Taskrabbit CEO Ania Smith on AI automation and the future of work

Today, I’m talking with Taskrabbit CEO Ania Smith. Taskrabbit is one of the original gig worker platforms that’s really focused on work you might need done in your home; it’s been around for nearly 20 years, and you might be surprised to know that the core product has not changed all that much. Taskers can sign up to offer services like assembling furniture, mounting TVs, and helping people move, and they get to set their own prices for that work, which makes it a more of a marketplace than something like Uber.

As you’ll hear Ania say, the difference between Taskrabbit and something like Thumbtack or Angi is that Taskrabbit manages the entire interaction from end to end — it’s not just a directory but also where people can leave reviews, get customer service, and manage all their payments.

Ania describes all this as matching supply and demand, and we talked a lot about where that supply of labor comes from and what it might mean that there are more Taskers than ever right now. That feels like a bit of a recession indicator to me — that’s a lot of people looking for additional income — but Ania had a more measured view and pointed out that there are some Taskers earning a very comfortable living on the platform.

Taskrabbit is also now notably owned by Ikea, which has a lot of interest in how the platform grows and what kinds of services it can integrate into the experience of being an Ikea customer. Assembling Ikea furniture is a core Taskrabbit service, after all, and I wanted to know if there was any pressure to specialize that service or prioritize Ikea work over other companies.

But if you’ve been listening to Decoder recently, you know that the heart of this conversation was about the next wave of user interfaces — particularly, the next wave of AI assistants that can book services like Taskrabbit, DoorDash, Uber for you. Google actually gave a demo of one of its agents going through Taskrabbit’s website a couple weeks ago, and I’m very curious how all these services are thinking about having those kinds of tools step in between them and their customers.

After all, if you’re just asking a voice assistant to get someone to help you mount a TV, you’re probably not using the Taskrabbit website yourself. That’s a big change, and Ania and I talked about it for a while — you’ll hear her say Taskrabbit’s end-to-end focus is the differentiator here and that other platforms will find it hard to compete with that.

Okay: Taskrabbit CEO Ania Smith. Here we go.

This interview has been lightly edited for length and clarity.

Ania Smith, you are the CEO of Taskrabbit. Welcome to Decoder.

Thank you for having me. Excited to be here.

I’m very excited to talk to you. It feels like there’s a lot of change coming to the service economy on our phones. There’s a lot of change in the economy in general that is interesting to talk about. Let’s start at the very beginning. TaskRabbit has been around for a while. You’ve been the CEO for about five years. How would you describe Taskrabbit today? How should people think about it?

So it’s pretty similar to what it has been. We connect people who need help around the home with a vast network of highly skilled and reliable Taskers who can help you with cleaning, mounting your TV, assembling your furniture, or a range of other tasks around the home. And I think we do that pretty well.

Taskrabbit was founded in 2008. Many things have happened since then. We’ve gone from a desktop paradigm to a mobile paradigm. Competitors have shown up. There’s Fiverr, which I think of as where you go to get a cheap logo. Do you think of Taskrabbit as expansive into digital services? You’re describing hanging TVs. Is it more of a physical services platform?

Today it is, and that’s really our focus. It is our vision to become the number one marketplace for home services specifically. There’s a lot of things to get done around the home, everything from putting down mulch and cleaning your gutters to mowing your lawn. Then there’s so much to do inside the home.

There’s also a lot to do with big moments in your life, when you think about starting a family or moving. Moving is a perfect example. You have to clean your old place, clean the new place, pack, unpack, find an actual moving service, mount TVs, assemble. So, there are a lot of use cases where these services are still very much needed.

One thing that’s interesting about that market in particular is that there’s no shortage of gardening companies, moving companies, or people on Craigslist who will hang your TV for you. There are AV installation companies at the high end. Do you want the companies who are doing customer discovery on the platform or do you want individuals?

Today we definitely focus on individuals. We’re getting to a point where half of workers in the US are actually gig economy workers. That means many things, right? Not all of them are doing furniture assembly, but many are using these platforms to find additional work and to have the flexibility to transition between different stages of life. That’s what we focus on, and those are the people who we want to help.

We’re here to provide a meaningful income to folks who need it. There are a lot of people who may be between jobs or students who have time on weekends or during the summers, and there’s a lot of opportunity to make a pretty good living on Taskrabbit. An average hourly pay is close to $50 an hour, and in some markets, it’s well over $50. So, there are a lot of great opportunities to make a meaningful income for yourself and your family.

When you say meaningful income at $50 per hour, that’s not a full-time work week at $50 an hour, right? That’s piecework. How much are people making on the high end on Taskrabbit?

We have people making well over $200,000 a year. It’s been a couple years now, but there was a Tasker I spoke with who was a student at NYU. One of his friends was moving and using Taskrabbit. He had never heard of Taskrabbit, so he looked it up. This was 2022 or 2023, and he said, “Hey, maybe I can do this while I look for a job.”

So, he got smart and figured out how to mount TVs fairly quickly just by learning it on YouTube and really optimized his jobs. He could sometimes mount as many as three TVs in an hour in Manhattan. This is all he did for his first year after graduation, and he made well over $200,000 a year.

Is that sustainable? I know a lot of people in the industry who have AV installation companies, for example. They go from that to “I’m going to design you a home theater system” to “I’m going to sell you the components of that at some high margin.” That’s how you build your business. But that’s different from “I’m going to install a lot of TVs every single day for the rest of my life.” Is something like that totally sustainable on Taskrabbit?

No, but how many people have the same job for the rest of their lives? We have Taskers who’ve been here for five, six, or 10 years, which is great. Sometimes they leave, they come back, and they leave again. It just really depends on what’s going on in their lives. We want to make sure that we provide the opportunity. If people do want to come and earn an income, we’re here to help them do that. For us, that means getting as many jobs as possible so that we can make sure our Taskers have jobs that they can do.

So, $200,000 is the high end. It’s a great, shiny number. What’s the median?

I don’t think that I know the median. It really depends on the market. There’s obviously liquidity and density questions, and it also depends on the category. Many of our Taskers are not working what you would consider full time. They are doing this as a stop measure in between gigs or just in the evenings. So, we don’t really track the median earnings because everyone has such a specific and different use case. What we think about is what are your weekly targets for earning if you want to work as much as possible in a week,and how can we help you reach those targets? We find that that’s what Taskers really resonate with.

What are the biggest markets for you?

New York City is by far our biggest market, and it has been probably since we launched there [in 2011]. But we have a lot of growth in our secondary and tertiary markets. Also, we’re now in eight countries. London has become a huge market for us, which is really exciting to see. Toronto is a big market as well. Obviously, the Bay Area is still our home, and so we have really great brand equity there. LA is a big market. We’re now in hundreds and hundreds of cities across the globe. It’s fun since sometimes we get into competitions to see which city can grow faster, but they all have different trajectories.

Is the action mostly in cities?

Today, yes. As you know, we’re owned by Ikea, so we also definitely think a lot about how we can support the Ikea business and support Ikea customers who want their furniture to be assembled and delivered. In the US, we also do delivery for Ikea, and so we make sure that we can cover all Ikea stores. Most of them are not located in urban city centers, and they’re much bigger in Europe. So, a market like Germany has 50 to 60 stores, which is the same as the number of stores in the US but the US is triple or quadruple the size of Germany. It’s much more densely populated, and we want to make sure that we cover all the Ikea markets.

I want to come back to the Ikea relationship because I think that is super interesting. It’s been several years now and I’m very curious how that has developed. But I want to stay focused on the cities for one second. I lived in New York City for a long time. Now I live in the suburbs. When I moved to the suburbs, I just started buying more and more tools. It’s just a thing that happens when you move to the suburbs. I didn’t have the space to have a bunch of tools and capabilities in my apartment in New York City. Is that why there are bigger markets in the cities? People just don’t have the stuff or the skills?

I think it has to do a lot with market density and liquidity. You want to be closer to the jobs, but I would say the suburbs are our fastest growing markets overall compared to cities because there is just a lot of opportunity. When we moved to the suburbs, my husband was also very keen to buy tools and start doing work around the house, and that didn’t turn out very well for us. Maybe it turns out well for others, but we still have Taskers here quite often because there’s just a lot of work to do. You have more square footage, and by definition, every single additional square foot will require more work. So, we see a lot of potential in the suburbs as well.

One of the things that is interesting to me about the broader economy, as you’re describing it, is the people who want to fill their time with work. They want to make more money and see the opportunity and that liquidity in the market. But that supply of effort requires time, right? People’s time isn’t being filled with their full-time jobs. They’re between jobs or they might be out of a job. Do you see a correlation between the health of the overall economy and the Tasker supply?

We do. For around the last three years, we’ve definitely seen more supply than we can handle on the platform. We now have waitlists in many cities because we don’t want to onboard a Tasker and then not be able to provide them with jobs. It’s sort of a false premise. I want to make sure that if you are on our platform, there’s work for you to do. As the economy has changed over the last few years — especially over the last couple of years as we’ve seen a bit more struggle — we’ve had thousands and thousands of Taskers applying every year. It’s 15 to 20 percent more than the year before.

Does that feel like a recession indicator to you?

I’m not an economist, and I read all of this stuff just like you do. If you do read that stuff, we should’ve had a recession last year and one the year before, and those predictions turn out to be wrong. So, I think the word is “uncertainty.” We just don’t know because at least over the last two years, we understood what was going on a little bit more.

There’s just a lot of uncertainty overall. It’s very hard to predict what’s going to happen with tariffs, with bond yields, with interest rates, with the housing market. All I can say is we see an increase in Tasker applications. I’m not sure that that’s an indication that a recession is coming.

Do you see an increase in the number of people who are buying services from Taskrabbit? Because that’s the other side of the equation. Is demand also going up?

Demand is going up, and that’s great to see. There’s a bit of a caveat to that, and it’s not just Taskrabbit. There are, as you mentioned, multiple players. For all of us, demand is going up. We’re still such a small portion of the overall market and its overall possibilities. Most people still assemble their own furniture, put up their own TVs, and clean their own houses. Even though we see a lot of growth, there’s still so much opportunity to continue growing. Even after 10 to 15 years, it still feels like the market is quite nascent and services are still hard to figure out. There’s a lot to think about, especially when it comes to how we match the perfect Tasker to the client.

I talk a lot about mounting. We have thousands of Taskers who can mount your TV, which is great, but can they be available on a Friday evening, mount it on a brick wall, and know how to hide the cables? That’s a different skill than mounting a smaller TV on a normal wall. There’s a lot of variables that go into an actual perfect job, and we still have a lot of work to do there. So, there’s still a lot of growth opportunity.

I love that we’re talking about mounting TVs so much because that’s totally my wheelhouse, and The Verge audience knows that. I love talking about mounting TVs. This is great. We’re just in the strike zone for me.

How do you get that data? How do you evaluate who is good at mounting a TV above a fireplace versus just on a regular piece of drywall?

There are many ways. First of all, we now have AI models and tools that help us do this even faster, but the Taskers themselves tell us what capabilities they have, and we ask pretty structured questions and they can actually explain. But the bigger piece of data comes from their experience on the platform and having done similar tasks previously.

So if they tell us that they hung up this type of TV, we can use that data to match them to exactly that kind of job again. Of course, the clients provide reviews, so we can search automatically just with words, and there’s then structured data and understanding their skill set. We take that all into consideration as we build out a more optimal match.

There’s a relationship there between what customers are saying, the reviews they’re leaving, whether they validate the skills, and the rates that Taskers can charge. Is that a correlation you can measure?

Taskers do set their own rates, which makes us very different from many of the other platforms that you see in the market. It’s why so many Taskers actually love coming to Taskrabbit. If I had to summarize, they fully recognize how to set their rates, so we provide a lot of guidance to them. We take into consideration where you are based, what types of tasks you perform, your previous reviews. And then based on what the market is doing, we have a rate we could potentially suggest. They don’t take our suggestions all the time, which is fine. They get to set their own rate.

To bring it back to TV mounting — again, my favorite topic — you could pay less money for someone with fewer reviews and take on more risk that they might not know what they’re doing. Or you might pay more money for someone who appears to know what they’re doing.

That feels like a pricing strategy that should exist on the platform. Whenever we talk about the gig economy, I get workers emailing us saying, “I just want to know how to make more money.” It feels like on Taskrabbit, the best way to do it is to just charge more because they can set the rates.

Of course, they can set the rates, but if they charge too much, the market is not going to reward them in a way they might want. So, you can definitely set your rates higher than what the market is willing to bear. If you’re in a market where there are thousands of other Taskers, that’ll manifest itself on the platform by having you essentially get fewer jobs if you’re pricing much higher than what clients are willing to pay.

Do you see that rates go up over time for an individual Tasker or in aggregate?

Rates are going up in aggregate. They definitely have ebbed and flowed. It also matters a lot on the markets. Certain markets are obviously much higher than other markets. As we see trends with inflation and what’s going on externally, the rates tend to follow because even supplies are becoming more expensive for Taskers. Then, their lives are becoming more expensive, so they need to charge more as well. So we do see, in aggregate, rates moving up.

Do you know the categories where rates have gone up the highest?

It really depends. Moving is one of our highest categories in terms of rates just because it’s a more complex product. You can take hours to help someone move. In terms of total job number, that’s obviously one of the higher ones. If you look at per hour numbers, it really varies by market. There isn’t a single trend.

Then in aggregate, the rates across the platform are going up as well?

If I’m an individual Tasker there for five years, is it generally true that I can just charge more over time because I have more reviews and more data to back up that I know what I’m doing?

I don’t think that’s generally true because if you have a lot more supply in that market, you’re competing all the time. You can’t just assume that you’re going to be able to raise your rates over time. When people start up, it’s a bit similar to Airbnb. When you’re a brand new host, you come in and you have no reviews and no ratings. You want people to stay in your home and you want to do a good job, so you may price a bit lower just to make sure that you show up in search results and that people book your place.

As you gain more experience, understand what your guests want, provide those things for them, and you continue to get really great reviews and really great feedback, you’re able to raise the prices. But there’s a limit as more and more Airbnbs show up in that market. So, there are many components that help you figure out how to price, and Airbnb, just like Taskrabbit, also helps hosts figure out what the right pricing schematic could be. But it’s definitely not just a science.

Airbnb is a perfect example because it lets people set rates and it has all kinds of pricing tools. The other end is Uber, which doesn’t let anyone set a price. If you’re the best Uber driver or the worst, it kind of doesn’t matter. Uber is just pricing for you. That dynamic has led to a lot of frustration over there.

It seems like the challenge in your case is that if you get good enough, get famous enough, and you’ve had enough customers, you might leave the platform. You can just do more effective marketing and take a higher margin on your own. How do you make sure you don’t have that graduation problem?

I mean you could, but it’s not just about marketing. That’s one of the biggest services that we provide, but it’s also about scheduling. It’s also about providing some level of protection so if things go wrong, we’re there to help, and to provide consistent work. So yes, you can leave the platform and try to do this on your own. We hope that we provide such great service to our Taskers that it’s harder for them to manage outside of the platform.

For example, if you’re working on multiple platforms or doing it on your own, you’re managing your own Google Calendar. It’s really hard to sync them all and think about how to manage your time appropriately. We can help do that for you. We can help figure it out and make sure that the tasks we’re providing are the ones most matched to your skill set because we understand your history and we understand what the client wants. We can make sure that the tasks that are visible to you are the ones that are closer to your home.

So, there are a lot of additional services and benefits that we provide to the Taskers where they want to stay on a platform because it’s easier for them. Also, if they set a rate of $100 per hour, they get to take home $100, which is very different from some of the other platforms.

Walk me through that. What’s the revenue model for you? How do you make your money?

We make our money just like many platforms do, where we have to charge either one or both sides of the marketplace for the services we provide. I just explained some of the services for Taskers. We also provide a lot of services to the clients. The way things work on our platform for clients is when they book a job, they pay an additional service fee to Taskrabbit for everything from quality assurance and fraud to just being able to find the Tasker.

You’re saying these people have been vetted. They actually know how to hang the TV.

There’s a lot of temptation in every marketplace like this to charge both sides. Why not charge the Taskers?

To be fair, we haven’t tried this. We haven’t experimented with this at least since I’ve been here. I believe this has been tested previously. I think that the Taskers themselves are keen to provide the best service that they can, and we want to make sure that we are a supply-based marketplace. As you well know, this thing doesn’t work without supply, and you kind of have to start with supply.

For me and for Taskrabbit, it’s really important that we have the best quality supply, and we don’t need to be charging the Taskers for that. We don’t use a lead gen model. We are instead making sure that the entire job gets booked from the very beginning. We’re there to hold it together with the client and the Tasker all the way through to payment as opposed to just giving a client a few names and saying, “Go try them out.” So, we believe that the service we’re providing is ensuring that the job happens and it’s high quality. For that, we primarily charge the client.

You’ve made a distinction here between what you call lead gen platforms where they charge for demand. You’re a plumber, you want to get some customers, you would charge some service to get you some customers. Describe what you mean by lead gen and how Taskrabbit is different.

So I’ll just talk about what Taskrabbit does and then perhaps compare it a bit to what other platforms do. Taskrabbit allows you to book a Tasker on a platform. I guess one word to think about what we do is “shepherd” this process all the way through. So, we make sure that the Tasker shows up, that the task actually gets done, and then we collect payment from you and the client. We then distribute that payment to the Tasker, make sure that you can write the review for the Tasker, and so on.

There’s just a whole slew of things that happen, including the fact that if something goes wrong, you can pick up the phone and call or you can send us a message and say, “Hey, the Tasker’s late,” or, “The Tasker canceled, what should I do next?” That’s different than just getting the names of three or four potential providers and forcing you as a client to call each of them and start negotiating on a price or the time before the person comes and does whatever task. If something goes wrong, it’s really just between you and that provider. With us, we’re there all the way through until the job closes.

When you say you’re not a lead gen model, that feels really important. That’s a principle that you need to hold onto because every other company that’s similarly situated does sort of charge both sides. Substack, for example, takes 10 percent of everyone’s earnings just to send emails. It would tell you it does a lot more than just send emails,

But it has a graduation problem. People leave the platform because you can get cheaper email distribution elsewhere. Substack has to build new services in order to justify its 10 percent for its highest-volume writers. You don’t have that problem. Does that pressure just not exist or does that pressure exist and you keep it away?

Of course it exists. So Taskers definitely leave the platform, and clients leave the platform.

Sorry, I meant the pressure to charge both sides of the marketplace.

No because again, I really feel that we don’t need to charge the Taskers to make the money that we make. We provide enough services to both sides, but the client is willing to bear that cost and that has worked for us. It doesn’t mean that it’s always going to work and that our models and the way we match are always going to be the same. We’re constantly experimenting. We are looking at other ways with many of the components that we’re talking about because we have to continue evolving. But right now, this works for us. So today, we don’t see changing the model to charge Taskers.

I think that brings me to Ikea. So Ikea bought Taskrabbit in 2017. It’s a complicated transaction, but the upshot is Ikea now owns Taskrabbit. Do you think it’s that ownership that protects you from the pressure to increase revenue by charging both sides of the marketplace? I feel like if you were a public company you might feel that pressure much more keenly.

It may be. I don’t know. I’ve never been a CEO of a public company. I can tell you that we have pretty big goals and high pressure from Ikea. Obviously, there is a component of stability that comes with having Ikea as a parent, and there are good and bad things about this. The good thing is we are not a public company nor VC-funded, and I don’t have to worry about how things are going to evolve financially in the same way.

But that can make you a bit more complacent. So we’re trying hard to ensure that we have very ambitious goals and a very big vision so that the teams work really hard to continue to grow. There are many ways that we can drive the business forward, actually build growth capabilities. Charging Taskers is not one of them, but there are many other ways where we think we have a lot of opportunity in terms of growth.

Let me ask you the big Decoder question here, which is expressed in two ways. First, how is Taskrabbit structured inside of Ikea? How does that work?

Well, we operate very independently inside of Ikea. I report to a board and the board is made up of Ikea executives and external board members. We have three board meetings a year because Ikea works in tertials, not quarters.

That’s very Ikea to make up its own complicated Swedish word. It’s perfect.

It’s so Ikea, but it works for me. So, we meet three times a year, we set targets together, and that’s it. Now, I do work very closely with Ikea in a different way — more in a business sense — where there was a lot of opportunity to ensure that Ikea customers have access to Taskrabbit. In some sense, Taskrabbit customers have access to Ikea, but Ikea has millions of people visiting its apps, its sites, and, of course, its big stores. We want to make sure that we have access to those customers in the best way possible.

So, we’ve done some great things there. Just recently in the US — but we had done so elsewhere before — we made it so that it’s much more seamless for Ikea customers to purchase, let’s say, a desk and assembly all at the same time and pay for it at Ikea checkout versus having to go separately to the Taskrabbit app, try to put in the very fun Ikea furniture names, and then try to figure out how to get it assembled. That’s a great example of how the two companies can really leverage their capabilities.

At the end of the day, Ikea purchased Taskrabbit under the premise that there are some potential Ikea customers who don’t want to be Ikea customers because they feel overwhelmed by the idea of putting together furniture. We want to be there to help to solve that problem, and I think we’ve done a very good job there.

Let me connect that to the supply conversation we were just having. I’ve put together a lot of Ikea furniture in my day. I feel like I’m pretty good at it. That’s a relatively specialized skill. Reading an Ikea instruction manual is like learning another language in many ways.

I think it depends who you ask. Some people are really good at this.

By the way, one of my dream stories is that we go to a conference of flat-pack furniture designers and the Ikea people come in dressed all in black and everyone’s like, “It’s the Ikea designers.” It’s obviously the easiest to put together, but it’s like a whole ecosystem.

It’s a whole thing. But I can tell you, watching our Taskers put together Ikea furniture, it is impressive. They know the furniture, so they can do this so well inside and out. They just do it with such speed. You may be good, but you’re never going to be an Ikea Tasker who is doing this day in and day out.

That’s kind of what I’m getting at. It’s a specialized skill. I would actually say putting together Ikea furniture is different than putting together the furniture you would buy at Target or Walmart. They’re different companies with different styles. The way that things fit together is literally different.

Again, my dream story is that we find rival flat-pack design gangs and put them head to head. I don’t know if that’s how it works, it’s just how it works in my brain. That’s a specialized supply. Do you keep that pool aside and say, “These are the people who are best at assembling Ikea kitchens and we’re going to make sure they’re available to Ikea”?

They get to choose. I would say the majority of the Taskers who assemble Ikea furniture are also really good at assembling furniture from Wayfair, Target, and other places. They want to assemble furniture. The way it works on our platform is there’s an Ikea assembly furniture category because, like you said, it’s specialized, but then there’s just furniture assembly where it could be a piece from anywhere. Those Taskers definitely cross.

Many of them are also handymen, so they will do minor home repairs, like if your doorknob is loose. They have great skills. Some of them want to just focus only on Ikea furniture, but I would say that’s the minority. Many of them cross into multiple categories.

Because Ikea is a parent, do you reserve supply or is that just not a problem?

It’s just not a problem in all the countries that we serve. I alluded to this earlier, but Ikea is much bigger in Europe, so a larger portion of our Taskers know how to assemble Ikea furniture in Europe versus the US because we get a lot more Ikea jobs in Europe. The Ikea business is just much stronger over there. So no, we don’t keep them separate. We let them choose whatever category they want. We do not have any challenges with finding supply though, even in Europe.

You mentioned that Ikea has its own pressures on you and its own goals for Taskrabbit. What are those goals?

The goals are always about growth. How do we continue to grow faster and how do we continue to evolve where we’re solving the right problems for the overall customer and the Ikea customer? We just recently acquired a delivery company here in the US that does delivery of big and bulky things so that we can help Ikea deliver furniture to your home.

It wants to make sure that we are able to expand to other countries. We’re only in eight countries, while Ikea is in dozens of countries, so there’s still a lot of opportunity for growth that way. So, there are definitely pressures, but they’re similar to what any other parent or investor would have.

And then there’s Taskrabbit itself. How many people is Taskrabbit?

Right now, we’re nearly 500 folks.

So, we are structured like many other companies. Today, we’re very functionally focused with a matrix overlay. We have [general managers] looking over the US market and GMs looking over the European market and each country. There’s always a bit of tension… or I would say maybe not a bit. I would like there to be a bit, there’s probably a lot of tension between some of the needs that are unique per market or per country versus having fully functional ways of thinking about the problems.

As an example, there are different competitive pressures in different markets, and there are different ways that customers go onto platforms and book their tasks. There are specific needs that customers in France have that are not applicable to the market in the US. How do we make that tradeoff? France is a much smaller overall portion of our business than the US, but you have to continue growing across the world. So, there’s definitely healthy tension across there.

Put that into practice. What Taskrabbit makes most of all is a software product. You say that customers in France have needs that customers in the US don’t have. What are those needs and how do they get expressed in the product?

For example, there is some local tax law in France that allows customers who do work on their house to get some tax benefit, which is a great thing. This isn’t something we have in the US, and so we want to make sure that customers have access to that. But to do that, you have to build out a different flow and allow them to be able to submit whatever tax documents that say that this is the work they’ve done. There’s dozens of these examples across markets.

Marketing campaigns also have to be different because people are different in every market, and they resonate with different types of themes and creative. How do we go broad but also have local capabilities? That ends up causing tension across the world, and that’s what makes it fun. Certainly we don’t have it figured out yet.

My joke on Decoder is if you tell me the structure of your company, I can tell you 80 percent of the problems. If you tell me you’re functional with a hybrid overlay, I can definitely tell you 80 percent of the problems.

Here’s the thing: I’ve gone through that a lot, but there is no perfect structure. If there was one, then we would all be doing that, and we wouldn’t have to have this conversation. If you have a functional structure, it creates certain sets of problems. You can sort of fix them by going to a GM or a business unit structure. Then, unsurprisingly, you create other sets of problems. So, to me, it is not so much about the structure. It’s much more about the skills and capabilities that people have, the incentives, and how they are uniquely positioned to work across boundaries, functions, or geographies. That’s much harder to do.

It really comes back to culture. What are our values and how do we get people to really think and row in the same direction, whatever the strategy may be? Those are the bigger problems, and I don’t think that changing the structure gets you closer to an answer. I think these types of adaptive problems are more difficult to solve and are actually more challenging. I don’t think changing structure is the answer.

Maybe one of the hottest quiet debates on the show over the past few months is whether or not structure is a proxy for culture. You seem to be very much in the line that it’s not. We’ve had some opposite answers, but it feels like the split to me is whether you have multiple kinds of businesses you’re running or you have one core product.

Once you end up in the multiple lines of business, the structure starts to reflect very different kinds of tensions and tradeoffs. Do you think that Taskrabbit has just one core business, or do you have multiple lines of business that you’re trying to operate?

Historically, it’s been one core business, but it’s definitely separate from the delivery piece. There are many different things that we need to think about. We need to think about A to B, which we didn’t have to think about before. We need to think about how to locate all sorts of components that are different from our core business. Still, it’s really important to us. We think it’s a great opportunity that is very connected to the moving business.

We are thinking a lot about what that means. Do we have a separate moving business business unit and what does that look like? Again, at the end of the day, it’s not so much about structure. It’s much more about the culture, how we can feel a sense of shared responsibility across the company, understanding what our priorities are, and what our vision and strategy are. Is it all aligned? Are we all able to row in the same direction? Is it very clear? Have we set the right expectations? These, again, are things that I think are more difficult to answer.

The other big question to ask everybody on Decoder is about decisions. How do you make decisions? What’s your framework?

I know there are many different answers here. I think there are different ways to describe them — one-way doors are decisions that I’m sure you’ve heard of — but the things that are really hard to reverse. So, closing down the office during COVID, reopening an office, or acquiring a company.

These are bigger decisions, but I really trust my team and then think a lot about how they are empowering their teams to make their own decisions. I spend a lot of my time with customers. I think it’s really important, whether it’s having Taskers at my house or spending a lot of time interviewing Taskers or clients. But day to day, I’m not as close to the problem as the people who I want to have the authority or the agency to solve those problems. They are hopefully spending much more time with the customer than even I am and understanding that specific problem much better than my team and I do.

For me, it’s really important to empower the team to make decisions. You have to test and try. The thing that we keep focusing on is making decisions quickly because the speed is what gets you behind. You can constantly be looking for more data and more certainty, and most of that doesn’t exist. You have to start using your gut — your instinct or however you define that — to look at your past experience. You have some pattern recognition. You kind of understand how things may go, you weigh them, and you make a call. Doing that quicker is the most important thing.

You just described being close to the customer. It’s interesting. Your history at Uber is as the head of courier operations, right?

You were on the supply side working with the courier.

Same with Airbnb. I definitely have a soft spot for the supply side of the business. Frankly, this is why I am doing the gig economy. When I was younger, my family and I emigrated from Poland. In the ‘80s, my parents were trying to find jobs with more hours. It was all hourly work, and it was hard to make ends meet. I often wish that these types of platforms existed then because no matter what, you can hustle and make more money. I definitely can relate to the supply side of the business a lot more.

How do you make sure that the Taskers are happy? If everyone is doing this kind of work because they’re not making enough money, there’s sort of an inherent unhappiness there. How do you make sure that they are happy and enthusiastic about doing this instead of having to do it because the economy is in turmoil?

I don’t know if the premise that Taskers are unhappy is right.

I think the premise that supply is increasing because the economy is in turmoil is kind of bearing itself out.

Correct. But I’ll tell you, I have Taskers showing up at my house every month or week, and these are people who are really striving to make an incremental income or their main income. They love having the opportunity to use their skills to do something or to help someone else while making a pretty good wage, generally speaking. Like I’ve said, we’ve done a lot of things. We try to provide a lot of additional benefits for them. We try to help them and make sure that they can call us.

We have a whole team that is focused just on helping Taskers make money. This team will help you think about how you set prices, think about the categories you’re in, how you can make sure that your write-up is correct, and so on. We’re providing a lot of AI tools for Taskers to do that as well, but having that human interaction really helps. So, if you’re here on our platform and you want to make money, there’s an opportunity to make money. We want to be the people who help them do that.

I’m very curious about what you have Taskers at your house to do every couple of weeks. How many TVs are you hanging up Ania?

[Laughs] No, it’s not TVs. Don’t tell my husband this because he still pretends that he can do these things, but he really doesn’t. He doesn’t notice that it’s broken and he doesn’t notice that it’s now fixed. The house just runs. This is okay.

I had someone come in here the other day and mount a bunch of pictures in my son’s room. I had someone come in to assemble furniture, like these bookcases behind me. I had someone come to help with my lawn. There’s an infinite number of ways to get someone to help you in your home.

What have you learned about your own platform that you wanted to improve by using it as much as you do?

So much. Make it simpler. Always make it simpler. Make it easier. I may not care to choose who the Tasker is. I just want to make sure that it’s going to be a person who can do the job really well, and I want to do it as affordably as I can. So sometimes, it would be appreciated to have a bit more help in choosing. So, we’re working on that.

You’ve mentioned AI several times. I’ve been asking basically everyone what happens when you have agents and more capable assistants built into your phones. I keep calling this the “DoorDash problem,” but you could call it the “Taskrabbit problem” or the “Uber problem.” The idea that I could just tell Siri, “I need to get a TV hung up,” and then Siri goes off, clicks around the apps for me, and brings a Tasker to my house is very powerful.

There are a lot of companies attempting to build their way towards that goal. Google just announced a bunch of stuff, Microsoft just announced a bunch of stuff. That would disintermediate you, right? At the end of the day, you would just become a commoditized service provider to an assistant that might be collecting all the user interaction and own the customer relationship. Have you looked forward to that problem? Do you know how you want to address it?

I absolutely look forward to the problem. I look at it a little bit differently. If those agents exist, you have to go to where the customer is. We have people that come to our sites every day, but if people want to go to Ikea and book a Tasker via Ikea, we’re not going to say no. If they want to go to Target and do that, we’re not going to say no. If they want to go to Siri and do that, we’re not going to say no. It’s very easy to fall into thinking, “I don’t own the customer” and “I don’t know what’s going to happen with the customer.”

But the customer will know that this is a Tasker, or we can make sure that they know that this is a Tasker. They can understand that they’re working with another party, and they can come back directly or they don’t have to. Then, there will be some economic way to figure out a system where it works for both parties. Siri, in this case, wants to be able to provide these types of services, and they can’t really do it without Taskrabbit because only Taskrabbit actually has a network of thousands of Taskers. We cultivated that network. We know who they are and we understand their skills. They’ve all been background checked.

Siri, Apple, or whoever is not going to do that. They’re just not going to be able to do that. We will always have the strength and supply, and we will continue to go where the customer is to make sure as many of them have access to our platform as possible.

Isn’t that the story of the internet though? Isn’t it the companies and the platforms that have the strength and demand that get to set the terms?

I think it’s a bit different when you look at a gig economy marketplace. At the end of the day, it takes a lot of work to bring together a set of Taskers or hosts or Uber drivers. It’s really hard to acquire these folks and to make sure that they provide the quality that you want, and I don’t see other companies being able to do that.

I also think it’s too early. There are so many ways to test this and how to make it work. The example that you provided, we’re still not quite there. There are many situations where it sounds great, but there are many that are still hindered from actually being a seamless experience. There’s payments, logging in, accounts, and all sorts of ways that all of this can and will be resolved. But that gives us a chance to test a lot of different models and understand what works best for the Tasker, for the client, and in this case, for the two parties involved.

I’m with you that it hasn’t been proven to work, but I also go to the events and watch Amazon executives demonstrate the next version of Alexa and it books an appointment on Thumbtack when the dishwasher is broken. I can see how you want Taskrabbit in that mix, but then you are next to Thumbtack and the agent might just pick the lowest price, especially if you’re charging the customer. You might just pick the service you were told about.

Look, we’re very different from Thumbtack because Thumbtack is a lead gen product that is very hard to close, whereas we provide the experience all the way through. In the demo that you described, the client still has to go and work with whoever the end provider is. In this case, it’s someone to fix your dishwasher. This isn’t something that we have to do. Our product would look very different because we are able to shepherd the entire experience through.

You close the loop. You actually send the task to the Tasker.

We close the loop, but again, we’re still quite far from where that could go. I think the possibilities are endless, and we have to go where the customer is. If everyone all of a sudden goes and talks to the phone and that’s how they order things or talk to Siri, I don’t know how we would not be there.

Do you think it’s different because you are not a lead gen model? I think about Thumbtack, for example. At some point, it’s just there to deliver demand to a bunch of HVAC technicians or whoever was in that demonstration. Maybe Siri delivers that demand, and Thumbtack gets totally disintermediated. You’re saying that’s not your model and you’re not worried about it, and you’re charging the customer anyway. So would you just charge a higher rate for an agent to solve the problem?

There are so many different economic structures that can work here. We’re in the very early stages. We may all think and guesstimate that this is the right structure, and then at year end we can decide, “Wow, this doesn’t work for us, the third party, or the client. We’re not getting enough people to actually do this.” There’s a lot of trial and error here, and it’s too early to be saying, “This is the right model and this is the right structure.”

What are the most promising early flashes you’ve seen?

I think the concept of having an agent do things for you is very interesting. If you take it a lot further, eventually in theory, we won’t even have laptops, our Kindles, or whatever. We might just have agents constantly doing things for us on other types of devices. I think it’s super exciting to be a part of that journey, but I think none of it has been proven and we’re still actually pretty far away. Although technology is changing so quickly. It’s quite remarkable to watch. It’s quite remarkable to be alive during this time.

I go to all the demos and I have the same reaction as you, which is, “Does this actually work?” They’re very convincing demos.

They are. I was just at a conference where someone was adamantly telling me that it’ll only be three years before a robot can hang up a TV.

I don’t believe that for a second.

I don’t believe that. I just don’t.

I have hung up a lot of TVs.

This person was very adamant. Anytime that I’m hearing about any work that is being done in a physical space, it’s still really far away. They’re working on having robots try to empty the dishwasher, and that’s a complex task. It’s really complex when you really think about it. There’s a lot that still has to happen.

One of the things that I’m thinking about a lot is where does automation come and where does it go? Taskrabbit doesn’t seem easily automatable from the jump, as you’re describing it. A lot of these tasks are really, really hard for robotics to begin to try to solve, like hanging a TV or emptying the dishwasher. Many rounds of innovation have to happen to make that viable.

But for other similarly situated competitors or peers you have like Uber, you can see how automation will change their business. The cars will drive themselves. That’s going to change the business in radical ways, and we’ll see how that goes. Do you see any automation that might change how Taskrabbit works in the near future?

From a software platform perspective? One hundred percent. But in the context of what happens in someone’s home? I really don’t see a robot mowing the lawn just yet. But it will happen.

We just reviewed a bunch of robotic lawnmowers that you can buy and they don’t work very well.

Right. They don’t work very well. The internet didn’t work very well whenever it started. The cellphones we had didn’t work very well. So, it will evolve, but I think we’re still quite far away.

On the software side, what do you see that’s going to change?

For Taskrabbit, how we optimize the match between the client and Tasker is going to change. We can get so much smarter. I think how clients will come to our platform will change. Just the basics of how SEO and SEM work. Those terms will barely apply [in the future], and that will happen much sooner than a robot hanging up your TV. I think that pace of that change is much more accelerated.

Do you think that’s where the growth is going to come from?

When you talk to one of the car companies, they think all their growth is coming from the cars driving themselves. They can see the massive margins that come from getting rid of the drivers. But you don’t have that opportunity.

Not yet. I understand their point. It’s pretty remarkable to take a Waymo in San Francisco. I was at Walmart many years ago, and Walmart’s offices are right across the street from YouTube. After YouTube became part of Google, there were a lot of self-driving cars — they were not called Waymos then — driving around, and they looked so foreign. Obviously, they had [test] drivers in them. And now look at it. So, this will happen. I understand the excitement. I just think we’re further away from that in home services.

Have you had any conversations with the big AI companies about being a provider to their would-be agent platforms?

We are continually exploring our options.

That’s the only non-answer you’ve given so far and I appreciate it. Does Ikea have a plan to build some great agent? Ikea has had many software platforms of its own over the years.

I’m not sure, so I don’t want to put words in their mouth. I know that AI is a critical investment and that it’s looking at it in all parts of the business, but I wouldn’t know exactly where it sees the most innovation.

When I talked to Dara Khosrowshahi from Uber about what I keep calling the “DoorDash problem,” he said, “Look, at first we’re going to build it and we’ll see if it works. If it’s cool and if it works, we’re going to charge high rates to make our business work, or maybe we’ll come to some other arrangement.” That was basically his point of view.

I think that’s a great answer.

So, you’re tracking there? Let’s just see if this is cool?

Absolutely. Let’s make sure that we’re involved. I think it is the right answer. We were part of the demo with Google last week. We’re definitely doing more than dipping our toes into it.

There’s a lot of ways to build those integrations. I went to Google I/O and I watched what it called Project Mariner.

That’s right. We’re part of that.

It’s running Chrome on a data center and AI is literally clicking around the website.

That’s right, and you can see it. I would say that’s still somewhat rudimentary. I think you can do better. But again, it’s just a first step. It’s okay.

That to me seems like a Rube Goldberg machine. APIs exist. We know how to make computers and databases talk to each other. This is fully ridiculous. Then, there are new age ways of doing it, like Anthropic’s Model Context Protocol, where the agents have a more structured way of communicating. It’s kind of just APIs, but they’re cooler APIs.

Which of those ways do you think is going to be the future for you? Are you just trying them all?

I think whatever is easiest for the customer is going to be the future.

Who is the customer here?

At the end of the day, it’s the people who are paying for these services. In our instance, it’s the clients who need things fixed around the home. I think whatever is easiest for them is eventually what’s going to work. I know Project Mariner kind of seems odd. You’re just watching this mouse move around.

It is. But at the end of the day, it’s just the first step. I think Google is going to continue to evolve there. I don’t want to say which way is the right way because eventually, all of these platforms will build products that will make it much more seamless for the customer to get things done. We just want to make sure we’re involved.

Ania, this is great. What’s next for Taskrabbit? What should people be looking for?

We have some very fun things coming up. Ikea is a partner and has been a very successful partner. We believe we can have other very successful partners. We just rolled out a suite of products that will allow us to speak to those partners via an API much faster to ensure that their customers can come and use Taskrabbit services. So, we’re really excited about that opportunity.

The other thing that’s really big is the continued evolution of how we match a client and a Tasker. The way you may see that as a client or a Tasker on our apps is just a more seamless way to know that you’re getting the right person to do that TV mounting task on a brick wall.

I’m absolutely going to sign up for Taskrabbit to mount some TVs in my neighborhood. It feels like the thing I’m doing this afternoon.

If you love doing it, you should absolutely be doing it. You live in a neighborhood where I think there’s a lot of people mounting TVs.

There’s going to be a lot of crooked TVs here in Westchester. Thank you so much for coming on Decoder, you’re going to have to come back soon.

Thanks so much. I really enjoyed our conversation.

_{Questions or comments about this episode? Hit us up at decoder@theverge.com. We really do read every email!}