Jobox is the largest managed marketplace for home services in the US. They use HyperTrack to find nearby pros to assign and track jobs, and improve job completion rates. This is an interview with their Co-founder & CTO, Kaushik Pendurthi.
Leigh: Can you please tell me your name, your title, and about your company?
Kaushik: I'm Kaushik, I'm co-founder and CTO of Jobox.ai and we are building the largest managed marketplace for the home services industry in the USA.
Leigh: What kinds of problems or issues were you facing before you started using HyperTrack?
Kaushik: See, I'm coming from the valley, this is not my first company, I've been a co-founder CTO for the last 10 years, and I have hired the best people in the valley. So my first instinct for solving a problem is to get my team from Stanford or Berkeley or really good engineers and solve it with first principles. And that's exactly what we did when we wanted to build a location-based service internally as well. And very quickly, we realized that there are a bunch of problems over there and the ROI was very low. No customer actually looks at the solution and compliments the location part of it, but everyone wants it to work perfectly well, so that they could have their service taken care of, in the right way. So we spent like a few months, and we at every point of time, felt like the ROI was dropping down. And we also felt like if there was any tool out there, we would absolutely jump on it and use it. And over a period of time, we tested a bunch of different tools, semi-optimal, some open-source software, then finally ended up with HyperTrack and that's how we got started with HyperTrack.
Leigh: Right, I understand. Thank you for that. So how were these issues affecting your company's performance or your KPIs before you started using HyperTrack?
Kaushik: Yeah, see, we are catering to that segment of people who really care about the battery life of the phone, like they are switching between multiple different types of phone sets and stuff like that very frequently, they're not using the best phone devices out there. So they are using older OS versions or some of the phones are not new, they're pretty old. So what happens in those cases is these pros, service professionals, who drive around the city service the customers. So they always have to put their phone on charge and keep it up to the mark and stuff like that. So one of the first problems that we faced was when we build something in house, we did not have enough time to optimize the battery and the efficiency and performance, it started killing the battery, we did not get a great outcome. But in fact, it actually started getting us negative reviews, because these people wanted that. And the second thing was they were also some of the nuances where we lost the detailing of where the location was. And sometimes we sent them jobs, which were too far from the place where we were assuming that they were because our algorithm stopped collecting the data and stuff like that. So these are the kind of problems that instead of pushing us forward, it started dragging us backward.
Leigh: Fantastic, how did it feel? Not having battery efficient app or continuity of location data?
Kaushik: See, as soon as this problem came up, the team really loved it. They wanted to solve a complex problem. They were looking at interesting algorithms and stuff like that. But I would say the biggest people who were worried about this problem solution was not the engineering team, but the product managers, because when they started seeing the effort it takes to perfect a solution. Engineers got super excited, they wanted to do more and more. But product managers who had like metrics that they had to drive, were more concerned about how this would all play out.
Leigh: Right, got you. What was the breaking point? What was the moment where you said enough's enough, we need to find a solution?
Kaushik: That's a very interesting question. You have to understand our business a little deeper for that. So for us when we get a job from one of our demand partners—our demand partners are like Yelp or Minute Key or many different people—who have jobs where customers come to their platform and request for a service and that job actually comes into our platform programmatically. And we find the best technician who's closest to him, and dispatch it to him and make sure that we give live feedback to the customer where he understands that, the pros, the service professionals, started driving and is arriving in five minutes and things like that.
The customer who comes to request our service is in distress, actually, this is the person whose car stopped on a highway or things like that. And he wants a service right away. So at that point in time, we are catering to someone who was already going through a little bit of trouble on that specific day. You don't want to have these logistic annoyances play out and they become even more painful. So fundamental thing when there was a customer out there, not able to anticipate the time it takes for a pro to reach or choosing the right pro or not able to indicate whether the pro is driving towards him, not towards him, these are the biggest stress points for the team.
When we are building it all alone internally, we don't have so much time to solve the last 5% of the bugs. And this is where the breaking point for the team was.
Leigh: Right, I understand. Tell us a little bit about finding HyperTrack what what was it like when you found them?
Kaushik: I mean, fortunately, I was friends with the people, investors who invested in HyperTrack were also investors at my company Jobox. So they asked me to look at this company, figure out if we could have some synergy and build it together and stuff like that. My first reaction was to build it in-house as a typical Silicon Valley engineer, and take the challenge head-on and build the best solution out there so that we could really be accelerating fast. But then when I met Kashyap, I realized that the problem that he was trying to solve and the resources that he was putting into the problem, were a lot higher than what we were thinking of putting in. And then I realized the complexity of the problem, wanted to try it out for a few days, tried out the dashboards, the product, everything looked extremely neat. And it got us hooked. It was a one-way street for us when we spoke to HyperTrack.
Leigh: Great. What do you feel sets HyperTrack apart from other companies like it?
Kaushik: HyperTrack is constantly innovating. And I think their product feature set is growing really fast. And most importantly, they are the only ones out there in the market who actually deliver what they promise.
Leigh: Great, thank you. In your own words, how would you describe HyperTrack in one short sentence?
Kaushik: See we use a bunch of services, I always use the sentence, like Twilio is for phone numbers, HyperTrack is for location services, this is how I look at it.
Leigh: Thanks. Can you talk to us a little bit about how you and your team are using HyperTrack?
Kaushik: Because we are building a managed marketplace, the platform metrics are all focused on the completion rate of the jobs. And completion rate of the jobs means we need to be able to dispatch our pro, service professional, as fast as possible. And we use a bunch of algorithms to make that happen. And HyperTrack forms the core of it because we are building all those algorithms on top of HyperTrack, understanding a bunch of their API and stuff like that. So it has a direct correlation to what we are working on as like a company's North-Star Metric. And that's how we use their dashboards, we look at the performance of the individual service professional seeing how they behave when I get when they get a job, do they start driving towards the job or not all these kinds of things through HyperTrack.
Leigh: What do you like best about HyperTrack?
Kaushik: I mean, mainly their feature set, and they're reliable, and also an ever-evolving solution.
Leigh: Is there anything that doesn't quite work for you or you'd like to see improved?
Kaushik: One of the complex things is the scale. Like I'm afraid that when a certain scale hits, maybe HyperTrack will have glitches, or maybe their dashboards, or maybe something would stop working. This is what we are constantly looking at. But we also have a throttling mechanism internally as well, where we are slow funnel releasing HyperTrack to our users. But until now, I don't see an issue that is hitting the scale.
Leigh: So how do you and your team feel after you started using HyperTrack? How did it feel overall for your team?
Kaushik: As I mentioned earlier, it's a one-way street, we implemented it once. And we have stopped caring about the location issues, battery issues, efficiency issues, not being able to track location issues, all those kinds of things. We have a fairly high-level satisfaction from HyperTrack, which we are enjoying at this point in time. And the second thing is the feature set that they are building is actually accelerating our product roadmap as well, the one that where we can share the ETA to also the customer-facing application from HyperTrack that benefits us a lot.
Overall, I would say the North-Star Metric that we initially aimed for, the completion rate of the marketplace, it's moving much faster with HyperTrack, it's making it easier. The location has become more like a commodity for us at this point in time building the features on top of it, rather than it being the central part for our company to focus on. But however, we are reaping the benefits in this process.
Leigh: What problems are your solving or what benefits have you realized by using HyperTrack?
Kaushik: The problems we are solving are trying to match the pro to the customer through our platform, we use HyperTrack to understand where they are, and whether they're moving and to sometimes even assess whether they're probably far from the car or something because they are in a stationary position or they are in a restaurant or they have parked the car and things like that. And also the other interesting thing that HyperTrack helps us with is when the pro arrives at the location and starts walking, then we know that he is a minute away from the customer or two minutes away from the customer. And we immediately notify the customer to look around for Joe or John who's going to come to them and solve the problem and stuff like that. And sometimes it's a very important use case because you have these people stuck on highways or like secluded areas and we don't want them to go talk to just anyone else who seemed like a service professional but with HyperTrack that gets us a little more clear.
Leigh: Great awesome. What features are you benefiting the most from and why? If you can just go through one or two of the features that you feel are most useful for you and your team?
Kaushik: The feature of identifying all the pros around the job location, sorted in an order where we could choose right from the top to bottom, this is a little complex problem, more complex than we usually think of. And doing it really fast in a scalable way, doing it 1000s of times in a second or a fraction of a second. This is something that we are highly benefiting from.
Leigh: Perfect. Do you have any KPIs or metrics of success that you'd be able to share with us? Do you have any numbers available that you could talk about?
Kaushik: We are a stealth mode startup, we do not like to share too much information out.
Leigh: Do you feel that the original problem that you are facing is solved or is on the way to being solved?
Kaushik: The original problem is actually solved. The fun part is with HyperTrack, we are becoming more creative, and we are creating more problems and solving them.
Leigh: Perfect and so the last question here is would you recommend HyperTrack and if so why?
Kaushik: Hundred percent, I would refer HyperTrack one for the reason that there are a lot of solutions out there that promise that they can really help accelerate the product development and they're reliable. But I've used a bunch of them. I don't want to name them right now. But they're very hard to solve the problem right. The solution is always somewhere around the corner, but it is not accurate. With HyperTrack. I have seen that it works in production. And I can just say that with 1000s of queries happening every second it still works very, very well.
Get in touch: To learn more about how to build an order assignment system along with battery efficient fleet apps for your on-demand operations using HyperTrack, please book a time to talk to us.