When you travel a lot and end up somewhere jetlagged and hungry at random hours of the night, sometimes you have to make compromises.

On a recent trip to Taipei, I found myself sober and eating a tea egg out of a plastic bag from the corner 7-Eleven at 1 in the morning. It wasn’t a Facebook moment, but my snack turned out to be surprisingly good (I severely regretted not buying more), and it was then that I began to understand the deep bond between the Taiwanese and their unglamorous yet satisfyingly functional everything-store.

When I spoke to Fiona Lau at Shopline — Batch 10 — I was surprised to learn that she and her team weren’t Taiwanese because I’d heard so much about the “great Taiwanese company” in our current batch. (It’s hard to fault my non-Asian colleagues for their childlike confusion, but really, there are enough of us now that everyone should be an expert.)

Shopline is a DIY ecommerce platform that helps offline merchants in Asia go online, with self-branded online shops that take a few minutes to set up.

The company went from 1000 registered shops and 10 paying customers at the start of Batch 10 back in August, to 6,500 shops and 140 paying customers this past week.

They grew their paying customer base by 14x in 9 weeks, and they more than doubled their conversion rate while continuing to grow their overall user base.

By the time Fiona, Ray, and Tony flew in to SFO, they had already established a presence in their home market of Hong Kong, but were also looking at growing in Taiwan.

I talked to Fiona about how a bootstrapped startup acquires users in the Asian Tigers, the one tool most companies are underutilizing for both growth and product, and why 7-Eleven might be the key to unlocking ecommerce in Taiwan.

What acquisition techniques did you rely on when you were bootstrapped, pre-Accelerator?

I actually spent a lot of time offline in smaller shopping malls, bazaars and other offline shopping districts. In Hong Kong, there are a ton of these.

We were running small experiments with Facebook, capped at $30 a day. Facebook was perfect for us because it’s so dominant in Hong Kong.

We had almost zero prior experience with marketing and paid acquisition. We were trying to place our first ad just reading the instructions off the package, in the help center.

We didn’t have a network of other startups at all, so we spent a lot of time Googling how-tos, reading blogs, using common sense and winging it.

There was no benchmark for anything we did, but somehow we managed to get those 1,000 signups before we joined 500, so it wasn’t an entirely bad start. But, we had a lot of learning to do.

Your team is from Hong Kong, which gives you home court advantage. How did you get into Taiwan?

It wasn’t our original intention to to target Taiwan. We started in Hong Kong because it made the most sense to us, but we knew it was too small.

8 million people live in Hong Kong, but penetration is still really low in terms of the number of businesses doing ecommerce, or just with a website at all.  People have been doing everything through web agencies. It’s small, but we still see lots of opportunity.

At the time, we were using Facebook to find users in HK, and we asked ourselves, where else is Facebook really dominant?

We ran some ad experiments in Taiwan, and those turned out to be really successful. Facebook is even more entrenched there, the quality of offline SMBs is very high throughout the country — many selling great products with almost no online distribution — and we can support the language too.

Now the name is out there, and we have people reaching out to us asking, Can you accept Thai Baht, Can you open this in Thailand?

What are the biggest growth takeaways from the past 9 weeks?

I know every 500 company says this, but MixPanel. Mat helped us set up funnels the right way, whereas before we were trying to track everything because we weren’t sure where to focus.

We thought it would be safer to track more stuff, but then we never knew what to do with it. Either that, or we were busy trying to look at funnels that were way too detailed.

We also did a lot of work with custom audiences on FB, a/b testing for pricing, using Intercom.io for customer relationship management, really thorough UTM tracking, and SEM because our product is so intent-driven.

But one surprising winner was live chat.

Most people think of live chat as a customer support tool. In this way, it’s being overlooked. For us, it’s also been hugely useful for feature development and user retention.

Customers will say, “I want this!” but they usually don’t know how to do product design — they’re not product managers — so we just listen, ask a lot of questions, and then translate. We’ve built our feature roadmap from live chat transcripts.

Every week, I pick and choose good customer transcripts and use cases and send them around to the team, aiming for at least 1-2 a week. Our whole team reads our live chat transcripts to understand what’s going on.

We used to think it would be annoying to show live chat to our users all the time in the admin panel that displays to our merchants, but our customers really love it!

They like to talk to us all the time. Before they never tried to talk to us because sending an email can be so painful. You never know if you’ll get a reply and when. Live chat helps us with user retention in moments when people are interacting with key parts of our product, and then we end up learning from them in those key moments.

TAKEAWAY: “Talking to your customers” once or at random feels nice, but it’s not game-changing. Instead, be persistent in collecting insights, and systematic in translating and sharing them.

It’s just good to talk to customers as much as possible. It can NEVER BE BAD. How has it shaped your roadmap directly?

Talking to our customers has been a huge time saver for a company like us trying to figure out what we’re doing, and it also taught us about 7-Eleven.

People in Taiwan like to pay for things at 7-Eleven, which PayPal doesn’t accept. It’s a specific to that locale, but it’s very important to our merchants because it’s how their customers want to do transactions.

Now, I’m not from Taiwan, but an early customer told us about Allpay.

PayPal kept rejecting people’s payments to our merchants, and we were stumped. One day a customer saw ad somewhere and reached out to us, “Have you heard of this? Allpay.”

Allpay is a payment service that’s tied to a bigger number of Taiwan’s 18 banks than PayPal is, and it lets people pay into their account from their local 7-Eleven. PayPal’s only connected to one bank in Taiwan, so Allpay has much wider adoption and is a better option for our users right now.

We looked into it after talking to that customer, and decided to integrate. Our merchants’ customers can print out the barcode and take it to their favorite 7-Eleven for payment for whatever thing they’re buying from a Shopline-powered store.

That really differentiates you from the international or less local competition.

TAKEAWAY: Early stage companies taking on established competitors need to addressvery specific pain points that bigger companies have missed. Ask customers & make friends — live chat, Qualaroo, phone interviews, 1:1 personal email.  

Batch 10 is almost over! What’s coming after Demo Day?

We love the environment and ecosystem here but you have to be where your customers are so we’re headed back to Hong Kong. We’re stretching ourselves thin with the time difference, especially for live chat!

There are so many resources here. You have people who failed, ones who succeeded, ones who are in the middle of it. You have all these other startups in similar stages. This is one of the things that’s so great about doing the Accelerator.

In the middle of program, I went back to Hong Kong to do fundraising. Even though our team is from there, all of my meetings were set up through the 500 network. I talk to people from Batch 3 and Batch 4 and it’s all the same family. They’ve introduced me to people I would never have been able to reach myself.

When we joined the accelerator, we had an intern who was so excited about it he paid his own way to SF. By the end of the program, we should have 7 people on board, and we’re paying his way back this time.