 Hello, everyone. My name is Gul Chin. Today, we're going to discuss with you how we can solve the marketplace supply shortage problem with the help of product. Before we move on with the topic, I'd like to give a very brief introduction of myself. Currently, I'm working as a senior product manager at Firmary eBay in New York, DaVinta. But before this, I've worked for the last eight and a half years in different tech companies including Uber, Rocket Internet, etc. Mostly focusing on the marketing previously and on the product management lately. So today, we're going to discover and explore what are the possibilities we can unlock with the help of product to help the market supply shortage problem. But before we move on with the discussion, I'd like also to refer again to the product development framework. As you know, most of us are the product managers. We have kind of a framework we need to follow to make sure that any product we build is successful. Where this product development framework usually starts, first of all, with us discovering problem space, meaning that before we start working on anything, we need to make sure that we have a clear understanding of the problem. What's the problem size? What's the problem? What's the root cause of the problem causing it to happen? Who are experiencing this issue? What's the impact? What's the size of the impact in terms of revenue or other business metrics? How frequently this problem occurs, etc. So all those questions and many more would give us a better understanding of the problem scope. And after we have a clear understanding of the problem scope, then we usually move forward with different phases, starting with ideation. If we have already a clear understanding of the problem, then we move forward with exploring different ideas, how we can potentially solve this problem, also sharing it with internal and external stakeholders. After the ideation phase, if you already have a few wireframes or options, how we want to solve it, we move forward to design some demos or options, and then we start validating it with our main customers to make sure that this idea we came forward with would work for customers and would solve their problems. Without validation, we can never be sure that we are working on the right idea. And after we pass this phase, we usually continue with the development or launch phases. But today, I wanted to once again highlight the importance of the problem discovery, because it's a very important very first step that everyone needs to check into consideration before jumping on the ideation or the solution phases. So if we are working on the under supply problem, it's very important for all of us to understand why we are supply constrained as a company. There might be several reasons, and some of those reasons might be, first of all, to start asking your colleagues or yourself if this is an acquisition problem. Maybe we are undersupplied because we are not acquiring enough suppliers. Or maybe we are not acquiring the right suppliers. We are acquiring the people, but they're not right fit based on our needs. Of course, those are the marketing questions, but you need to collaborate with all the departments if you are given a problem of under supply and you are trying to solve it. So then you need also to check since when this is happening, why this is happening, if this is an acquisition problem. Is this a budget problem? Maybe you did some change with the budgeting lately, which impacted your under supply problem. Or are there any supply limitations in the given area? Maybe they were a limited number of the suppliers available in this area and maybe you already reached this number. Therefore, you are unable now to extend it further. But what if it's not an acquisition problem? You double checked all the lines and asked even more questions, did more analysis, and found out that none of those are causing your under supply problem. Then the next question is to look into is, if there is a problem with acquired supplies conversion rate to the next jobs, meaning that let's assume that you acquired a million people every month, and then you start noticing that only 1% of those people converting to the next job while previously it was higher or historically it has been lower. Then the question is why? Why the minimum number of the people are converting to the next jobs? Why this rate conversion rate is below the industry benchmarks or your company benchmarks or market benchmarks you are operating in? What's happening there? Why people are not converting? Maybe it's a UI problem or maybe there are more steps in the signup funnel or supplier funnel that's causing them not to convert to the next jobs or maybe they find some bugs, experiencing certain challenges with a converting. So it's very important for you to discover if this is a conversion problem and secondly, what's the reason causing it? But let's assume that it's also not a conversion problem. Then third group of the problems that may cause your under supply situation can be average supply being nothing as engaged as you would like. For example, let's say that any supplier is providing a monthly targeted number of the jobs. For example, you expect an ordinary supplier for Amazon Flex to do at least seven or 10 deliveries per day or 30 deliveries per week. But when you look into the schedules or track the data, you start noticing that some proportion of your suppliers and the biggest proportion has dropped below this benchmark you have as a company might. Then it means that your suppliers are not as engaged as you'd like. Jobs per user is dropping. Then it's very important to understand why this is happening. Maybe you can do an extensive research with your UX department or you can do a survey or you can go out and directly ask thousands and hundreds of people just to learn more what's causing this problem to occur, which is a very important metric to keep an eye on. However, let's assume that none of those three problems are causing your under supply issue. All of them are really healthy for your company. Then the next question you need to ask and validate if there is any retention problem in your company, meaning that your turn rate could be high or higher than usual. And again, when you compare it to different companies, competitors, businesses, market benchmarks, you might say that your turn rate is quite high. It's above, for example, 15, 20 percent. Again, depending on your industry or depending on the markets you operate in, this might be a problem to investigate and understand why this is happening, why people are churning. Maybe they're unhappy about their earnings or maybe your suppliers are unhappy with some procedures with a level of customers, quality of the jobs they're receiving or their in-app experience is faulty. So there might be dozens of reasons, but it's very important to understand if it's a churning problem, then why people are churning. And finally, let's assume that none of those are happening or maybe two or three of those problems might also be occurring at the same time for your marketplace. It's also important for you to respond to the fifth group of the questions. Maybe any of the factors here could also potentially impact, could be impacting your supply situation, which is the seasonality and the holidays. For example, if it's a specific season and the holiday happening in the country, it's more likely that your market or one or more of the markets are going to have to differentiate its stocks or sudden change in the graphs for certain metrics, again, including supply availability. Another important factor is any change in the competition landscape can also impact your supply situation. For example, even when the competitor is doing the price increase, price decrease, or new competition is coming to the market or existing competitor is leaving the market, it's all might change your supply situation. So it's important to keep an eye on it and double check verify it whenever you are investigating your under supply problem. The third and the fourth factor is that if there is any change in the company and legislation's internal procedures or governmental legislations in regards to your supply situation, it might also impact your suppliers. For example, if you recently started to acquire new document from the suppliers, new ID check, or any other validation, then because government asked you to do so, then most likely you are going to lose certain proportion of the suppliers in the funnel or their existing suppliers might also leave because they didn't have this document before. Unfortunately, those change that you are driving internally they need to be discussed ahead. And if those happening, then you are most likely going to experience a churn or lose certain proportion of your supply, existing supply. Basically, those are some of the problems that might cause under supply issues. But again, it doesn't necessarily mean that if you are under supplied, it definitely causes one of those problems. However, those are the most frequent problem types causing under supply issue. Now let's assume that you already made your discovery and you have a clear understanding of what's causing your supply, number of the suppliers to drop. If you already have this clear overview of the problem space, then you are good to move to the next product development phases, which starts with the ideation. Now that we know the problem, let's start thinking how we can potentially solve this problem. Here, I'm going to give you some examples based on the hypothetical problems we discussed. But again, those are the potential solutions on the ideation phase. Doesn't mean that those are the right solutions for your company or for your business. You always need to you always need to dive into the problem, discovery problem root cause, understanding it, and then finding the right solution which works best for your customers. But for some of the cases, I can share ideas that might work for bringing more supply into your platform or engaging your existing supply more. Let's start with the first problem group, which was an acquisition. So let's say that the problem was that we are not able to acquire enough supply at a targeted value. This means that we have enough budget, but we don't want to spend more on the suppliers than we should. Given that we have ROI target, we always want to spend less than we expect to learn from those suppliers. Therefore, there is always a budget. Marketing teams prefer to stand with them. And given the situation, if we have the limited resources or limited channels, you might think of bringing new acquisition channels on the board with the help of product. One of those channels is a referral channel. For example, in Lyft, Lyft is one of the companies which uses referral channel where you can go and refer different drivers. And as a driver, if you refer a friend to drive with them, then if your friend completes a certain goal within the given period, for example, within the 19 days, if your friend completes 295 rides, then you will get $350. This is a friend referral program and works in the terms that the company has set for itself, which knows that the company is going to be ROI positive if this goal is hit. But before you start building a referral channel, you need to take into consideration several factors. One of those important factors is the finance. So you need to make sure that whatever budget you set for your referrals, it must be within the company caps. It needs to take into consideration the caps where you would still be ROI positive from the marketing and the product perspective. Also, you need to make a few decisions about what type of referral campaign you need to build. For example, you can build a campaign where only referring person would be rewarded. Or both parties, both referrer and the referring person, referred person, sorry, would be rewarded in the end. This is usually the most popular one where both parties are rewarded because then both parties feel engaged and they want also to try the platform, they're keen, they have a goal, financial goal to achieve the results. The third type of the referral campaigns is when only one party, but referred person is being rewarded. It's the least preferred one because usually you need to motivate the referring person to refer the friend. However, if only referred person is being rewarded, then referrer is less likely to feel engaged to participate in the program. And finally, there might be some referral programs where none of the parties are rewarded. This is also happening sometimes, especially when some companies, they know that they have valuable enough acquisition streams. They just open the channel for inviting your friends, but they don't position it as a referral channel. You still have the functionality to invite friends to the platform to let them use it. However, it's not a referral program because none of the parties have any benefits of referring other person. Another important decision you need to take into consideration when building a referral program is referral value. So what benefit do you want to give to the referring and the referred people? There are a few options without deep diving into financial value of those selections. For example, you can reward them with cash. It can be in credits or in balance. Most of the companies are using this option because cash is the most engaging, has been proven one of the most engaging tools in the referral programs for the last decade. However, some of the companies also prefer using premiums and free product upgrades, especially for referring parties. For example, you might see in internet companies they are doing it really often. If you bring a friend on the board, then you get one month of free internet. Another value type for referral campaigns can be selective benefits. It's when you have two or three options of different benefits available for the referring or referred person, but you let them select out of those options or pool of options that you have. So they get the chance of selecting, which is also one of the elements customers really like. And finally, the forced type of referral benefits can be a chance of participating in the competition. You can run the referral campaign, but you can tell that with referring each friend, for example, you will earn certain points and in the end of each month, based on your points, you will get the chance to participate in the competition where you can earn something big. Here, you know that it's a game of chance, but some companies still prefer using this option. However, there is no data, whether it's engaging or not. It needs to be tested. Basically, those are all the referral channel ideas. However, there is also another important acquisition channel, which has proven to work really well for supplier acquisition or generally for user acquisition, which is called Enablers. It's when you create as a company or as a product, you built new B2B portal for Enablers. You onboard active partner organizations who can build their businesses on top of your business to engage in your supply and they earn money out of it. For example, one of the good examples that I really like is Uber. Uber has built an Uber fleet platform. This is a platform where if I'm a car company or if I have, for example, dozens of cars or hundreds of cars, I can just register my company there and my cars and then I can hire drivers myself on a fixed rate or stable rate. It totally depends on me as a car park owner. And then I would register them in the platform, which is called Uber fleet for Uber B2B Enablers. And then they would monitor their employees. They would bring more people on the board because they have more cars and then they would train them. They would earn money out of it and they would be arranging the payments themselves. So as a company, Uber is winning an optionality here to skip the user acquisition and to skip the user training parts because enablers are already doing it as they are building their own businesses on top of existing platform. So with enabling this new channel, companies may save some time and some resources from bringing the suppliers themselves. Instead, they bring the companies and the partners on the board who are going actively to work on bringing new suppliers. This might be one of the bad valuable channels. Now about the conversion problem. So let's say that we want to solve conversion and some of the conversion problems because our acquired supply doesn't convert to the next first or until the 10th jobs, which means that if this is the funnel you have, let's assume that people come, they sign up, then they get their account activated. If they have all the documentations in place, then you expect them to download an app to start working. They receive the first request, they accept it, then they start completing the jobs first until the 10th one. So you usually want in a healthy funnel job completion rates to grow after the first job or you also expect to sign up to first job conversions to be really high. But if in your funnel, you start noticing that somewhere you have huge dropouts in different phases, then definitely in the product discovery phase, you should have already investigated the root cause of it, why people are lost somewhere in those funnel and why this proportion is so high, what's causing people to leave. Let's say that you already have some ideas why this is happening. For example, one of the ways, one of the problems if people are lost in the funnel is because app is too complicated to navigate within, then you might come up with a solution of showing them enough education when they sign up first. For example, this is from one of the apps in the internet where you can see step by step what's happening when you sign up to the platform so you can understand how this platform is working. This can also be done for your suppliers to make sure you save time on training and you move the training phase into the app directly. Besides that, if people are failing to move through the conversion funnel, you need also to double check your existing processes and existing UI. Maybe there is something in the UI side that makes people to feel challenged and they don't know what to do next or they feel it it seems a very complicated thing to use your product and that's why they're leaving. If you found out any of those barriers and you know why this is happening in the discovery process, then here in the solution phase it's very important for you as a product manager to come up with certain solutions to make improvements in the UI and at least to experiment those solutions to see if this is changing the conversion funnel to the positive. The third group of problems are turn-related problems. So let's say that your supply situation is worsening because your turn rate is really high and the suppliers that you are acquiring and they're converting to the next jobs they're gone after a certain period and you wonder how you can potentially solve this problem. Probably you already made follow-up discovery and you talk to your suppliers and you have enough information why people are churning. One of the churn reasons might be because people think they're not earning enough in your platform. So there might be two different situations here. First of all, if people are really not earning enough on your platform or they are earning money but they don't have a visibility on it because when users don't have a visibility on their earnings they assume that they are not earning enough and they start complaining. So it's very important to show the visibility on the earnings so your suppliers would be happy and then be keen and motivated to see their earnings so they wouldn't be churning often. This is at least an idea first to be experimented for example in Uber each driver can see their weekly earnings they can see their the graphic of their weekly earnings and how it has been changing and they can also see some basic statistics to relate to. Another important challenge for drivers or for suppliers sorry to churn might be because they don't know how they can potentially earn more money. If suppliers are keen to earn more money in your platform you can do in-app education potentially to help them to earn more and this might be one of the ways to experiment and see whether it would be impacting your churn rate or not. Sometimes they just think that it's not possible to increase their earnings and they try to look out different ways or to move to competition for increasing their earnings and earnings is usually one of the most important factors impacting suppliers situation with two-sided market places especially in their real-time ride-sharing companies. Another problem causing the retention issues might be your suppliers and again related to the earnings they might think that they're not receiving enough job requests or they're not receiving enough customer requests so they can earn more. This might be caused from different problems but first of all you need also to double check and verify that your search or matching algorithms are performing really well. Those are the algorithms companies such as Amazon Leafed or Uber has to make sure that customer gets their request done. For example, when you search on Amazon for any keyword the search algorithm is ranking all the results based on your criteria and also based on different criteria market-based or based on the popularity quality etc and then it's trying to show the best results for your search on top of the page. Similarly to Amazon search algorithms for example Leafed or Uber's matching algorithms are trying to look for drivers whenever you are searching for a driver based on the location and also based on the quality for their rating etc to make sure that you are matched with the best and the quickest possible nearby driver. So those algorithms are the basis for the success of those companies and the products. Therefore as a product manager if you want to double check your retention rate or to improve your turn rate and make your suppliers happy you need to make sure that those algorithms are working really well and they're fair to your suppliers. Of course you need to validate and double check and there might be no bugs on these platforms in these algorithms but it's not enough just to maintain those. You need to make sure that you have a team always working and improving search and matching algorithms because you cannot just build it once and leave there because it can get outdated as your business is growing the more cases you are experiencing the newer rules you need to bring into those algorithms or new technology you need to bring in that you don't want to miss. If you don't keep improving the search or matching algorithms one day you might be super outdated and your suppliers might just leave because they're not getting really enough jobs or there are some use cases you are not covering with the help of the rules that's going to cause your business eventually fails. Therefore it's very important to validate if everything is working right with those algorithms and then making sure you keep improving those based on the feedbacks you always validate and double check if there is anything needs to be improved here. Then the next thing that might help with the supplier retention and to make sure you can help them to earn more is smart pricing strategy. So we talked that one of the most important reasons for suppliers to turn is related to their earnings and one way of improving their earnings can be improving your pricing strategy. Of course pricing strategy can sometimes not fall into product management area but sometimes it does because there are so many things we can do with a pricing as product managers. As you know there is always a base pricing. It's a basic way of how price for your product is being calculated but there are also different pricing strategies. One of them is dynamic or search pricing. The way how dynamic pricing works is that depending on the seasonality on the demand and supply situation in a specific period of time you can set an algorithm which can define your pricing estimating how much demand or supply you are getting and it's going to show different prices for different timings. And again dynamic pricing or search pricing in the ways that we know from different companies it's taking into consideration all the market specific characteristics. So it uses real time data sometimes also the future data to understand if there is enough supply for the demand and in case if there is less supply available for the increased demand then it's going to automatically be increasing the price. When the price is increased it's expected demand to decrease and when the demand is decreasing then the marketplace equilibrium is happening meaning that there is enough demand to and supply to meet with each other. This is why dynamic pricing is being used however we also need to be careful with dynamic pricing there needs to be the highest caps and the minimum caps to make sure that it doesn't keep growing all the time and it doesn't hit the unbelievable numbers for the market places. But it's usually preferred because especially during the holidays the peak hours when there is a there are going to be high demand or higher than expected then those are the times when dynamic pricing might help to decrease the demand so the suppliers would get enough jobs and customers wouldn't get dissatisfied. At the same time it's valuable because it's increasing the suppliers earnings and they're getting more money from the jobs they would get within the normal hours. There is another pricing strategy that might potentially help to increase the suppliers earnings which is called personalized pricing. So suppliers might earn more when we charge customers more but of course our aim is never to charge customers more our aim is to make sure that there is enough supply in the marketplace to meet the demand with personalized pricing we show different prices for each customer taking into consideration user specific details for example sometimes the web browser customer using the device customer using or a historical data of customers behavior locations customer traveled we can estimate whether this customer can pay higher or less amount for the given product or given service and then we can create a user based pricing personalized pricing for this user and charge users differently. We need also to be careful with personalized pricing because there are some ethical and legal questions we need to double check and make sure that we are operating within this frame however it's also one of the ways that can improve the pricing strategy in a smarter way to make sure we have enough suppliers to meet the demand not to just dissatisfied customers and at the same time to make sure we help with the supplier earnings so they don't churn. One of the best ways of predicting preventing churn is to predict the churn before it happens. So it's really the smartest way first of all we in most of the companies we already have churn prediction algorithms in place meaning that we have we can estimate any user's probability of being churned before this user is churning and depending on the probabilities we can target those users with different communications or with different benefits and then prevent them from churning. With this we can decrease our churn rate as it's easier to resurrect churned people before they churn versus then later. And finally one of the churn reasons can always be product related we always need to take consideration that maybe there are some barriers on the UI side or there is a boggish experience for users that are causing them to churn. And if this is one of the problems that you validated during the discovery phase probably you need to work improving UI to make sure that users don't leave the platform. And finally the last problem is the engagement problem as we discussed earlier sometimes jobs per user is lower than expected so we want active supply to be very engaged and also very dedicated to the product we want them to complete at least x number of the jobs within the week or within months however if they are performing less than expected for example if they're doing only single or two jobs per month on average then there are certain ways with the help of the product we can incentivize them to change their behaviors. Changing user behavior is not easy but it's very much possible with the help of product. For example one of the ways we can use is incentives or reward programs. Incentives are really easier way of changing user behavior we can just say suppliers give them some targets and say that to do weeks to get why so do certain activities for example complete x number of the trips to earn $200 by the end of the week. So those are the how incentives works sometimes you can make it smarter, automated etc which is going to make it easier to manage and another way of it is reward programs where you set some kind of a gamification or more engaging way of working with your suppliers to make sure they get engaged with your platforms they become more loyal and then they start working towards the goals on monthly basis so they can reach out to the targets for the company. But there are different type of reward program types for example there are perks sometimes supermarkets are really loving this way of working they give users some benefits with every single shopping that they are making it's called perk and it's a reward for each purchase then there are programs such as earn and burn with every single shopping you get some kind of a point or discount that you can use in the next shopping where you burn it another way of reward program types is tiered programs this is where you assign different levels to the users you can say that for example you're a gold diamond platinum etc and then depending on those levels you can give them different goals or different reward types first type of the reward programs is gamifications and this is the favorite my favorite way of engaging existing suppliers or any type of users because gamifications is usually making it fun making it more engaging for the users assigning them badges making smart goals dynamic goals etc and the fifth way of fifth type of reward programs are hybrid programs meaning that you can merge two or three of the above versions above program types into one and build a hybrid type of program there are some criteria for reward programs it's very important to take them into consideration if you are building one the very first criteria is from the product side and from the UI side to make sure that your reward program is easy to use and it's easy to understand of course when building a reward program and deciding on what the reward should be you're going through multiple complicated processes of agreeing on the financial part agreeing on the budget making sure the budget you spend per each supplier is within your profitability cap meaning that it's ROI positive you take into consideration all the future predictions financial predictions behavioral predictions and you keep validating yourself so it's very important to make sure that when you build and when you agree on those benefits or when you agree on the terms they are not complicated and they're going to be easy to use and make it easy for others to join and use this program second important part is to make sure that the reward you are giving in the end of program is very relevant for the users so this part also requires a lot of research pre-work to be done before launching the program itself so you need to make sure that any benefits people are going to get out of this program is going to be relevant for them it's important for their day-to-day lives jobs, etc so for example you cannot give a manicure service for the Uber driver right as a benefit if there is a reward program for them you will really need to make sure that this is what they would be using on their day-to-day lives and relevant for them and the third criteria is to make sure that the reward program you are building is real-time meaning that any behavior users are doing or suppliers are going to show in your platform are going to be automatically reflected on their performances in a way that they can see it somewhere inside app this is one of the most important criteria based on the research made on reward programs that users would want to have for the reward program to succeed and finally it's very important that this program is trustable users are going to turn and leave your platform if they win anything and they don't get it in the given promised period or if they're communicating something inside the product but it doesn't happen in the end of the month so this is something we need to avoid to make sure that we establish trust with our users and they feel safe within this program I want to share one of the examples one of the companies that are using gamifications as a reward program type for example this is Uber Uber has built Uber Pro reward program for the drivers drivers here can needs to hit different goals to fall into one of the tiers which is gold, platinum and the diamond for example with every single trip they complete or depending on their cancellation rates and many other factors they earn points and based on those points they can fall into one of those tiers the program that Uber is using is more likely to be a hybrid one because there are tiers there and there is a gamification inside so it might be considered potentially a hybrid program in my opinion and here for example you can see the badges and badges and the goals how it's being communicated inside an Uber app for drivers each time people are earning the points they're going to see how much more left and then with those they can always unlock the very next level and earn more so basically this is type of the gamification you can also use for your suppliers in the platform to get them to complete more jobs within the months to earn more points and to unlock more earnings for themselves another way of increasing the supplier engagement let's assume that if this is the maximum suppliers can do they cannot complete more than 10 jobs per week but you want them to do 11s within the given time frame then you need to think some of the ways that can help them to optimize existing situation how can you optimize existing supply let's say that any supplier can if any supplier can provide only 10 jobs in an hour why not to take the 11th job in a queue and let the customer know that this job is going to be completed not within the hours they wish but within the hour you suggest them which is going to be a little bit later for example there are food delivery companies market shopping grocery delivery companies in Turkey this is one of the screenshots I got from from the app called Stegalsin here inside the app if you are searching for a product and you want it to be delivered for your home then you can schedule it for any time which is offered to you depending on the supply availability so you can also make sometimes a real time order but here because the supply is not going to be available for that kind of a late night hours they give you one hour gap instead of saying that's going to be delivered for sure within 10 minutes after you request it so basically the customer's expectations are being managed existing supplies optimization is happening because supplier knows that the job doesn't need to be completed right now but the job can be taken into queue and completed after each other within the suggested time frame this is one of the ways that can be used for optimizing existing supply and in the end I want to highlight one of the things that Karl Marx said about the supply and demand situation supply and demand constantly determine the prices of the commodities but they never balance and only constantly and this is true but the cost of the production for its part determines the escalations of supply and demand here are the three most important highlights from the days that I'd like everyone to make is first of all demand and supply problems are going to be ongoing challenges for most of the businesses but you should always start with a problem discovery as a product manager to make sure that you understand the problem space before you jump into the solutions and finally to solve this under supply problem you need to use relevant tool to improve the situation and always validate this tool testing your ideas and then measuring the impact with the right people or right customers or with the help of right metrics well thank you everyone please feel free to reach out to me through LinkedIn if you have any full-on questions but I will be happy to connect with you thanks a lot