How to Achieve Positive Unit Economics in Grocery Delivery: Yango Tech Webinar

so yeah the subject of today's uh webinar is how to achieve positive unit economics in the grocery delivery so basically in e-grocity business and today's agenda is the components of the unit economics and why they matter then we will discuss the strategies for achieving positive unit economics and uh tips and actionable insights for benchmarking including the percentage breakdown Etc uh of course feel free to post your questions the questions are very welcome uh in the chat now or in the comment sections after all feel free to send a direct message to myself or to Alexa after the webinar and we'll do our best to share the materials and respond uh to your valuable questions and during the course of the webinar we'll go into comment section from time to time to uh tackle the questions okay so now that's a quick slide about us my name is evgeny I'm a chief operating officer of young Daily Tech so I have experience in writing technological businesses and heavy operational businesses in the UK and Central Asia also have been lucky to be part of a few Erp implementations and uh Alexia is our head of implementations of Jango dele Tech and he's got a great Consulting background from well-known company McKinsey and he is actually in charge of all of our implementation and using the Best Consulting practices to implement our technology into other partners file uh brilliant thank you again everyone for uh for coming and whoever else have uh just joined so I'll give a quick uh overview of uh uh what is uh our uh company history and basically why we uh are uh incredible enough to talk about Virgo what we're going to be talking about so our first dark store we opened in May 2019 and uh in February 20 after the MVP we started to expand to the new cities and then uh in December 2021 we expanded Beyond The Home Market in uh and uh in 2022 we already have achieved the positive unit economics and uh basically we made that business profitable and then we decided to offer the same technology which we are using for uh our own business uh to our partners all around the world and those would be the uh egrossery operators uh who could be the conventional retailers or dark store operators and uh pretty much anyone who could benefit through it and we already live in the Middle East in a few countries and uh hopefully in many more countries to come uh this is like a quick snapshot of the business it's uh now uh around five uh uh 500 dark stores uh of course our operating uh is uh almost uh five stars uh the accumulated gmv is uh over a billion dollars now so it's a very large uh business of very large uh scale 99. 9 of our orders come in full uh without the substitutes and uh we managed to turn our coolers up to five times uh per hour and I will talk about more about this on uh the later stages of our webinar our infrastructure is uh extremely powerful extremely strong capable of handling uh hundreds of thousands of orders uh uh through our Cloud infrastructure now um again I will repeat for those who have just joined like uh what we do in a nutshell is that we provide end-to-end solutions for the uh e-commerce which help to improve the uni economics and build the profitable uh e-grosity uh business uh this is a quick overview of our full technological stack so uh we offer the white label Client app so basically everything what is uh related to the uh getting those orders and processing those orders as well as we have a product information system which basically feeds all the information product pictures ingredients etc etc into the client app and it could be our client up or our partners Client app or aggregator Client app we have already made Integrations for their uh different client apps we have pick it up and WMS system which automates all the procedures and all the operations inside the store and of course we have the last mile solution uh which runs uh all the the couriers so our inventory system uh Works in real time so basically you can see every single item in real time and all of them all operations are fed into our pick it up so that the pickers don't need to think much just follow the instructions or fully automated system the questions on this slide are very very welcome but I think we'll progress further to the unit economics now basically what is the unit economics uh What uh what is that uh so uh basically that's the calculation of how much money you make or lose uh per order so uh really the revenues less the cost per order or per store or per any other unit and uh I guess uh everyone on uh this uh call knows that why that uh uh this is uh important uh oftentimes uh uh companies uh the the unit economics become the business becomes more and more efficient with the scale right so many companies uh and especially the in the EG grocery segment they tend to burn massive amounts of money without really knowing when they are going to achieve the scale where they will become profitable and the cash burn becomes uncontrollable which is pretty dangerous uh for the company we of course have already achieved our profitability so we can relax a little bit and for those who want to increase their profitability or who is now uh still uh not profitable uh this webinar uh could be helpful and of course we're happy to answer further questions on how to do it so the main reasons is uh basically to make business uh uh profitable now uh What uh the unit economics of the e-grossery or of the dark store which delivers orders to the online what it consists on and this is of course a simplification we have much more complex model which we are happy to discuss like on one-to-one basis uh on the later stages with everyone uh interested uh here is the simplification like you're raving you basically the money in uh what it consists from so the average order value so your basket size this is your absolute Top Line and then the delivery fee how much you charge your customers for the delivery fee right so those are the only money which you are taking from your customers of course there is back margins and uh some promotional uh money coming from the suppliers vendors and uh Etc but we won't be covering this uh in the scope of this uh webinar now everything else is essentially what takes money from your p l uh which is the cost of groceries basically how much you buy those groceries which you sell afterwards the cost of fulfillment how much uh every single uh Peak or the Picker or how much it costs to prepare the order pick back the order and then once the order handed over to the delivery driver to deliver it to the customer how much it costs uh per delivery so and basically uh the revenue minus the costs uh equals your direct contribution so uh from here I will uh open the very simple uh spreadsheet which helps to um uh understand like how it works uh in uh in practical in Practical terms I'm just looking at the chat Window Guys and uh if you have uh any questions whatsoever uh feel free to post them and I will be very happy to answer them at the end of their uh webinar uh I see now that you can see uh my screen which is a simple spreadsheet so we'll try to build like a simple p l uh simplification of the p l here now the first item of course is uh average order value so basically this is how much we sell our average basket for uh now the second item uh will be the delivery fee so how much are we charging for the for the delivery fee and this is essentially all of your Revenue so we can see that the basket is a major part of your revenue and delivery fee is a smaller part of of your Revenue now uh then you will analyze the cost and the basically how much is the cost of your groceries so how much you sell the groceries for it you answer this question you you will need to know uh what is your markup on your goods right so let's say for the sake of this uh example uh we uh markup uh 25 on the groceries and uh that will mean that uh on the average basket of 21 uh the cost of our groceries is uh about 17 right of course if uh our markup is higher then our cost of groceries is going to be to be lower right now the next line in our uh simplified p l would be uh the cost of fulfillment basically the cost of picking so uh how do we calculate this uh so we take our typical dark store or gray store uh whatever the facility or the microfulfillment center whatever the facility you're fulfilling the orders from from so where you are doing the pick packing so let's say that store is doing 200 orders uh per day and uh you have like some pictures on that store and depending on geography your cost of Labor in that store could be different let's assume you are somewhere in the Middle East that means that the cost for business is anywhere from three to three and a half dollars uh per hour of the thicker labor if you are somewhere in Africa that probably would be uh 1. 3 1.

6 1. 7 dollars per hour uh so in different countries it's uh different so how many orders can one ticker pick per hour so let's assume that he can do five to ten orders whenever in uh our Benchmark stores uh the time your item is anywhere from six to ten uh seconds per SK even on the bigger dark store so that's really fast so if uh you you can calculate how many or how many items is there in your average basket and from there you will determine uh how from there you will determine what is the average time per basket and uh from there you can determine how many picker shifts you need uh on your store so let's assume uh that to cover the day uh all the store which is doing 200 orders per day uh you are doing you need eight shifts right so that's eight people coming for whatever uh so I mean you let's assume you need five shifts of eight hours so let's assume you have the shifts of eight hours and that means you you will need like uh five people at different times of the day to cover the uh workload of the store uh so total uh picker cost per uh day then uh will be about uh 120 dollars through the simple formula and uh then uh cost per order of fulfillment is going to be about 60 cents right so this is really a uh really effective uh uh really cheap so on every order you spend only 60 cents uh for the Fulfillment for the picking uh for the sake of this simplification we are not looking at the cost of uh keeping the warehouse utilities etc etc now the last line which is like uh very significant uh is um and which is very different so basically what makes the e-grossery different from the uh conventional retainers is that you have the delivery which uh you don't always have in conventional retail business so how do we calculate the cost of delivery that's a very simple uh uh formula we take the cost of Courier per hour and the cost of Courier per hour tends to be higher than the cost of picker simply because they have motorbikes they need more skills like driving driving license Etc and this work is actually harder than the work inside the store because sometimes you need to be in the heat sometimes in the cold so it's uh usually more expensive then how many orders uh per hour The Courier can deliver and this is one of the key uh operational metrics which uh you would always strive to improve uh because the business cost per hour uh of Labor of Courier is fixed so if you can get this cooler to turn around more times that means that you can split his uh cost of Labor between more orders so that means the higher this number the lower going to be the CPO so in this particular example uh the CPO uh is going to be 1. 86 dollars but imagine if you can get this Courier to deliver not 2.

2 orders per hour but let's say three or four orders per hour and then your CPO will drop significantly so you will save money on that so now uh let's have a look at uh your uh direct contribution so basically we will need to deduct all your costs from all your revenues and in this example our direct store on 200 orders per day is already profitable of course there is going to be some other cost uh but that means already 12 of your total revenue is actually your contribution so uh this is a the example of the direct store model and guys please send in the chat if you have any questions I don't see any questions yet uh I hope everything is clear and everyone is uh finding this uh webinar useful um um I'll come back to the slides now uh uh and uh proceed with the further uh explanations and of course this simplified model and more complex model we can share with you in the uh spreadsheet uh after the webinar please feel free to send us the direct message so what are the strategies for achieving like positive unit economics in operation and in Commerce I will be covering more of the operations part and the Alex a uh will take over after to cover the commercial part now uh the cost uh per order of couriers so how can we actually improve uh this metric right so uh what is our levers what are our methods so number one is of course routing so we can send our couriers to the most uh optimal uh route dispatch this is the time of how long it takes uh for uh for example for Courier to uh get to get the package from the pickers so here our very sophisticated technology comes in place because it's literally instant sometimes couriers need to kind of wait when the aggregator app will receive the uh uh uh uh location and the parameters of the mass of the order before they can collect it those things they come together uh basically routing dispatch and batching uh batching is when you actually assign few orders to the same Courier so if your orders are uh located close to each other that means that uh you are on the way then that means that without losing inefficiency in uh without losing time The Courier can actually drop few orders on the way and once you hold in real time the information about all of uh the allocations of where the orders are coming from also once you have in real time information about uh what are the baskets being filled up at which locations uh and when they are going to be checked out you can really be very intelligent and wary you can use very smart predictive algorithm to batch orders in most optimal optimal manner so sometimes it's better for The Courier for example sometimes it's better for Courier to wait for one minute uh longer to get another order received and packed so that he can actually drop it on the way so let's say for example you don't have enough uh you don't have coolers in the store right now and you have one order ready but you know that the other order is being packed or the other order is being checked out now so you get your Courier to wait for another minute so that you can batch another order which is on the way to that one and that is the one of the very important ways of optimizing the performance of the coolers and oph orders per hour per quarter now um you can of course use own couriers or you can fall back on the third party Courier so uh typically the uh we have like the peak times so the demand for the groceries or whatever your uh distribution from your micro fulfillment centers is not uh flat so uh it's some hours of the day you have more uh demand and other hours you have like less demand and then it might Spike again so usually the way how it looks like is about lunch time you have the spike in demand then before dinner it's less demand and then it's again higher Demand right and while The Courier shifts and you don't always have like enough couriers to match this uh this demand so usually uh the sometimes you would have more couriers sometimes less Courier so you might want to have uh just enough coolers to service this level of demand and the peak demands you might want to fall back into the third party couriers or into the taxi applications or something like this but there is like even smarter way to uh offset the Peaks uh and much the supply and demand and uh I will I will show this uh in the next slide um in the this is the example of search we're all familiar with the search term from Taxi business when basically there is too much demand the tax Affairs are higher than usual and then you have a choice either to pay more or to wait for the fare to go down and this is exactly the same concept which we're using here so once we have too many orders and not enough couriers we can introduce the delivery fee or the minimal basket fee and that means that some of the customers will choose to not place the orders now but we'll choose to place the orders later so we will offset this demand to later and here is the graph of how it's working the Practical terms so you have certain amount of orders per hour Target you have certain amount of orders per hour and then you have the spike in demand so you need to do more orders per hour right uh and then your Implement search which is the delivery fee or the minimal basket so certain percentage of orders are becoming search so you make as a business you make more money on those uh on those orders so and we look here at the result and what happens so there is less uh canceled orders haven't changed so you see it's like the same the same level of canceled orders as before we apply the search and at the same time our delivery promise is no more than 15 minutes still so we actually have done best of both best for both parties so Source customers who are not uh price sensitive they still place the orders and those who are price sensitive they Place their orders after so we satisfied both types of our customers as well as made more money for the business and this is one of the superpowers of our uh career management tools and definitely prerequisite for the profitable business so uh the other uh levers here are the shifts which is pretty simple make sure that uh your uh Courier shifts are as close as possible to your demand kpis where you pay bonuses to the couriers to work better and faster there is the whole science behind it of how to manage your uh Workforce and how to continuously improve them uh the dashboards like the ones which I demonstrated on the on this screen which your preparations team can be monitored and can be improving and can be can be monitored and controlled in the the business of what happens so those are the some of the main tools of improving The Courier efficiency and of course don't forget to ask questions uh now this one is the Fulfillment so how do we improve the uh uh fulfillment how do we uh how do we get our fulfillment more efficient so uh the Logics is pretty similar the less speakers needs to pick back more orders so the same is less couriers needs to deliver more orders same Logics in the fulfillment in the Fulfillment center so number one I think is the real-time inventory what does that mean first is that whatever the client sees on the app is exactly what we have in stock so there is no discrepancies and there is no missing items so your pick Packers are not going to be wasting time calling the customer asking oh can I substitute Coca-Cola zero for Coca-Cola light because the customer who is exactly will exactly see what you have in store and if you don't have uh one of the items in store then it will simply be removed from the client app so it wouldn't be possible to order it now install routing we send other pickpockets through the most optimal route inside the store so they never need to go to do the circles because we know which what is on which shelf so with your pick packet will only wait one way to the most optimal route and save himself uh time we do have a robot fulfillment which could be uh very handy for the economies where the cost of Labor is high so that helps us to actually almost bring the cost of Labor uh it it could possibly make the cost of laboring uh uh developed countries same as uh in developing countries automated tasks uh is a very important part of our WMS and however pick it up that uh what helps uh that means that your pick Packers they don't need to think much about what they need to do for example when the item is coming up to the expiration the automated task will come to the pick it up and we'll send the Picker to a particular shelf to take this item and write it off we also have the automations like where we for the fresh items where the Picker will see the pictures of the apples for example and number of stains on the Apple which are acceptable and not acceptable as per your company policy so if there is two stains it's acceptable we keep it on the Shelf if there is four stains or it's a little bit yellow we write it off but not only that also we don't actually want to write them up so just before the expiration date we would automatically send the item on promotion and try to sell it uh to sell it off so this usually gives a uh 15 to 30 decrease in uh cost of uh fulfillment uh and uh one uh the other like large labor is uh actually our uh fulfillment optimization or the Distribution Center optimization this uh is uh uh gives you the savings of like kind of three to five percent of the gmv this is massive on large scale so how does it actually work uh the usually you would have like the delivery schedule from uh your uh suppliers and you would have the minimal order quantity from the suppliers and oftentimes what you actually want to have to replenish is per your demand is different to the what your supplier can deliver because you could possibly use like say uh 20 bottles of Coca-Cola and the minimal order is like uh say 40 and they can only deliver once a week uh so as we see the inventory in real time we connect our artificial intelligence algorithm to it and we know exactly what is going to be consumed when and then the automatic internal transfer or the purchase order is being sent to Distribution Center or to the supplier making sure that you will replenish just as much as you need just in time and after some time the algorithm actually learns to predict better than any human can do and on large scale that's to be honest the only way to run the business and I'm happy to dive deeper into the logic service and how it works but now I'm going to hand it over to Alex a to talk about the commercial uh that many topics in operations thank you a lot let's see what we have in Commerce um I want to start with Commerce incentives basically These Are the discounts which we give to our customers here we have two big levers to improve commercial incentives in particular to do to decrease them to the level of one to three percent of topology first of all based on our experience we think that Dynamic pricing and dynamic discounts work well for example you can give a discount for close to expiry date products or you can introduce some games for example discount in the evening or instead in the morning so this helps a lot to drive sales and at the same time don't give discounts to everybody for everything but be smart in the sense also we think that Banos can work quite well bundles how do they work so basically you give a discount for example ten percent once you buy two items of specific Goods if you don't buy one of them you don't get the discount in this sense again you drive sales at the same time you can play smart uh to introduce this discount and drive um total level of incentives down I think we can move next yep front margin actually I'm front margin basically this is the difference between um the price you sell to the customers and the price you pay to the suppliers and for sure this might depends on the bargaining power um and quality of Assortment bargaining power with your suppliers so you need like economies of scale at the same time to introduce quality assortment to introduce the items which actually have white high front margins we need specific tools for the analytics of your category managers and um we usually on a daily basis I would say even on hourly basis direct performance of every SKU on every store to make mindful decisions about which skus we should keep in our assortment and we should be replaced for example we even build cluster clusters of Assortment in specific part of the city to introduce these sqs and other parts of the city have been through these different Studios so to drive front margin and based on our experiences could be up to 45 percent total we use live commercial p l reports with high Precision of data to track performance of every SQ um on the next slide I also want to mention writers um writers actually could be quite High especially if you focus on price good pressure certain which is very important to that to drive retention to drive a frequency of your customers that's why it's very important to control write-offs in the sense what actually we do uh we think that demand planning based on artificial intelligence that is outer replenishment is super super useful uh there are different algorithms and actually they can decrease your total rate of to less than one percent also Dynamic discounts so as I mentioned before um for close to expiring the products you also can introduce discounts to drive sales and actually get rid of items which are soon to be expired um and also let's talk um a bit about average order value super important quite difficult kpi to increase and uh there are two big levers here for average order value first one is number of rescue use number of unique items you keep on the shelves to provide to your customers is a circle so here we observe the following effect so if you imagine Corner Shop with average assortment approximately 2005 100 items and if you increase this assortment by 100 new skus you can drive your average order value by one percentage point at least so this is just very modest estimate two uh but our space in this story is quite restrictive that's why actually we also use demand planning tools based on AI and this helps us to keep less items of the same SK in the store and put an address to you also we employ some shelves layout with improvements to make shops hmm feet in the store in a better way and it also saves space so we don't have over stock we don't have back stock at the end of the day you can increase [Music] average store by 700 SQ unicorns if you do it the right way based on our experience and another lever which drives average order value um is on shelf availability this is the next one uh on sale availability this is basically whenever your customer opens your website or app uh he or she can find a specific SQ and if you say for example that 100 USA is 100 this means that anytime person opens app or website he or she can find a rescue which you have in your active assortment um we also observe that this is quite important uh to drive many metrics and in particular to drive average order value uh by 0.