come to the dashboard, grab the responses and go to the service I use to manage the responses and paste it one by one.. my time is precious!!"

Don't worry, we've got you covered! You can integrate your form with Zapier, and if you are more tech-savvy you can also integrate with your own webhook.

We have a guide to help you set up your Zapier integration. It allows you to send your form responses to over 2K apps! #notbad

If you need to send the responses to your own webhook, you can go to your form editor and add it there. We make a POST request, passing the response as JSON in the body of the request.

Spam Protection

Worried about receiving spam? Captchas to the rescue! Our captcha is a simple text challenge that your customer will have to answer to be able to submit the form. We use our own captcha solution instead of the popular Google reCaptcha. This way your customer's privacy will remain protected and they won't have to find crosswalks, traffic lights and busses.

You can set it to Always to trigger on every response, Suspicious Content which will be evaluated by our system and trigger depending on a number of factors, and to Never display it, so your customers don't need to bother with it. To set up, go to your form editor and you can find it on the left side, along with the other form settings.

File Upload

We have added a simple file uploader that will let your customers upload a file when submitting a response to your form. You can now add it as a field when creating or updating your form. The file format supported so far is the same as our file library, but we will in the future support more formats.

What's Next?

We will still keep improving and polishing it every day! Make it as user friendly as possible and add more features along the way. Needless to say, we are very happy to be delivering custom forms to our customers and we hope they can be very useful to each one of you.

We are also happy to hear from you and your needs. If you think of anything that can be improved, please let us know.

To stay in the loop about new updates, you can subscribe to product updates via the product changelog.

Landen is now Umso!

Fri, 06 Nov 2020 05:22:48 +0000

Same people, same company, same product, new name!

We’re excited to announce that as of November 2020, we’re rebranding! Part of rebranding for us will be changing the name of our product and company from Landen to Umso. Our vision hasn’t changed and we're still on the same mission to build the best website builder for entrepreneurs.

The new name doesn't really mean anything, and we won't pretend otherwise. It's a short and sweet imaginary word. We've spent a lot of time trying to come up with a new name and had a lot of different ideas going in different directions. In the end, we had 56 candidates with affordable .com domains, some of which we're embarrassed to even say out loud now.

You may wonder why we’ve changed our name. Some of you really liked the old one. However, we felt that it created a preconceived idea of our product that didn't align with our vision of a well rounded website builder solution. The name Landen makes many people think of "landing pages". That means different things to different people, but it roughly translates to single-page websites at the end of a marketing funnel.

What we have been building over the last two years initially started as a landing page builder. Fortunately, it has evolved into so much more. Our goal is to build the best possible website builder for entrepreneurs and startups, because we want you to succeed. Our new name frees us from the limited idea of what our product does.

Q&A

Why do you say "we"? For a long time, Landen was a company of one while I bootstrapped it to a point where it would cover my living expenses. In May 2020 , the I became a we when I hired our first employee, Simon. In August, our team grew by another great mind, João. And now we’re a team of three!

Did you raise any money? No, we are completely funded by our amazing customers. Thank you to all of you for trusting us and for being awesome customers.

Can I keep my .landen.co subdomain? Yes! We’re not forcing anyone to change their free subdomains. If you currently have an xyz.landen.co subdomain, then you can keep it, However, all new websites will get a .umso.co subdomain. Whenever you're ready you can switch your .landen.co subdomain to a .umso.co subdomain in your dashboard (for free of course).

What about product updates? Glad you asked! The name change was just a little side project for us. We’ve spent most of our time improving the product itself. You might not have seen much of our efforts lately, but that will change soon. Keep an eye out on our changelog!

Anything else? Just email us at help@umso.com and we'll answer your questions!

Tool recommendations for technical founders

Sat, 15 Jun 2019 00:24:31 +0000

These are the tools I use to develop Landen. That includes front-end dev, back-end dev, database management, product design and everything else. You don't need much these days to develop software. I use the cheapest MacBook Pro and a lot of free or relatively cheap software, but it could all be done on a $200 laptop with completely free software. Most of the software I pay for also offers a free trial or limited version.

Landen is not affiliated with any of these tools or companies and I get no money for my recommendations. I've included links for your convenience.

Visual Studio Code

Free

I've been using Sublime Code for the longest time, briefly tried Atom and finally found Visual Studio Code, which really got me excited. It's free, so if you haven't tried it yet, I absolutely recommend you do.

There's a healthy plugin ecosystem around VS Code, and I use a few fairly basic ones:

Go: absolutely wonderful for Go development
Import Cost: Display JavaScript import size in the editor
One Dark Pro: Color theme
Prettier: opinionated code formatter
Reactjs code snippets

VS Code Website

SQLEditor

$79 once - 30 day trial

This is an indispensable tool for planning and maintaining SQL data structures. It generates the SQL statements for you for PostgreSQL, MySQL and a bunch of other dialects. As my databases grow more complex it helps me see everything at a glance. SQLEditor Website

Postico

$40 once - free version available

Postico is a PostgreSQL client for macOS with a beautiful and simple interface. It supports filtering, automatically finds data with foreign keys, has an SQL editor with autocomplete and comes with dark mode. It's developed by a software developer duo in Austria. The value for money for this tool is amazing, they could easily charge 3x as much. Postico Website

Querious

$49 once

Querious is similar to Postico but for MySQL databases. It's even more powerful and also a bit more expensive but if you regularly work with MySQL databases, it's well worth the price. Querious Website

Quiver

$10 once

There are probably as many code snippet managers as there are programmers. Quiver has been reliable, easy to use and organize. It doesn't come with tons of bloat and you don't have to pay a recurring fee for it. Quiver Website

PostMan

Free to $18/mo

If you work with RESTful APIs this tool is great. It let's you send network requests with a simple graphical UI. You can also save and organize requests in groups and regularly test them automatically. Another noteworthy API tool for macOS is Paw ($50). PostMan Website

SVGO

Free & Open-Source

SVGO is a command-line tool for optimizing and minimizing SVG files. If you export SVG from a design program such as Sketch, it'll often contain lots of unnecessary data which can be safely removed before using it on the web. There's also a SVGOMG, a web UI for svgo. SVGO on GitHub

Sketch

$99 (one year of updates)

Sketch is a design tool. I used to design UIs in Adobe Photoshop and I'm very glad that's no longer the case. Sketch is much simpler and geared specifically towards UI design. There are also some other very modern and noteworthy design tools such as Figma, Framer and InVision studio (free). I simply use Sketch because my requirements are simple and that's what I'm used to. Sketch Website

Todoist Premium

$48 / year

This one's a to-do list app which hits right right balance between being powerful and simple. When you're a solo founder you don't need too many team features and especially no productivity apps that take more time to manage than they free up. Todoist has a free tier as well. Todoist Website.

Firefox

Free

Firefox has caught up with Chrome in most ways. I now work mostly with the FF development tools, and only go back to Chrome for a few things that FF doesn't offer yet. Firefox Website

TinyPNG

Free

This tool is like SVGO but for PNG and JPG files. Before you use any PNGs or JPGs on your website you should compress them to reduce their file size and make them load faster. TinyPNG does so easily and quite effectively, sometimes compressing image files by around 80%. TinyPNG website

ScreenFlow

$129

I use ScreenFlow to record GIFs and product videos. It's much easier to use than most pro video editors. It is to Adobe AfterEffects what Sketch is to Photoshop. Much less functionality, a lower learning curve, cheaper price and still everything you need for simple videos. ScreenFlow Website

MetaBase

Free & open-source

MetaBase is a handy tool for querying and visualizing your SQL database. You can run it on a server or on your local dev environment. The setup is super easy and even though it still has a couple of bugs, it's relatively easy to use. MetaBase Website

These are all the tools I can think of right now. If you have any recommendations or feedback, let know on twitter.com/felixgurtler.

Analytics Update

Tue, 11 Jun 2019 16:26:01 +0000

For a while now the Landen dashboard has included some basic visitor analytics like page views and unique visitors. Of course your website should also have a good conversion rate, that's why conversions are now being tracked by Landen as well.

Along with the new conversion tracking, the analytics dashboard has gotten a complete overhaul. Everything that's new is described in more detail below.

Conversion Metrics

The conversion rate of a marketing website is one of the most important, if not the most important metric. If your content gets a lot of attention but visitors don't convert to signups or paying customers, you have a problem. That's why Landen now includes conversion tracking with advanced analytics (available on all paid tiers). The feature automatically tracks contact form submissions, email list signups, Stripe checkout clicks and link clicks.

At the same time it's important to respect the privacy of anyone who visits your website, that's why Landen never saves any PII (Personally Identifiable Information) along with engagement data.

One more great addition to the conversion tracking is the support for custom events. You can include a small, fast and secure tracking script on your online store or web-app website to track events such as signups and purchases. Learn more about custom events.

Page views & Unique visitors

Page views and unique visitors aren't new to the analytics dashboard, but they have been improved, how displaying the total for both counts and a cleaner graph. The numbers are also more accurate with more bots being detected and filtered out.

Now page views and visitors are only counted if the a page has been open for more than one second. Browsers also need to have JavaScript enabled. This will automatically eliminate a lot of bots, but most importantly it helps us get more accurate conversion numbers (which are also tracked with JavaScript).

Acquisition Channels

Instead of just showing all referrers in a list, the most important sources are now split up into: direct, referral, organic and social. Social describes all traffic coming from social networks such as Facebook, Twitter, Reddit and so on. Organic traffic comes from search engines (Google, DuckDuckGo, Bing, ..). The referral category contains all other traffic that comes from outside websites. When a visitors types your website address into their browser's address bar or visits your website via bookmark, it's counted as direct traffic.

This new acquisition channel grouping should give you a quick overview of your traffic sources. Below it you will still find a list of all referrers. On that list you can now also see all the exact URLs linking to your site when you click on a domain.

Next steps

The rest of the dashboard remained mostly the same as before in this update. Ultimately the geographical data will be displayed on a map in addition to the list. It is also planned to make it possible to filter data by event, referrer, acquisition channel or geography.

To stay in the loop about new updates, you can subscribe to product updates via the product changelog.

Bootstrapping to Ramen Profitability in a Year

Wed, 29 May 2019 07:00:00 +0000

One year ago I launched Landen on ProductHunt. Since then it has slowly become profitable enough to cover my living expenses. I'm sharing the relatively unspectacular story to provide you a glimpse behind the scenes and to hopefully inspire other entrepreneurs.

Before the idea

I've worked as a freelancer building websites for a few years before starting to specialize on data-heavy business tools. I've worked with some medium sized companies and made a decent living, but I soon realized that instead of working for clients, I could build something for myself. At the beginning, building a profitable company on my own still seemed pretty far-fetched, but then I discovered the IndieHackers community. I was inspired by all the success stories, and seeing other people who were just starting out and sharing their experiences. Not long after, I was convinced that I could do it too.

The big question: What do I build?

There were a few ideas at the start: a habit tracker, a load testing tool, a website editor, a web-design course and bunch more. Most of those ideas came from my own unhappiness with existing solutions. Ultimately, I decided to start a website editor for the following reasons:

I had some experience making websites as a freelancer
It addressed my own needs: I would need a cool product website either way
It seemed the most promising to become profitable

The third and last reason was the most important one: I want to be able to live from it. Are website builders my passion? Not really. Will this change the world, make a dent in the universe? Probably not. This is my first attempt at starting a product business, so I viewed it more as an opportunity to figure things out.

Making a "Plan"

Website editors are definitely something that has been done before, with many huge established players. The great thing about creating a product in such a market is that your idea is basically already validated: people want websites and are willing to pay for them.

Now, the only question is, how can I differentiate Landen from the many potential competitors? To be honest, I probably didn't put as much thought into that as I should have. My main idea was to focus on two things: make it easier to use than any of the alternatives, and produce higher quality outcomes (more beautiful websites that load faster).

Now, the thing is, my potential competitors have millions of dollars, hundreds of employees and much more experience. How could I possibly create something better? I decided to not focus too much on what the competition was doing. I had my own vision in my head and it seemed different enough from all other solutions.

Minimizing Risk

Since I didn't do much research beforehand, I needed to minimize potential losses in case of failure. Launching early is one of the most repeated pieces of advice in the bootstrapper community, so I decided to set myself a 100 day deadline. I was also able to keep my expenses extremenly low by building and designing everything myself and keeping the infrastructure simple.

The Prototype

With that "plan" I went ahead and started programming and designing. At this time I still had the full-time job, so I was very limited with what I could build. Even though I worked nights and weekends, it wasn't an overly stressful time. I was previously working the same hours, only then it was unpaid overtime at my job. Putting that time into my own project instead was a lot more satisfying.

To save time, there were no fancy features, I used the simplest hosting setup I could think of, and focused 90% on the core builder concept. I only used tools, languages and frameworks I was already familiar with: React, Go, and AWS. You can read more about the tech behind Landen here if you're interested.

Another thing that saved me a lot of time was not setting up a subscription system. Even though Stripe makes it realitvely easy, it's still a lot of work. Instead Landen offered a free tier and a premium tier via a one-time payment which would give customers access for one year.

During the development I also posted some questions on IndieHackers and started sharing prototypes where I received lots of valuable feedback. As it turned out, many of my fellow entrepreneurs were also potential customers, which helped me get some early test users who were genuinely interested in the product.

Launch Day

Looking back I'm not sure how I got the prototype done in 100 days, but somehow I did, and on May 29th, 2018 Landen launched on ProductHunt. I hunted Landen myself, put up a short video and some screenshots. There was an email list I started with about 40 people in it whom I emailed at the day of launch. That's it.

It made #3 product of the day. It was a pretty stressful, but on the evening of that day, I got my first paying customer - a huge milestone! Landen had about 250 signups that day and the next (most of them on the free tier).

Refining the Pricing

A while after the launch, I realized the pricing was not right. Like many first time founders, I priced Landen way too low, at $49/year with a generous free tier. There's nothing wrong with that price itself but it didn't fit the vision I had for Landen.

As I received more feedback and did more research, it looked like the ideal pricing model would be a monthly recurring subscription. I didn't bring back yearly plans until much later (another mistake). Yearly plans are a great way to raise money early on. By now, the pricing has been overhauled three times, getting more expensive (almost) every time.

Other than the pricing model, the product seemed to be on the right track. Overwhelmingly positive feedback motivated me to keep working on Landen, and I continued developing things from my own wish list and customer requests. I also wasted a lot of time.

Going off-track

There were things I knew wouldn't be important to the projects success, but they seemed fun to do. So I did them anyway and developed a basic analytics feature, integrated email lists, and kept polishing the UI.

Landen might have grown faster if I had kept my focus, but I still do the same thing today. Sounds like an easily avoidable waste of time, but I don't think it is. Starting a company is stressful, and doing those fun little detours has helped me deal with the stress and prevent me from burning out.

I was only able to work as much as I did because I enjoyed what I did. There are always some aspects which are less enjoyable, but in order to do this long term, the bulk of the work needs to be something I like doing.

Positioning

When I was done with the fun side-side-projects it was time to focus on the core roadmap again. I tried to find the right place for Landen in the website-builder market. With so many different use cases, industries, and a big range of potential customers from solo entrepreneurs to mid-size businesses, it was hard to say exactly who Landen is for.

I had set out to make a new website builder but there are lots of different kinds of websites. What should Landen do and for whom? At the beginning I just thought: the more the better. I wanted to expand it and make versions that would work for restaurants, hotels, service businesses and almost everyone else who wanted a website. For at least half a year I didn't realize what a big mistake that was, and kept expanding that list:

Then one day I was talking to my girlfriend about how starting a company poses so many questions I don't have the answers to. I was trying to articulate what I wanted it to be and said something like "I just want to make it easy for startups to make websites". That's basically how I made design decisions for Landen from the start but I never really realized that.

From then on I decided to scrap any plan to expand into other industries (even though its still tempting) and focus entirely on building something to help startups make websites. This hasn't just made it easier to plan the product development but also to market Landen. The limitation alienates a lot of potential customers, but for startups it becomes a lot more attractive.

Making realizations like that is one of the most exciting things about running your business. You get to figure things out as you go, your product becomes more defined and your vision gets clearer. Some people might have this figured out from the start but I don't think you have to.

Going Full-time

About 5 months after the launch, Landen made $1,000 MRR, and I decided to go full-time.

Wait.. what? Only $1k MRR after five months? And going full-time?! Waiting for a higher MRR and keeping the day job would've definitely been the safer bet, but my client work was starting to make me miserable. Landen gave me the freedom to live out my creativity and work on something new instead of realizing someone else's vision.

I'm lucky I started out with no dept (yay European universities), no dependents, and a safety net of friends and family. My freelancing work allowed me to save up enough money. In short, the stakes were fairly low, so I took the "risk".

It might still be too early to tell, but I think it was the right decision. Even if I had waited longer, it would've just been a matter of time before I made that transition. Making the jump sooner, freed up my mind from many "what if's".

After going full-time I kept doing more of what I was already doing: developing features and improvements. However, it also dawned on me that I'd have to start marketing soon, but I wasn't really happy with the product yet. Not having done much marketing led to pretty slow but stead growth. Most new customers came from word of mouth or from the ProductHunt launch.

It took me another half a year to feel like Landen was ready for to be shared with the world, but I had no idea how. I planned another ProductHunt launch with all the new features, and tried out paid acquisition (which spectacularly failed).

To this day, I am mostly clueless about marketing and grateful for all the amazing customers, reviewers and enthusiasts who keep telling their friends, colleagues and communities about Landen.

Ramen Profitability

A few weeks ago, almost one year after starting work on Landen, it has finally reached "Ramen Profitability", meaning it can sustain my minimum living expenses. It's been a great journey, even though it's been relatively calm. There's so much hype out there about overnight successes, startups making millions and "the next big thing" that I just wanted to share a more modest story.

Reaching that milestone gives me faith for the future of the company, even though that doesn't mean Landen will ultimately be successful by many standards. My definition of success in this case is to match my previous income with a healthy and happy customer base.

What's Next?

I will continue working on Landen full-time and try to learn as much as possible along the way. A detailed post about the future vision for Landen will follow soon.

Thanks for reading! If you have any comments, please share them over at the IndieHackers thread.

How Landen Works: The Tech Stack

Wed, 15 May 2019 03:32:58 +0000

Hosting websites for thousands of customers as a one-man operation can be a challenge. The systems have to be scalable and reliable to provide high up-times with fast responses. At the same time it, can't be too expensive or time consuming.

There's only one way to achieve that: simplicity.

I built Landen with the technologies I knew when I started: React, Go, PostgreSQL, AWS. Here's how that comes together. Note that I've included a lot of links for easy reference, I'm not affiliated with any of the services mentioned.

Front-end

Landen's front-end is a "thin" client, that means it is a JavaScript application which runs in the browser, and fetches its data from a RESTful API (the back-end). There are many ways to host such a JavaScript app. I choose Netlify to host it, for it's performance and ease-of-use.

The React app itself is built with create-react-app. Here are some interesting dependencies I'm using:

date-fns for date formatting (much smaller than moment.js)
frappe-charts a beautiful and lightweight charting library
marked for markdown
react-loadable for code-splitting (keeping the app size low)
react-select for select menus (even though the file size is huge)
react-virtualized for performance in big lists
slate for the blog editor
bugsnag-react for error reporting
fuse.js for filtering lists

When adding new dependencies I'm always very careful to keep the bundle size low. Even with minifying and compression, some dependencenies can easily 10x the size of your web-app, and thus it's load times.

In terms of UI, I didn't use any frameworks because I'm a big fan of experimenting with UIs and like to write them from scratch. That usually means things take a bit longer, but it's a rewarding experience. The only 3rd-party UI library I'm using is normalize.css.

With the React app itself I try to re-use components as much as possible, and keep complexitiy low. At the time of writing, the entire frontend contains about 25,490 lines of code (without the dependencies dependencies).

Back-end

This part of the stack is there the data is stored and processed. It's a little more complicated than the front-end and consists of 4 main parts:

API server: sends data to / and receives data from the React App.
Site servers: host the actual sites created with Landen
Certificate Server: for managing SSL certificates
The database: where important data is stored

All of the back-end servers are written in Go and hosted on AWS EC2. The database is a managed PostgreSQL instance hosted with AWS RDS. Not having to manage the database server itself is a great relief for a solo developer team.

The API server is what some would call a monolith: it handles nearly everything when it comes to data processing, handling, authentication, and whatever else the React App needs. It currently provides 99 REST-ful endpoints, which is still a very manageable amount. The API server lives behind a Load-Balancer which makes it possible to scale up, if the need ever arises. However, to this day a single server has been able to effortlessly handle the entire workflow.

The site servers are responsible for figuring out which website to show when a visitors goes to any Landen site. They are also hosted behind load-balancers to increase redundancy and handle traffic spikes. Just like the API server, there hasn't been a need yet to scale up the site servers, despite the amount of traffic they are receiving. Since they are mainly serving static content, the computing power they require is very low.

The certificate server works in the background to request SSL certificates via Let's Encrypt. There are some ways to do this more easily without a separate server, but those come with limitations. Because Landen has multiple site servers which all need the latest certificates, it's best to handle all of that in one place instead.

Finally, the database is a simple AWS RDS instance so there's not much to manage from my side. There are no NoSQL databases in Landen's stack at this time, and maybe never will be. NoSQL sure has it's place in many applications, but it's often just used because it's hip and cool while the data it stores might never exceed the limitations of a basic SQL setup. Since I need to keep things simple, I'm trying to avoid them for now.

To help manage and extend the functionality of the stack I use:

Hyperping for uptime monitoring and status pages
Postmark for transactional email sending
SendGrid for all other email
Imgix for image hosting and processing
CloudFront for other static assets
Sentry for error reporting
AWS S3 for larger data such as site files
Pusher for real-time applications

That's it! A pretty simple setup for now, let's see how long it can keep up with growth. If you're curious about any aspect of the stack in particular, or need some advice for your startup, let me know on via Twitter.

Landen 2.0 🎉

Wed, 08 May 2019 23:41:29 +0000

Finally, the day is here! It's been a bit more than a year since I started work on Landen. Today I'd like to introduce Landen 2.0 with a second launch. There have been a lot of changes to the core product and some powerful new features:

Core Product

The updates for Landen's core product: the website generator, editor and hosting have been countless. From a new load-balanced server infrastructure, over performance updates, icon & asset pickers, to new sections. If you want to find out more, have a look at the changelog.

Blogging

Content is an important part of marketing and building a brand these days. Landen has support for blogs on the same domain as your homepage, which is great for SEO. That also means that you don't need to wrangle with Wordpress setups, servers, plugins, updates, or blogging platforms with their own agendas. On Landen you own your blog. There's currently support for multiple authors, scheduling and backdating posts (if you want to move an existing blog). Some more updates are on the way: tags/categories, design editor, blog analytics, email subscriptions.

Analytics

Privacy is one of Landens core values, so it only makes sense to have a simple, privacy-focused analytics system right within Landen. Data is collected server side, on the same domain as your sites without calling any external URLs. Currently available stats include: pageviews, top sites, referrers, countries, cities, browsers, devices, and operating systems.

Email Lists

Sometimes you might just want to validate an idea without setting up a complicating mailing system. With this new feature, it's now possible to create email lists right within Landen. Then your visitors can sign up with your email capture forms, all without requiring any 3rd-party services. You can export your lists as CSV file at any time.

These improvements and new features are just the beginning. A lot more cool stuff is on the way, and you can see some of it on the roadmap (where you can also request features).

In a way, this is the first real version and 1.0 was just a beta test. If you're currently working on a product you haven't launched yet, I highly recommend you do the same: Launch once your product gets the point across and start validating your idea right away. If this were the initial launch and Landen didn't have any paying customers yet, the pressure would be overwhelming after a year of hard work.

Thank you to everyone who has used Landen in the past, paid or free. And thank you to everyone who's provided valuable feedback.