Hello! My name is Valentin Staykov, I’m based in Toronto, Canada and I’ve been working on SaaS startups for the past 4 years. Along with Haneef Ghanim and Aloke Pillai, I’ve been building Pastel for just over a year. As CEO, I handle sales, marketing, customer support, and business development.
Pastel is a feedback and communication tool for people building websites. It allows designers, developers and product managers to collect feedback from their clients and teams by leaving comments directly on top of their website without having to install anything or sign up. We have a subscription business model and currently have 3 different plans.
Set up takes seconds, only needing a URL to turn any website into an interactive canvas. Users can share that canvas with their teams and clients to get feedback, then easily export that feedback to their favourite project management tools.
You can see a demo on the ConvoPanda homepage here:
Before starting Pastel, I worked on a SaaS startup in the real estate space with Haneef. We stumbled into the space knowing very little about the industry. We made a ton of mistakes and wrong assumptions because we didn’t know our target customers as well as we should have.
When we decided to move onto something else, a priority for us was to try and build a product we would use. Being your own target customer makes things a lot easier.
So, the idea for Pastel came from our own frustrations. We’ve been building web products for some time and getting feedback has always been a pain. We’d typically see screenshots shared in emails or on Slack, with long back and forths. Feedback wasn’t contextual and often difficult to act on. We wanted to build a tool that would allow the person giving feedback to say “this button doesn’t work,” and the person acting on it to get all the information they need automatically.
Pastel went live mid-July 2017, and a week or two later, someone posted about us on Designer News. The reception was less than stellar. Many people voiced their concern that our price of $25 per month was too high. We also had a few messaging issues around our team plan pricing, which caused some confusion.
While there were some supportive voices, the overall tone in the comments was very negative.
We were initially discouraged, but then we started seeing a bunch of sign-ups coming from the Designer News post. And pretty soon, we had our first paying customer. We had four people pay us that month, and by the next month we had ten paying customers (nine of whom are still with us today).
We’ve tried a lot of different things since our first customers, many of which didn’t work as we’d hoped. My cofounder wrote up a blog about our early attempts at getting customers and what we learned. Something that made a big difference early on—and continues to help us now—is talking to our customers using a chat tool. For most of our paying customers, we had some kind of interaction on our chat tool (we currently use Intercom). Whether they asked a question or reported a bug, they were much more likely to convert when we got back to them and resolved their problem quickly.
Overall, our sales process has been relatively low-touch. We tend to have less than two or three interactions with users before they convert to paying users, and those are typically less than five minutes long each. We also have quite a few customers that start paying without ever talking to us.
This is something we’re still trying to figure out. Most recently, we’ve been testing content marketing as a way to get in front of people. We’ve also experimented with engaging people on Twitter and are currently sponsoring a design contest. All have had varying degrees of success and we’re continually reflecting and trying to optimize our efforts.
We did try direct sales a few times, but at that time, we only had an individual plan at $25 per month, so the unit economics didn’t make a ton of sense. We now have team plans starting at $99 per month, so we’ll probably give direct sales a shot again sometime soon.
When setting up our free trial, our main goal was for users to get value as quickly as possible, with minimal effort. The first thing you see when you sign up for Pastel is an input field asking for a URL. Once you paste a URL, a canvas is created and you can immediately start using Pastel. You can open the canvas or share a link to it, and immediately start getting feedback.
We knew getting people to add a new tool to their workflow would be difficult and that’s why we made Pastel look and feel simple. There’s tons of stuff going on in the background that allows us to deliver that experience, but when you use it, it just works. People can immediately understand the value without any upfront costs to set it up, and that makes it much easier to adopt.
It took over a year to get here, though. In the early days, we ran into a lot of issues with websites not working properly with Pastel. We’ve been fixing that as we go along, and now we see reports of something not working much less frequently.
We try to get on the phone with each of our customers after they’ve used Pastel for over a month. We have a conversation with them to learn more about how they use Pastel and what they’d like to see in the future. We still find surprising use cases that we haven’t thought of before. Most recently, we found out that one of our customers that does SEO consulting uses Pastel to essentially deliver a report of what their clients need to change on their website. So instead of getting feedback from clients on a website they’re building, they give feedback on an existing website.
Talking to other startup founders that are further along in the journey has been really helpful. Learning about how they did things, the mistakes they made, and what worked for them has helped us approach our problems in a smarter way.
Two books that have also helped us are Intercom’s book on marketing and Traction by Gabriel Weinberg and Justin Mares. Everyone on our team is a product person, but we’re aiming to be as methodical on marketing as we are on product. Both these books have gone a long way in helping us frame our thinking.
If you have and idea you’re thinking about, just start it. But be smart about it.
You don’t have to quit your job; there are tons of things you can do to validate an idea part-time. Talk to your target customers. Put up a landing page with a waitlist. Combine a bunch of existing tools to try and deliver the value manually. This won’t work for everything, but for the vast majority of ideas there’s a way to test them with minimal resources and time.
At the same time, knowing when to jump in with both feet and fully commit is tough. We decided to build a product for a market we knew very little about and made tons of mistakes. I learned a lot, but I would definitely do things differently if I could have another go.
You can read more about our learnings and journey on our blog over at blog.usepastel.com.