Using Data-Driven Marketing to Grow My Business to $8,000/mo
Hello! What's your background, and what are you working on?
I'm Roman Kalyakin, a software engineer and a travel addict.
In early 2015 I co-founded Flystein, a flight search service that connects travelers with air travel experts. We have a group of experts with an impressive background in getting the most out of airlines, airports, airfares, and frequent flyer programs. At Flystein they help our customers build the best and the cheapest flight itineraries.
Since we launched in 2015 we have helped almost 2,000 customers save on airfares. Recently we've averaged $8,000/month in revenue.
What motivated you to get started with Flystein?
I had been using "flight hacking" tricks to book cheap flights for myself for a couple of years before I randomly met Vlad Protasov in January 2015. Vlad, who'd traveled to Australia from Europe on a 400€ (~$450) return airfare, had been doing the same thing but on a slightly bigger scale. He had a group of happy clients that were regularly paying him in return for tailored itineraries that were not only convenient but often considerably cheaper than anything they could find on Expedia or Kayak.
After chatting for a couple of months we decided to turn the little "one man orchestra" business into a service available to anyone. I had a solid experience in building MVPs (minimum viable products) for start-ups, and I built a few tools to automate flight search. Vlad knew way more loopholes and tricks than me, and had experience selling this to clients. He also had been hanging out on frequent flyers forums and maintained a decent network of connections there.
The best thing was that the idea had already been validated by Vlad. We just needed to wrap it into a nice web application, bring together a group of "flight experts", and start getting customers. It took us some time to decide how to present it. We toyed with the idea of creating an invitation-only club for frequent flyers or building some serious automation tools, but finally decided to start with a simple pay-for-the-service model. This was a good decision because we both had full-time jobs at that time, so venturing into something that would take months to build would probably only help Flystein never see daylight. Shortly after, we decided to give it a go.
What went into building the initial product?
I badly sprained my ankle while rock climbing, and could hardly walk. So for those few weeks that I was out of sport and social life, I spent my days doing contract work and my evenings and weekends building an MVP for Flystein, going through paperwork needed to start a business, and discussing initial product features with Vlad. Vlad, meanwhile, had been recruiting our first experts.
I started with my usual stack for building MVPs: MongoDB with Node.js on the back-end for the API and batch jobs, Angular.js on front-end for the two web apps (one for clients, the other for our experts).
It took us just under three months to launch the first version of the platform.
We briefly talked about getting seed funding from incubators and angel investors, but decided that it would take us less time to build something simple first, try it out, and see if we needed more capital.
How have you attracted users and grown Flystein?
Initially Vlad sent all of his personal clients to Flystein and started a massive guerrilla marketing campaign using our personal Facebook and Twitter accounts. We got our first paid customer the day after the launch, and this covered the initial costs of setting up the business, buying domains, and hosting.
Once we had annoyed all of our friends and reached as far as we could in our own network of connections, we moved over to more traditional marketing methods. We launched a blog and a mailing list, and started reaching out to bloggers and striking partnerships with other growing businesses in the travel domain. We offered affiliate commissions and discounts in return for cross promotion.
Half a year later, one of our customers secretly submitted Flystein to ProductHunt. Our analytics recorded 17,000 visitors. This month quadrupled our revenue, and taught us a lesson to use a CDN (content delivery network) early to speed up our server's response times and reduce load. We also decided to create a pool of candidates for the position of "flight expert".
Since then, we've experimented with different marketing campaigns: monthly newsletters, Reddit AMAs, and paid ads. But we've found email retargeting and cross-promotional blog posts about Flystein from influential bloggers to be the most effective drivers of sales.
What's your business model, and how have you grown your revenue?
Our business model is a cross between a traditional travel agency business and a marketplace platform, like Uber or AirBnB. We maintain a pool of experts who know how to find cheap flights and connect them with clients who need to fly but hate paying a lot of money to the airlines. Clients pay us for this service and in return receive an personal advice and step-by-step instructions on where and how to book cheap flights. Experts receive their share of the fee clients pay us.
Before we launched we had ran the idea of a human-powered flight search platform through our friends. We already knew that our business model might be hard to understand for an average traveler: why would you pay someone to search the internet for flight tickets when anyone can search internet nowadays?
From day one we offered a guarantee: if our experts can't find a cheaper flight than the client can find (minus the fee we charge), then the service is absolutely free. This simple concept has turned out to be hard to communicate. We've experimented with wording, layout, and checkout flow to convey this message to new clients. But two years later, we still receive a lot of emails with the same question: "What will happen if you cannot find a good enough deal?"
We use Stripe as our payment gateway. It fits well into our "guarantee based" business model. To launch a flight search request we require everyone to enter credit card details. Stripe pre-authorizes the card for the amount of our fee. The authorization is then removed if we cannot fulfill our guarantee of finding a flight cheaper than what the customer has found.
We have been experimenting with offering a B2B service. This has proven to bring a good and stable revenue stream. However, winning businesses has turned out to be really hard when done in parallel with running an existing B2C business. We've realized that a dedicated sales person and a separate platform for managing corporate clients is needed.
Our pricing scheme already includes a reasonable margin. We've therefore never attempted to tweak it significantly, fearing that our revenue will fall. At the same time we've been noticing a strong correlation between revenue fluctuations and successful marketing campaigns. There is also a slightly-visible correlation between revenue and traveling seasons. However, this correlation has become less prominent as we've started acquiring clients from different countries and continents.
Our average revenue over the last 3 months has been around $8,000/month (or $6,000/month over the last 6 months).
What are your goals for the future, and how do you plan to accomplish them?
Our short-term goal is to perfect our workflow. Flystein is largely based on two things that can be quite unpredictable: human interaction and the volatility of flight prices. Often it takes a good amount of creativity to solve problems that arise from miscommunication, misunderstanding, and sudden change in ticket pricing. We've been getting there slowly.
The long-term goal is to expand further into the B2B area and get more corporate clients. From our experience, this massively reduces customer service effort and brings a more steady revenue stream. Businesses that need to fly their employees often do so throughout the year, whereas occasional travelers use Flystein only once or twice a year.
If you had to start over, what would you do differently?
When I was launching Flystein I seriously underestimated the time and effort it would take to deal with people. This includes maintaining a productive relationship with my co-founder, as well as keeping our experts and clients happy.
If I had a chance to start over I'd seriously consider finding a mentor who could mediate between me and my co-founder when we have a disagreement. Alternately, I might try to find a third co-founder to create another point of balance in the decision-making process.
Even though we now have a dedicated customer service representative, I would search for someone for this position at an earlier time. It's better to train someone for customer service when fewer actual customers have signed up and the load is lighter.
I also slightly regret the time and effort we spent on chasing and establishing partnerships that, at the end day, did not bring us any value. We had a lot of those in the beginning, but after we got more experience with it we became more selective.
I would seriously recommend anyone launching an internet-based B2C start-up without previous experience to begin by seeing if there's a way to target other businesses rather than consumers. If it is possible, go with that.
What were your biggest advantages? Was anything particularly helpful?
The biggest advantage we had from day one is that we did not need to go through the product validation stage. Having done this before on a much smaller scale, we already knew that we would be able to break even in a short period of time. This helped us launch and go through the first few months with our own effort and without external start-up capital.
We've been quite lucky with some marketing decisions. And I must say that the outcome of some of them I would not be able to identify.
The disadvantage of spending time and effort on disagreement with my co-founder in fact comes from our different approach to things which I believe is a serious advantage in itself. Vlad always takes a purely opportunistic approach to marketing, sales, and day-to-day operations: if something comes our way, suggested or imagined, it should be tried — preferably straight away.
I always tend to keep a steadier pace, evaluate, refer to numbers, verify facts, and plan things. These two incompatible approaches often result in good decisions to either give something a go or avoid wasting time.
What's your advice for indie hackers who are just starting out?
Starting a business might be overwhelming. With such little feedback and so many directions to take in the beginning, it is very easy to start implementing many features that will never be used or prove in the end to be insignificant. Start simple. Don't build a document management system until you absolutely cannot move forward with Google Docs.
Outsource as much as you can to clear your way and focus on evolving your business. It is very easy to find yourself on the slippery slope of getting buried under day-to-day operational tasks.
Where can we go to learn more?
If you have any questions don't hesitate to ask them below. Thanks for this opportunity to share our story!
