The ‘Startup Growth Series’ showcases stories of entrepreneurs that have overcome obstacles in order to expand their businesses in Canada and abroad.
The early days of a startup can feel like a constant slog. Milestones are used to track development and maintain sanity in the slow trek towards becoming a real company: MVP, launch, first customer, etc. Each milestone reached feels more like a mark of survival than success.
“The best companies are those who can run the most number of experiments.”
Then comes the $1 million in annual revenue milestone, significant because it means you’ve built something that attracts early adopters, and they are willing to pay you for it. It’s a signal to the market that you’ve figured a few things out and might be ready for the big leagues.
Furnishr, a Toronto-based company that designs and assembles ready-to-move-in rooms for relocating professionals across North America, recently hit the $1 million annual revenue milestone. Founded in 2016 and predominantly bootstrapped, the company focused on two things to get to $1 million: customer feedback loops and targeting specific growth channels.
Speaking to BetaKit, CTO Karen Lau, CEO Mike Van, and recently-departed head of growth, Vahid Jozi, explained how the company went from a Craigslist ad to a million-dollar revenue milestone.
Building a data machine
As Furnishr acquired its initial customers – mostly expats and young professionals relocating for work – the company received the feedback that they wanted ready; made room designs, but also to add their own unique touches to the design.
“They might just change one or two things in a room, but they still want to have that aspect of uniqueness and customization,” said Lau. “They don’t want their home to look exactly the same as their neighbour.”
This insight led Lau and Van to offer a design consultation call for every customer. On the call, the team could ask customers in-depth questions about where they found Furnishr, their design challenges, and what customizations they might want. But Lau said there was another goal beyond customer service, she wanted a clear point in the customer journey to gather feedback and data.
After designing and building hundreds of rooms, the Furnishr team used the insights gleaned from the design consultations to build a design quiz, which gathered information about the colours, fabrics, and materials customers liked in a room. Each completed quiz gave the design team structured data and insight to start the conversation with customers, adding more depth to the design consultation call and more insight for the Furnishr team on what customers wanted.
Building a data machine to understand what customers wanted would also help Furnishr with its next important step: targeting channels to grow its customer base.
Investing in targeted growth
Rokham Fard, founder and CEO of PsychologyCompass.com and an entrepreneur in residence (EiR) at the Ryerson DMZ, said that rapidly testing growth channels, and turning off the ones that don’t work for your business stage, is critical to success.
“The past 20 years have shown that the best companies are those who can run the most number of experiments, in a given time period,” said Fard. “That framework also extends to finding out which growth channels work best for you.”
“Stressing that high-touch customer service, we get to ask a lot of these questions.”
Fard said that the three best growth tactics a startup can perform are testing multiple different channels (prioritizing the ones where your customers live), focusing on one or two channels at a time so you don’t get stretched too thin, and identifying business process bottlenecks so you can automate as much as possible.
Furnishr started with only one growth channel – asking customers for referrals. When a customer had a good experience (Lau said the company boasts a 99 percent customer satisfaction rate), Furnishr would ask for a referral to a friend, coworker, or family member in a similar situation. This channel worked for the ultra-lean startup in the early days because it provided a direct connection to buyers, leveraged other people’s social credibility, and required no additional investment.
The next step for the team as it grew was content marketing. After listening in on hundreds of design calls, the team noticed trends in customer questions or issues, which became the topics of each blog post.
“From all the questions that we get, we see a pattern of what information people need in regards to furnishing a place,” said Lau. “A lot of them could be about what to do with a sectional sofa in a small space, or how to lay out a tiny bedroom. So that’s where we get our content topics.”
Blogs are distributed through niche Facebook groups, such as groups for expats moving to a certain city, and other specific online communities, said Jozi. He added that while common metrics like traffic and clickthrough rates are useful, the real metric of success is if customers found value in it. He said he checks the quality of Furnishr’s blogs by asking customers if Furnishr’s content addressed a question or concern they had.
Lau said that asking customers if blog posts answered their questions gave the team two benefits. One, it increased customer trust in the people behind Furnishr, which contributed to the company’s high satisfaction rate that fed their referral funnel. And two, it provided Jozi with insight on which blogs were more likely to convert new customers. Since many relocating professionals have similar questions, if a blog post solved a current customer’s challenge it was likely to be helpful in attracting new customers as well.
“Stressing that high-touch customer service, we get to ask a lot of these questions, contextually, throughout the process,” said Jozi. “The feedback loop verifies that some of [the blog posts] do work, connecting the dots on a general scale.”
But one thing you won’t see Furnishr doing, at least anytime soon, is paid ads.
“We ran paid ads in 2019 for a bit,” said Jozi. “We basically overloaded our design team.”
The team realized that if they overloaded the design team, the company would lose its ability to satisfy customers, validate the quality of their content, and know enough about the customer to make future advertising or sales partnerships worthwhile.
Bigger base, bigger experiments
Now that Furnishr has growth channels that work, the next step is to build the backend infrastructure necessary to run tests at a bigger scale.
“We realized that we need to hit a different level of scale,” said Jozi. “So that has been most of my focus in the last year to allow for this.”
The machine learning platform leverages user insights to predict what kind of custom rooms customers would want.
One major necessity for Furnishr to scale is building out its tech. The company started with phone calls and moved to a quiz. Now, it’s building a machine-learning powered interior design platform.
While the quiz streamlined data collection and provided insight, it still required a designer to predict what customers wanted and put everything together. The machine learning platform leverages user insights to predict what kind of custom rooms customers would want. This means Furnishr’s design team can handle far more customers, something that Van is excited about because it means the company can think about bigger customer growth channels.
“This is something that hasn’t actually been done in interior design a lot just because the market hasn’t reached that stage right now,” said Van. “But our system is capable of outputting designs, understanding what this customer would want and what some of the main products are that they need, and it figures out a lot of the secondary and ancillary products with it.”
Four years in, Furnishr isn’t fundamentally different from how it started. But the company built a machine to replicate, automate, and scale the service it provided its very first customer, allowing Furnishr to reach the $1 million milestone.
One million in annual revenue isn’t the stopping point for a startup, however, so to continue scaling Furnishr will need to continue building a better machine.