Lesson from 3billion+ marketplaces - Josh Breinlinger, Jackson Square Ventures

[Music] hey everybody excited to be here so first of all I'm I'm glad I'm speaking after lunch studies always show that right after lunch people are generally happier and so you get a much better reception it even works for parole hearings so if you ever need to know that information get your parole hearing right after lunch and I'm from San Francisco so I think it's 3:00 in the morning in San Francisco so if I say anything completely incoherent just blame the jetlag other than that excited to tell you about my experience in marketplaces so I'm a GP at Jackson Square Ventures where of early-stage fund in San Francisco 240 million under management and it's all early-stage SAS and marketplaces so what we mean by that is we right one to five million dollar checks in companies that are typically somewhere between five to thirty million in valuation I don't like to use the words seed and Series a anymore because as you've heard tons of times you know people are now raising like three or four seed rounds and the price is going up and it just doesn't mean anything so we focus on just dollars in and valuation oh there we go okay so these are just some of the companies I've been involved in up work I was on the founding team at the time it was called oDesk has anyone used oDesk or up work okay fantastic so I was on the founding team when it was for people 2004 I was there from 2004 to 2010 and basically I ran a little bit of everything so I ran sales product community operations marketing business and along the way I made a ton of mistakes and luckily we got a few things right and so a lot of the lessons that are in this presentation are some of the insights that I gained from trying to grow that marketplace and I'll be happy to share all all the stories of the mistakes I made but of course the company is now public went public on NASDAQ about a month or two ago so no forward-looking statements only retroactive looking at the mistakes also a founding team member at Reve which does translation and Auto audio transcription captions and subtitles and so it's what most people are calling a fully managed marketplace these days and then my job for the last eight years is a marketplace investor so as a VC I've invested in contently which is a b2b content marketing SAS plus marketplace play bus comm is exactly what it sounds like it's a bus marketplace Thank You dan Strava some of you may know or anybody a Strava user okay we've got a few we should get some more if you run or ride you should use Strava it's really a social network for athletes that to track all their rides and runs offer up is one of the largest mobile marketplaces in the world certainly the largest in the US and its really a Craigslist competitor so it's peer-to-peer classifieds take out your phone snap a picture of something you want to sell and there's over 100 million items listed every single year and I think they publicly announce that we're doing over 25 billion in annual transaction volume so they've reached a really big scale rented is in the real estate space connecting property managers to property owners it's kind of like a b2b version of Airbnb craxy is another one commercial real estate marketplace kindly care marketplace for elder care givers and Aalto pharmacy is is basically like an amazon for pharma and so same-day delivery of any pharma anywhere in the Bay Area or LA at this point so that just gives you a sense of where I'm coming from we have done typically either the seed or series a even though I don't use those words in almost all of these deals so whoops this button okay so these are the the rest of the presentation is going to be all of the lessons that I've learned in these marketplaces and so three of these have reached over a billion dollars in gmv work there public you can look at their statements offer up is doing over 25 billion and craxi or is doing well over a billion but they cheated a little bit because the average transaction is five million dollars so they got to that scale pretty quick if you're interested my blog is at a crowded space and pretty much all of the detailed lessons are on the blog somewhere and I'll post the presentation up there right after this okay so lesson one you've seen a ton of these sort of marketplace checklists before so these are typically the things that in my mind make a perfect marketplace if you have all of these you win and they're really dynamics of the industry and the product that you're building I'm you can read most of these but I'm just gonna point out a couple ones that I think are a little bit different than maybe most lists that you see the first one is really around productivity gains and this is something I think a ton about you know when you hear about b2b market places that are doing things in pen and paper what they're really talking about is the transaction cost and so it might be fairly easy once people have discovered each other to complete that transaction but what I think you need to understand is all of the things that happen outside of the typical platform and so the good marketplace is nowadays are taking more and more of that on themselves and really just reducing the overall transaction cost for a buyer to to meet a seller and complete the transaction and so we're always looking for a market like but bus comm is actually a perfect example if you want to go charter a bus it's a massive pain you have to go find a whole bunch of different suppliers call them up get rfqs in place it's all emails and calls back and forth and they do it self-service online and so suddenly that transaction cost has been radically reduced and that's why it's working so you've heard all about massive markets obviously high fragmentation one other one I'll point out that I think a lot of people sort of get burned by is this notion of irregular you know quote non-monogamous usage and this is I'm gonna give you an example about nannies in a minute it marketplaces function much much better if you're interacting with a whole bunch of different suppliers if you're if you're really just a buyer that's looking to find one seller and then you're done the reality is the marketplace just doesn't add much value after that match and so that creates all kinds of problems later on for the platform the last one I'll point out is the no disintermediation so this is a problem for pretty much everyone and one of the things I always really look out for is there's certain marketplaces where once a buyer and seller meet and once they've established their own trust they actually both want to get off of the platform and that causes just these massive problems sometimes they outgrow the marketplace and that just kills you because your best users are leaving your platform so it's really important to understand the disintermediation characteristics okay anyone have all ten of those if you do come talk to me later so I would love to learn more so a couple examples of ones some of these can be deal killers almost nobody has all ten of them perfectly but I'll just point out you know some marketplaces that have always seemed to make sense to me but then I think they fail one of these criteria so nany's is a perfect example I have two kids at home we have a full-time nanny and I've absolutely no need for a marketplace once I found her right I pay her directly iMessage her on whatsapp there's simply like none of us want a marketplace in the middle anymore and so when I have that relationship if I just have someone employed full-time for a year and they're local and on site that's a disaster for a marketplace so I don't think we're ever gonna see a marketplace you know there's there's marketplaces that might be about the introduction and you see things like care comm they have to basically try to bleed you dry to get a monthly subscription just so you can contact a couple mayonnaise but then they're done and I have no reason to ever go back to them so nannies fails another one I've seen is aircraft maintenance which at first I thought was an incredibly interesting marketplace it's a surprisingly large market I think it's about forty billion dollars and it's a huge problem because obviously your if your aircraft is grounded at all you're losing tons of money very very quickly and so the the speed of maintenance is incredibly important but the big problem here is there's no fragmentation you have really just a couple dozen airlines that really matter and so the marketplace in the middle can't make any money they just don't have any pricing power and they get forced way down by the big airlines so this one fails on the fragmentation model and then locksmiths this is one of my sort of easiest passes I've ever done so when uber came along you know they created the whole on-demand economy wave of startups and everybody was trying to do the uber Forex and one of them is a couple people trying to over for locksmiths and at first it sounds great right wouldn't that be nice if you need a locksmith you know click a button boom you have it but the problem is just the frequency I think I've used a locksmith twice in my life and I've paid about $100 each time and so there's just not enough frequency of usage in the locksmiths category to ever build something interesting so that gets me to my next slide and apologies you probably can't read the the small writing in there but this is how I basically chart all marketplaces that I see and these I think are two of the most important variables which are basically frequency of usage and it's specifically frequency of match so while I might use my nanny every single day I'm not matching with a new one every day so it's really the frequency of the match and the size of the transaction and this isn't your classic you know two by two where you just want to be in the upper right what this is is really defining what type of marketplace you should build based on where you are in that two by two so things that are way on the right super high frequency things like taxis and meals which is why we have uber and postmates sind or - or whatever's big here and those models should all have a marketplace that has ultra low transaction costs that's why it's what I call a supplier picks model I open an app I press a button it's first-come first-serve and it just happens very very quickly and then at the other extreme you have things like buying a home which is a well in San Francisco it's a couple million dollars no but he wants an uber for buying a home you don't want to open an app press a button then you're out two million dollars because there's no standardization it's very custom there's tons of different variables that go into it and so what you need in in that model is more of what I call a buyer picks approach which the buyer can scan and browse millions of listings and you have things like Zilly Zilly and Trulia the charge more on a lead gen model and then you have all the things that are in the bottom left and my opinion is that in those categories verticals usually don't make sense so that's where you get things like locksmiths and plumbing and handyman and carpenters and I don't believe a vertical will work for any one of those things I think you need to group them into a horizontal platform which is why you have things like thumbtack in the US that is by far now the largest home service provider so they they have Home Services across 30 different categories and I think that's the only way because once you group those things what you've effectively done is increase the frequency because I can buy across 30 different categories so I'm a big fan of that the other one is offer up which is actually started with just baby and kid stuff and the reason they started with that is because if you have kids you know that you're buying for different strollers two different cribs a changing table clothes you outgrow every three months and so there's just naturally this very high frequency going on in the baby and kids category but that wasn't enough and so very quickly we grew out of that category and now have about thirty different categories and it's literally everything electronics sporting goods you name it so if you're whatever industry you're looking at if it's in one of those things that's relatively low frequency compared to some of these others I think you need to adopt a more horizontal approach and then you have the top right which is sort of empty at the moment this is where I think a lot of the opportunities are going forward and you've heard a lot of people talk about b2b marketplaces I think a lot of the the high high spend high frequency things tend to be in the b2b cases the challenge is they work a little bit differently and you can't take 30 percent all the time but but that's where I think a lot of the opportunities are okay so I was just talking about rakes this is how I view the rake factors I think the average I see is probably about a 20% rake for marketplaces that's you know kind of somewhere in the middle and these are just the things that I've observed that will either drive you very high or very low and I think this is really important to understand by the way the sort of key takeaways is at the bottom high rake is definitely not necessarily better some of the best market places have exceptionally low rates like Alibaba is an average of 3% but if you have if you're in a smaller market or a vertical you absolutely have to have a much higher rake so sometimes it's good sometimes it's bad but these are the characteristics that drive it way up if you have a fully managed marketplace you're taking ownership of the quality and the deliverable you get to charge really high rates as I've seen 50 60 70 percent in some cases and then the next level is you're not taking full ownership of it but you have lots of things like insurance and compliance and you can do scheduling and availability on your platform those get a little bit higher and that is it's going down you start to have you know in the 10 to 20 percent you start to have more transparent fees and but you're not taking as much ownership of it and you tend to have higher price things like Airbnb stays and things like that and then the tiny rakes and the ones that you know oftentimes start at zero and then build up from there are the ones that tend to meet in person you know you can think of the nanny's example or peer-to-peer goods or use cars or things like that often start completely free but the markets are so massive that you can start completely free and then figure out how to add value added services on top to get your effective rake up okay so next thing we're going to talk about is match types so this is a big pain and suffering I had it o desk in the early days is basically what we thought of is the fill ratio and the fill ratio was what is the number of people hired over the total number of jobs posted and that was just an incredibly important metric to us because obviously that is directly correlated with revenue but it also is correlated to just customer satisfaction on both sides like the they came there to find someone and so the higher that is the more satisfied they are the problem is that the type of match that you do has a massive impact on that overall fill ratio and so the lowest is what I call double commit I have to post a job candidates need to apply and then I need to interview and and sort of pick someone and then they need to actually accept the work so there's a lot of back-and-forth and you know it violates every conversion funnel thing you've ever seen right you have just tons of opportunity to fall out of the process so that one unfortunately is the worst for fill ratio but sometimes it's necessary because it you just don't have any standardization and there's no way to do the you know click a button and someone takes the job so the next best one is buyer picks which is basically how air B&B works I can scan through thousands of listings and you've seen over the years they've moved from basically when they started it was all the double commit model I'd see a place I said hey looks great I want to stay this weekend and then they'd have to approve me and you know decide if they wanted to rent their place out to me what you've seen is Airbnb is pushed so hard to get to the instant booking model and that's exactly because of this issue it drives conversion ratio way up you don't have that drop-off and so buyer picks with a single commit you know as soon as I see it if I want it book it that's that's the next best thing for filler ratio and then the best one is supplier picks and that's basically what over and lift have so I just say I want a ride from A to B and then I don't get involved at all it's really just first-come first-serve or sought ematic Aliyah signed and that's in my experience the only way that you can get too close to a 100% fill ratio is if you have that supplier picks model I can almost guarantee if you have a if you have a healthy network and you have a first-come first-serve anybody can claim the job it will get claimed so if you have the opportunity to go to that type of model you absolutely should okay so this is the the next really big learning I had at the early days of Oh desk and now op work and this is about quality so quality is obviously incredibly important for every marketplace you know for a lot I've seen quality be the number one priority for the entire existence of a company and this is how I think about it so first of all it's never static the quality if you're in a services marketplace you know there's gonna be people coming in and leaving all the time and everybody's of a different quality so I think you need to obviously attract high quality people to your network do some sort of screening testing verification whatever it is you decide to do and then the more important ones in my mind are the next three so optimize who gets what work you you as the marketplace founder or operator get too heavily influenced that if not completely decide on your own and then retain the best people make sure nobody is ever leaving your platform and then of course expel the worst people so what I have observed is a lot of people spend way too much time on screening they think you know hey we just need to make sure we get good quality in and we have all these great tests and only 10% of people pass our screen and they're all excited about you know they have these rigorous standards but my belief is we're actually terrible at screening I'm sure most of you have hired lots of people in your careers and you've made terrible mistakes and you've had to fire people so we're just bad at it even though you did the best job you could you source candidates you did all your interviews we still just suck at screening it's very hard to predict who's gonna end up being great and doing great work and so I think we tend to waste a lot of time on the screening process and we're potentially filtering out some of the people that will turn out to be the best like a lot of times I see people just doing simple screens on hey they need to have at least five years of experience well I don't think that's a particularly meaningful metric so why are you screening based on that you know obviously we all started somewhere so my usual recommendation to people is stop worrying about screening so much let way more people in but then be much more aggressive about the next three we or the optimize retain and expel and it's actually a question I asked a lot of founders when when they're pitching of ok how many people have you kicked off the marketplace and if the answer is zero I get very nervous right because they're just not monitoring the quality and what will happen is if you let the bad actors stay on the marketplace it will actually cause this vicious cycle of quality so the good people that are on the marketplace will just get annoyed you know like attracts like so high quality people like the this room you want to be around other really high quality people and if you see a whole bunch of buffoons hanging out you just don't want to be there anymore so you leave and that's the worst thing that can happen obviously is if you have your best users start leaving and and we had some of these problems you know back in the early days of oDesk but we course-corrected and so we got way more aggressive about kicking people out if they had bad feedback scores or weren't showing up or whatever it was we got way more aggressive about kicking people out and we got way more aggressive about retaining people and the way we did it is really just helped them increase their wages as quickly as possible so made sure that the best people on the platform one had as much work as they could possibly want and had wages that were constantly increasing because they we knew they were great so stop worrying so much about screening ok and and this actually gets to sort of the next point which is what what I call the desperation curve so we found this this is exactly about that screening issue it can actually have the opposite effect you're intending so we put in all these hurdles in these tests and certifications and then what we found is the best people just didn't bother going through them at all and so it actually putting in those hurdles lowered the quality of the network even faster so again you know and it's the same sort of thing great developers or engineers in Silicon Valley they never apply to jobs they just get recruited and so you need to design your marketplace and workflow and screen in an onboarding process for the highest quality people and just make sure you have a really good way to get the low quality people out ok that was so the next one is about teaching your users and this is something I think is just incredibly important and I've discovered a few tactics over the years that I think work incredibly well so first of all every marketplace is completely different the policy is the way you use it they're all different and so you need ways to kind of educate your users as they're coming onboard and a few of my favorite ones here matching new and repeat users so what we found we actually had this thing called the noob ratio so the newbie ratio and Edo desk was trying to make sure that that ratio was as close to zero as possible we never wanted a completely new freelancer to work with a completely new buyer because we just knew statistically the odds of failure were close to a hundred percent but if we had an either an experienced buyer or an experienced freelancer you know odds of success went way up and so we actually heavily encouraged new buyers not even to like review candidates that were completely new we sort of hid those from the interface a lot of times so do whatever you can a lot of times to match new and repeat successful users and that is a good way to onboard people community forums is the next one I'm a huge fan of community forums I think they've kind of gone out of favor and a lot of people don't have them anymore but I think it's a great way to get product feedback so I encourage everyone to go back and launch community forums and then the last one is is one of my favorites that some of my companies do you know you might have a marketplace I'm sure obviously if you're the founders you're you know eating your own and dog food or champagne or steak depending on how you like to phrase it but one of my favorite tactics is actually to delete every employees account each week and so every employees starting on Monday has to resign up and go through the onboarding process and it's just a great way to make sure that everyone is experiencing the same thing as your users over and over again and I guarantee once once you put engineers through this process a number of times and there's any annoying step it'll it'll quickly get fixed so that's one of my favorite ones just start deleting people's accounts see what happens so network effects most people think network effects look like something on the left where you know hey I get more users and I fall is it Metcalfe's law I think it just keeps getting better and better and better over time but I think the reality is nothing like that I think the reality plateaus if Facebook adds another hundred million users in China it doesn't change my experience at all right so the reality is that the network effects taper off pretty quickly and and we found that it oDesk and we found that an offer up we found that a craxi we found this in pretty much every market so if you think about it from a job posting perspective if you post a job and you get ten qualified candidates that's probably great that's all you need and we kept going and we kept getting more and more scale and then we were getting up to a hundred candidates per job but that actually became worse than when we had ten because then people were just completely overwhelmed had no idea what to do and so you need to actually constantly reinforce the network effects and I think that's that can only come by using your scale to do new things and unfortunately I hate to say it sometimes but Eber has done the best job of this if I was an investor I'd be more proud of it but I missed out on that one so uber has done a great job they had the black cars and then they had enough scale that they could do uberx and things got a lot cheaper and that plateaued again because there's only so far you can take it there is some effectively fixed costs of the drivers and gas and there's only so much you can do so the next thing they had to do was a pool so again completely reinvented the service offering once they had enough scale that they could do something completely new and for what it's worth if if you guys know nfx it's one of the marketplace investors in the u.s. they have a great blog you should search on for reinforcing network effects I think it's one of the best pieces out there about how to consistently do this and if you don't I can almost guarantee someone else is going to come along and invent a new service and I think that's that's one of the ways startups actually get disrupted fairly quickly is they they just count on the existing network effect in the initial product offering too much without looking to completely new ways to scale things and introduce new network effects okay the next one is on tearing and this was a this was a huge learning for me personally at oDesk we we had this vision which is great which is hey we're gonna build this amazing global meritocracy and everyone's gonna get these feedback scores and we're just gonna display all this information and the cream will rise to the top and everything is going to work out fine and of course we were completely wrong it sort of works but you really need to introduce segmentation I think there's another talk about segmentation in users for marketplaces and I missed that one unfortunately but my experience is it really follows the classic 8020 rule once you know that 20% of users are doing 80 percent of the work and 80 percent of most importantly the good work create a different tier they're a different segment I only have five minutes okay so just a really quick example what we do at Rev where we're doing audio transcription is once we know people are really good we promote them to what we call a river Pro tier and they get certain benefits they get early access to jobs so it's first-come first-serve to claim jobs but they get to see the jobs an hour before everybody else does and so that guarantees that that does the optimization thing we know that the best workers are going to be working on most of the jobs because of those segmentation tiers so disintermediation this is just very simple sort of graphic this is the way I think most users perceive the value in marketplaces they perceive a ton of value upfront and then usually it declines rapidly over time as you know things like trust in the matching sort of fade from memory and you're just thinking about why am I paying this marketplace 10% still so you know at some point users are gonna try to disintermediate and so you really only have two choices increase your perceived value or lower your fees and sometimes you don't have any good way to increase your value so I think more and more marketplaces will be lowering fees over time doing things where you know once you've spent $10,000 the the rake drops and so people are going to be doing more and more dynamic pricing over time on feedback scores so I hate five-star feedback systems I just think they're they're fundamentally inaccurate and what you should notice from this is that most of these are sort of five centered like eighty percent of all the scores or five stars so it doesn't really give you any sort of signal by the way and Yelp it's all four stars for whatever reason but the point here is that humans if you look at a five star distribution chart you're gonna have a few ones where people really hated the experience a few fours and a ton of fives and that's what every distribution chart tends to look like and you have no twos or threes so as humans we just think of more on a three point scale either I loved it I hated it or it was just okay so I think more people should move to just asking like would you like to use this again yes or no would you like to work with this person again yes or no it's more actionable and more tangible and the other thing is definitely do this as private feedback we found that when we did a completely public and you know I was rating you and you were rating me there was just this massive great inflation where everybody wanted to be nice and so everyone just gave everybody five stars as soon as we made that private then we started getting much much better signal in the data and actually feedback scores dropped by a full point but that was important because we could then act on it and actually do the retention and and expulsion that we needed to do so by the way the other I think future of feedback systems is to focus more on behavioral data so what do you actually do with the users do you have repeat transactions that's probably way more meaningful then did you give somebody five stars the repeat transactions is just way more important and the other thing that I think will eventually happen that I haven't seen anyone do yet is sort of what Google Pagerank does right a link from CNN counts way more than a link from my blog but in feedback systems everybody's feedback counts exactly the same and I don't think that's the right way to do things you know there's some buyers or some sellers where their feedback is just much more meaningful because they're experienced users they know the system that give good signal and good ratings that should matter way more then then every other user I think I'm almost to the end so I'll make it in time so there's just kind of a fun observation I've seen in every company that doesn't marketplace it's some point you start off and you have no trust and safety team and then at some point the fraudsters kind of find your platform and they figure out how to take advantage of it and they're very good early adopters and they talk a lot in these communities and they tell everyone else how to exploit your marketplace so you have this massive spike in fraud and then you overreact and you introduce a trust and safety team and they clamp down and they add all these security measures and ID measures and things like that and then the steady-state is you have a growth team that's basically fighting your trust and safety team and the important observation though is that if you have zero fraud you've probably still over corrected so some amount of fraud is okay and I think it's important to balance the the growth and and the fraud things so this is a new wave of marketplaces that I've seen is I think of as the master marketplaces and I I think there's commodity marketplaces and master marketplaces and they actually cannot mix in my opinion so you know masterclass is one that I'm sure a lot of people have heard of that type of content could never be on something like udemy or Coursera they just do not mix the buyers and suppliers are looking for something completely different so I think in many cases where there is a big commodity marketplace player there's actually a new opportunity to build sort of the master one or the super luxury one let's I guess what we've basically seen with Amazon and Farfetch'd right a lot of the far-fetched products and brands would never participate on Amazon they just don't want to be there and the buyers are looking for something completely different as well and then the last one I'll leave you with is really for marketplace founders out there in thinking about liquidity hacking so it's the number one challenge of pretty much every marketplace how do you solve the chicken and the egg problem the quick pop quiz does anybody know which one came first the chicken or the egg raise your hand if you think it was the chicken okay a few brave souls how many think it was the egg okay a little bit more so if you believe in evolution then you believe that mutations happen only in the embryonic stage and so there is an animal that was not a chicken that hatched the first egg that then became a chicken so the egg came first but this is how to do it in my mind these are the tactics that I've seen work to hack liquidity and get to those initial transactions the first one is finding big aggregators so a lot of times whether it's on the supply side or the buy side there's some sort of aggregators that you know could be a big office campus could be one big enterprise client that has millions of jobs there's a lot of these aggregators out there that you can find that may not really fit with your perfect vision of hade massive fragmentation I need millions of users all around but it's a way to get you off the ground the other common way is just sort of the what most people call the single player mode provide value to one side and that is sort of what github has happened with github basically they provide a tool every developer uses it and now github isn't doing it directly but tons of recruiting platforms are now built on top of github using that data to become a marketplace for for developers with offer up it was really about completely narrowing your focus I mentioned before they of course they have a vision they want anyone to be able to sell anything to anyone else locally like that's a huge vision but you can't start there so they started with just new moms ages 25 to 35 just in a few sip codes in Seattle just buying and selling baby and kid stuff so a lot of times you can just be intensely focused on a very small piece of the market and that gives you enough liquidity to get the ball rolling and get to the next step the next quick one is curation your mechanic does a nice job of this they didn't have a big network but they really only showed you two or three highly qualified options and then you just pick from there that works sometimes and then the last one is basically what I did at oDesk which is the brute force method and you know you sort of heard that from I think it was tor radar you know basically going around knocking door-to-door at every tour operator and just like basically not leaving until they sign up we all know that doesn't scale but sometimes that's the way you have to start and with oDesk basically I would be on the phone with somebody looking for web developers then I would start chatting with my friends in Russia like hey Demetri do you know any developers that can work on this project so of course it doesn't scale but I was basically a recruiter for the first year and eventually you get enough liquidity going in the marketplace that you can sort of have it work on its own the we actually did not have a marketplace for the first year and a half at oDesk we only had the tools and the only way to get connected was to call us and then we would like scramble to try to find you some people in Russia or India Ukraine wherever so that's those are my thoughts on liquidity hacking hopefully that helps the entrepreneurs out there and I think we have a little time for questions [Applause] hi Josh thank you for the presentation just a quick question you were quite against the nanny marketplaces but you've invested in kindly care which is basically the nannies for the elderly so what was your thinking on that one because you know there is no market and you know there's Network effects it's an introduction business the CAC is crazy is there anything that you know we're missing and you know Igor I guess it's ego right the founder has done unique there and based on your answer I might have a follow-up question okay fair enough thanks for calling me out on that one so the I think there is some interesting components to the Eldar space verse the the nanny space and a lot of it is around basically insurance compliance and actually various employment laws so what's interesting about the nanny thing is you sort of have two options in the space you can go to Craigslist route or you can go the like care agency route but the problem is those are basically both very expensive it's like Craigslist is suboptimal or agencies are incredibly expensive the best way to do it is actually to become the employer yourself so if I needed a caregiver for my mom if I become the employer I can save about $20,000 a year which is huge and so kindly care is actually helping families become the domestic employer which basically I don't think they could do on their own yeah do you believe in so you said you don't want to you know you don't want a nanny marketplace and in less than that because you can hire the full-time nanny what do you think about the the part-time the carpooling for the nannies you know I have someone have a you know a van pick up the five kids from the same school and basically are sharing the code because nannies and elderly care and all those sort of services are kind of very expensive right and a couple of grand a month and you you think those carpool you know higher utilization of same work first whether you have same people doing you know you get the idea what do you think about so I've looked at a bunch of those and and I haven't made any investments there and it's largely because I think the the uber for kids is uber and we've already seen it starting to happen and it's kind of like coming down like it would it started as just adults and then I went and talked to a bunch of high school teachers and guidance counselor's and they're like yeah everybody just uses uber and so I think the the like dedicated kid ones or dedicated elder ones are getting their niche is shrinking rapidly so I just don't think there's a great opportunity there and I think all the big players have their eyes on that too like they're obviously aware it exists so I think the the big players will win that one well it's too late stage for me now yeah I mean I think it I think it surprised a lot of people because I think people underestimated the market size of it and it definitely has some of the challenges of like hey if I have a dog walker or something like I'm just gonna use the same one every time it has some of those challenges but I think the markets actually big enough and the the needs are varied enough that they've certainly done a great job and carved out a great niche and there's there's enough other services that they can expand to like they have a sort of in my mind a mini horizontal right it's I think they're doing everything now dog walking sitting and whatever it's called the overnight care and grew and they'll probably do more grooming and they'll probably get into more dog food and veterinary health services and things like that so I think they're building a reasonably nice platform there so yeah I think it's a great company but too late come gracias what do you do when when somebody like Nass per se will let go jumped into the categoría and starts pulling money like crazy do you get you check ready to you know make another round yeah Naspers anybody from Naspers here ok good so Naspers is my nemesis if anyone doesn't know I'm an investor in offer up which you know is the marketplace in the US the largest competitor is let go which is funded by NASA's and Naspers is announce that they just invested another five hundred million in the business and they do things that are kind of crazy so they invested in TV ads and at one point they were spending twenty five million dollars a month just on TV advertising so you know that's sometimes a frightening competitor to have somebody that that's spending that aggressively on on acquisition what we've really done is is basically just focus on what we think are the things that really matter and and for us it's the things basically liquidity and gmv and user experience because I think anyone can buy downloads you know you put money into a channel and you'll get downloads but if you're not providing a good experience in providing the good cohort retention you're basically just wasting the money and so we think that's what our competition is doing generally but but yeah I mean the there's enough big funds out there that I think you know irrationally spending competitors don't don't scare us too much some back there yeah thanks for the presentation you mentioned how important it is to keep the quality high by kicking certain marketplace participants out of the platform can you give examples but this applies to customers as well or do you only think this is relevant for the supply side that's that's a great question a hundred percent is relevant for both sides and yeah we absolutely kicked people off the platform on the customer side and it's always a little bit harder decision but I'm confident it's the right decision to kick people off and you know it's the same thing you know if you take the the oDesk example no great freelance developer wants to work with a really obnoxious buyer that doesn't know what they're doing right that's just a bad experience for the good people so you're hurting yourself you're hurting your marketplace if you keep them around it's a little bit harder to identify sometimes because the the feedback and behavior signals aren't there but absolutely you should fire your customers if they're like causing problems in your marketplace so I think it's just trickier to identify okay well thank you so much and I'll be around [Applause]

Channel: Marketplace Conference

Views: 12,613

Rating: 4.9867549 out of 5

Keywords: The Marketplace Conference, Marketplaces, Speedinvest x, Autotech Ventures, Market One Capital, Point Nine Capital, Josh Breinlinger, Jackson Square Ventures, Speedinvest

Id: _xHsHWblK6U

Length: 43min 47sec (2627 seconds)

Published: Thu Dec 27 2018