Drew Wrangles is Head of Product & Design at Taskrabbit. He began his career as an analyst in the financial and insurance industries before taking a customer insight and marketing role at Zopa, a peer-to-peer online marketplace for lending money. Drew then spent over a decade at Tripadvisor, where he led a 50-person product globalization team. In his most recent roles, he served as Head of Product at Drizly and CPO at Niche before moving to his current position at Taskrabbit.
In our conversation, Drew shares his experiences leading product localization at Tripadvisor, his learnings from working with market development experts, and some of the differences he found in expanding products in different countries. Drew also reflects on what it was like working at Drizly during the COVID-19 delivery boom and how they had to adjust their operations to keep up with exploding demand.
It’s all about starting with the basics. Begin with a high-level vision. There will be various nuances depending on which market you apply it to. For some markets, there will be a similar problem, and for others, a slightly different one. Within those markets, you’ll need to consider who is the competition? Is the problem you’re addressing being solved by somebody else? All of those things come into play.
Once you’ve created that vision and know which markets it makes sense for, start in one market but begin to look for synergies in others. Look into where there are similarities between markets, where it makes sense to enter, and where you can launch your product again with minimal differentiation. It’s fine to do some differentiation and changes, but if you need to approach it with an entirely different strategy, that makes it less of an obvious fit.
You also have to balance both the supply and the demand on both sides of the marketplace. It is important not to force your product into a market where that problem doesn’t exist, if there’s already something solving it, if it’s not a problem at all, or if there are other constraints that might impact the product launch and holistic vision.
It’s about looking at data internally and externally and doing that market analysis. At Tripadvisor, we had users globally. Our product and our destinations were all over the world. Internally, we had data from people on our site, which markets they were in, what geos they were looking at, how often they clicked, and what our online travel agent (OTA) partners were paying for those clicks. We knew how much demand we were capturing internally, so that was a good driver for us to understand a revenue baseline.
Externally, we were part of Expedia for a lot of that time, so we had information on where their international travelers were looking and searching, as well as how much money they were generating. We had flight data for every flight in and out of every airport in the world, so we knew where people were traveling to and from. We did a lot of research to map that back to our geos and understand the markets.
We’d look at Google search trends data to understand relative demand from country to country, as well as how much those clicks were worth. We did a lot of research with external partners and locals from those markets to understand how those people would travel and what their patterns were. We put all those inputs into a big model and looked at the cost of launching and the cost and complexity of language translation. How do we think about which markets to go into next? The model we build would give us a recommendation of which markets to enter, as well as the projected ROI. We paired that with the complexity and the time of launching in those markets to give us a launch plan.
When I was at Zopa 20 years ago, we had a UK-only site and we wanted to launch a new market. We went for Italy and the US and tried to go into those markets very early without any data. Neither of those were particularly successful. That was a good learning for us — there needs to be that market fit and boots on the ground. We were too small to try and go international at that stage, whereas Tripadvisor did have the volume.
More recently, I was at Drizly and Niche, which were both very US-based. Both companies had international opportunities, but at Drizly, there were legal and regulatory challenges within the alcohol space that differed by country. Niche, being an education company, requires understanding of which markets are a good fit for international expansion because each country has different education models.
At Tripadvisor, we were trying to launch in Brazil. We wanted to build a local market site and app, so the first thing we did was translate the site. We built SEM and all the basic stuff to launch into a new market, but we were finding that we weren’t getting traction. None of our metrics were changing. We weren’t growing the market for local travelers at the rate we’d expected.
When you think about attractions and recommendations on Tripadvisor, they’re all driven by Western travelers going to these markets. If you looked up restaurants and things to do in Brazil, it was all the popular, touristy things. We didn’t have much for locals there. Brazilians travel differently compared to the rest of Latin countries. It’s such a big country, so many people don’t leave Brazil. They travel a lot domestically and have vacation spots within their own country.
We did a deep dive with our market development experts. We realized that we didn’t have certain accommodations and vacation types. Brazilians stay in fazendas, which are farm hotels. We didn’t have any of these. We were learning all the different categories that we needed and the gaps we had. We worked with Experian and credit reference agencies to get a list of everyone who’d filed a license to be a restaurant, for example. We got hundreds of thousands of new restaurants and tourist locations as a result. We looked at multiple sources to make sure that they were all legit, put them on the site in the local language, and then set off to get reviews.
Reviews weren’t coming in, so we ran a campaign to change that. The biggest needle mover was partnering with the local airline, TAM, and their multi-plus points program. We partnered with them to incentivize and reward Brazillian travellers for reviewing local destinations. We launched the campaign on a Friday and got more reviews in Brazilian Portuguese the first weekend than we’d had in the entire three years prior.
At Taskrabbit, we’ve got teams from London to the west coast. And when I was at Tripadvisor, we had folks all over the world. It was hard. One of the biggest things I found helpful was to be flexible with my own time. It’s easy for the head office to set the hours and have everyone adapt to that, but I wanted my team to feel like I was adapting just as much as they were.
Our Monday morning meetings were at 6 a.m. in the metro-Boston office. I’d get in the Uber at 5:30 a.m. for the meeting, which was 6 p.m. for our team members in Asia. They weren’t staying too late, but I was getting up early. That showed that we’re all in this together. Also, it’s helpful to establish key hours so team members know when certain people are online and when they’re not. That information is important for scheduling Slacks or emails so that you’re not pinging people at all hours. I found that helpful for respecting boundaries and the work-life balance.
In-person brainstorms are great for that. If you can’t be in-person you can consider using tools like FigJam that allow for cross-functional, innovative brainstorming sessions. In market planning, if we get to the point where we have a multi-market problem, we do our individual country brainstorming, come back, and share those. What did we learn? How should we approach this for the Japanese market? How can we approach this for the Brazilian market? They’ve been brainstormed independently, but ultimately you bring them all back together.
This is also a great way to learn from other cultures as well. Everyone is very internal thinking and fixated on their own specific market. Taking the time for some divergent thinking across different markets, and then converging again toward a single perspective was valuable for us on the product side as well.
I touched on the internal or external quantitative data and how that informs what’s happening on our product, how people are using it, and where they’re dropping off. That’s informing things for us, but pairing that with qualitative data too is the key. At Tripadvisor, we were working with our team in Asia and their feedback was, “People in our markets love shopping, but our product doesn’t support their demands for shopping tourism.” From a Western perspective, we said, “We’ve got all the shops. You can see everything on the site.” They said, “Yeah, but in the west, people go shopping when they’re on vacation. In Asia, people go on a two-week trip just to shop. They’ll pick a market, go there and shop for two weeks, and then come home.”
We had qualitative research from people on the ground in markets, like China, Japan, Russia, and the Middle East. We looked at how people were behaving on the site and saw that people from those markets were spending more time on our shopping pages and landing more sessions into our shopping flow. There was definitely a signal internally, even with the poor experience we had, that we could do that as a business.
Based on that data, we were able to build a business case that if we improve the pages and have a more curated experience for specific shops in specific geos, we can increase traffic by X percent. That will then drive more revenue. That led to a change in our international strategy on the shopping side to launch a brand new shopping experience on the site in certain key markets where we would curate the shops that were most important within a region of our 30 shopping cities. We completely rebuilt the shopping experience for those markets in those geos. We saw a great uptick in engagement and revenue.
Tripadvisor is a good example because that’s where we had the most international market differentiation. Understanding the different motivations of users with regard to content was important to us. In Brazil and Latin countries, people love writing reviews as long as they know why they’re writing them. They also need to have the right motivation: to help fellow travelers rather than help a company in the US. They were very honest. The reviews were detailed and long. But, we found in other markets, that wasn’t the case.
In Japan, people wouldn’t write long reviews. We had minimum review lengths because we wanted good content, but Japanese characters also impacted that, so we had to shorten it further. Culturally, Japanese users don’t typically like to criticize other businesses and people, so the majority of the reviews were positive. That’s when we did an airplane mileage campaign in Japan as well, and it worked to a certain extent. It didn’t get to the level that we were seeing in Brazil, so we investigated with our Japan team to figure out how to get users to come back and build trust. It was all about the quality of the localization.
Turns out, in Brazil, you can make a few mistakes. They didn’t mind too much. However, when a slight translation issue occurred in Japan, it had a huge impact on the trust and credibility of our site. We were risking losing the trust of the Japanese market, and despite all our efforts to drive more usage and metrics, we weren’t getting results. In Japan, the emphasis was around feeling local, being local, and having high-quality translation to get users to come back, as well as all the basics around our reviews and content. In other markets, it was more around incentives and rewards for writing reviews. That was a big challenge for us.
It was a wild time. Drizly was one of the few companies to benefit from the pandemic because we were uniquely positioned. The demand we saw from people staying home and ordering alcohol was incredible. It accelerated our business three years into the future in terms of volume, and we weren’t ready for that.
As a leadership team, we were meeting every morning at 8 a.m. to debrief on the day before. We paused all of our product roadmaps across nine teams to manage the volume and expectations and keep the site up. The whole company was jumping onto customer service calls and helping our retailers. It was a war room every day. It was a stressful, busy time, but also a fun one. When we got a handle on the capacity, we started investing in driving the business forward. We shipped just over 80 new releases in six weeks. Teams were jumping on to help wherever they could.
That growth led to our acquisition by Uber, which was great. I was there for 18 months after we announced the acquisition. We evaluated how we could effectively integrate with Uber. What parts of our stack, technology, and knowledge base do we want to help Uber with, and vice versa? We had a very different business model in terms of how we sold. Most online delivery was store-first — you’d go into the app, pick a store, and then you’d shop from that store.
On the flip side, for alcohol, being item-first was more valuable. A user would say, “I want to buy a handle of vodka and a six-pack of Bud Light,” and then we would be able to do the matching on the backend, find all the stores, map those products within the stores, and present those to the consumer as “Here are 10 stores that can deliver you those within an hour. Here are the prices and the best delivery.” It’s a very different model for selling alcohol, which we believe was the right one. People didn’t tend to have loyalty towards any particular liquor store, they had loyalty toward brands and the alcohol itself.
We partnered with Uber on how to help them learn about the alcohol sales. What aspects of our stack and their stack would be helpful? We worked on a good integration plan, and it was great. I really enjoyed working with the Uber team and learned a lot.
LogRocket identifies friction points in the user experience so you can make informed decisions about product and design changes that must happen to hit your goals.
With LogRocket, you can understand the scope of the issues affecting your product and prioritize the changes that need to be made. LogRocket simplifies workflows by allowing Engineering, Product, UX, and Design teams to work from the same data as you, eliminating any confusion about what needs to be done.
Get your teams on the same page — try LogRocket today.
Want to get sent new PM Leadership Spotlights when they come out?
Matt Moore talks about how new technology, such as AI, is changing healthcare for the better and improving patient outcomes.
This article delves into the two predominant statistical frameworks utilized in product A/B testing: Bayesian and frequentist methods.
Luc Lam shares what innovation looks like in words versus in execution within teams, and how to manage stakeholder expectations for each one.
Invisible metrics like trust, satisfaction, perception, emotional engagement, and stickiness are essential for creating successful products.
Learn more →
