# Why 90% of Membership Programs Fail (And How to Build One That Doesn't) | Interstellar ft.Anuj Rathi **Creator:** Stellaris Venture Partners **Platform:** youtube **Duration:** 44m **Source:** https://www.youtube.com/watch?v=XzNN96cMdpg&t=3s ## Summary Building a successful membership program is complex and requires strategic planning, clear objectives, and an understanding of customer behavior. The discussion emphasizes the importance of defining desirable benefits, understanding pricing strategies, and maintaining a long-term view on profitability while continuously adapting the program based on user feedback. - Membership programs are akin to startups, requiring a clear strategy and understanding of customer needs. - Successful programs must offer attractive, predictable, and profitable benefits tailored to specific customer cohorts. - Pricing strategies should consider customer behavior, with a focus on minimizing churn and maximizing retention. - Companies must think in terms of long-term horizons rather than snapshots to assess the impact of membership programs on P&L. - Differentiation from competitors is crucial; experiential benefits are more memorable than simple discounts. - Flexibility and the ability to adapt the program based on market feedback are key to success. - Founders should assess whether their product can support a membership program before launching. - Examples like Swiggy and Amazon Prime illustrate the importance of integrating membership into the core business strategy for sustained growth. ## Transcript [[0:00]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=0s) I never realized life is so complex building a membership program. >> Membership programs are grown-up companies again trying to become startup in their own right. The first thing is the ability of the company to package benefits that can be given to users that are desirable, predictable and profitable. When we had like no MOV, minimum order value. Many users started ordering papad and samosa and like 19 rupees things for free delivery and so on completely unprofitable. Generally Indian consumers are bit value conscious. How do you go about defining pricing as in what is attractive enough for a customer? [[0:27]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=27s) >> We cannot think of P&L in snapshot. We have to think of them in horizons. What we started observing was that even though 100 rupee coupon is like a more valuable coupon would still choose the option of free delivery coupon plus it is kind of a pain point. A 100 rupee coupon is a delighter a free delivery coupon is like a almost like a painkiller almost like I hate paying delivery fee. I'm removing this thing >> instead of spending a c of,000 rupees on a new customer. Why don't I create a membership program between the two the repeat customers lesser negative on unit economics. [[0:53]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=53s) >> It's the strategic clarity of the founders or CEO. This is the shape of P&L customer cohort wise that I want to see over the next 1 year or 3 years. >> Let's talk about a regular startup founder. They are thinking that I want to start a program. How should they think about whether it's even a good time to start a program? Do you have or can you create a set of benefits that make people turn? If you pick any of those things that I talked about, each one of them becomes like a recipe for failure if you have not done it right. [[1:23]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=83s) Welcome everyone back to our interstellar podcast where the objective is to talk to people who have played a key role in scaling up some of the startups that we've all been part of. But an interesting anecdote is that a lot of these people by the time they decide on coming to the podcast and the time we release this episode also become founders. So today I have with me Anoj uh you know Anuj has an illustrous career. He spent time in really well-known companies in India. Started with Flipkart and then became part of Walmart. Head did product and multiple roles out there. Also was one of the you know I think the person who started Flipkart first. [[2:02]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=122s) >> U then joined Swiggy spent seven years there again as head of product and then multiple roles on the business side instrumental in launching Sugi 1 as we know it today and then uh CBO at Clear Trip and as I said also now a founder in his own right. So I know first of all thanks for coming on the show. >> Thank you so much for having me Rah. Yeah. And the objective today which is you know as lot of startup founders when they're building a company there is a recipe to grow the business which is to acquire new users but I've seen time and again people don't have a very defined way of knowing how to retain users right and loyalty membership these are all ways to retain customers at least that's my understanding >> uh you have been part of building one of the more successful membership programs in India the Swiggy one and idea is for us to hopefully gain some knowledge from your experience. Uh so let me start with the controversial statement uh which is we were having this chat just before this podcast that 90 95% of membership programs fail. [[3:06]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=186s) >> Mhm. >> Uh can you explain that statement? Why do you think that happens? >> All right. So you started with the damaged question and let me if I um explain that all if I had to ask you how many what percentage of startups fail would that number be in the same ballpark? >> Yeah. Yeah. So, so membership programs are grown-up companies again trying to become uh a startup in their own right. It's startup B2.0. So, let me actually start by defining what what membership programs are, right? Like um and what does it need to succeed them? So um the ex okay the leadership or the executives in the company need to have uh a set of predictable benefits that can be packaged, positioned and sold to a set of customers in the target cohorts to drive change in retention, loyalty, frequency and a bunch of other uh changes. [[4:03]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=243s) um and the ability for them to see through it as the program evolves. So all of this is required to do that and if you pick any of those things that I talked about each one of them becomes like a recipe for failure if you have not done it right. So let's start with first does your company or your product even have the recipe of of building something that can be packaged together uh and served to customers that is attractive to them. That's point number one. B is even if it is attractive is it attractive to the right cohorts of customers that you have actually uh that you want it to be attracted uh attractive to C is can you run this predictively uh >> can your benefits be given predictably and is are the benefits memorable and easy to tell others but the most important one is can you run this profitably >> interesting >> so you can you may be able to run this predictably so I've seen both kind of programs one is if they are dud on day one because nobody loves it because the benefits were too underpowered or they're so good that everybody came on your program and now you're so close to death because you're running out of money. So, uh those ones so that's kind of one reason why it fails. The second one is uh in my opinion companies which start so companies which start with say understanding of pen in a snapshot way. [[5:23]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=323s) >> Okay, March this will look like that, April it will look like that and so on are usually not good candidates for starting anything like that. companies when they're ready to think of BNL in not in snapshot but horizons >> that in AMJ for this Xcross Y cohort users who are doing like this AOV frequency and behaviors because of my membership program I will acquire and retain them and they will start doing something else uh so who can actually build that snapshot that horizon in the minds they are the right ones so I've seen companies fail because they are their in inability to do that the third one is around um just the membership programs are going to be a drag on your P&L in the beginning >> by definition. They will turn your most profitable customers into your maybe most unprofitable customers if your P&L already does not support it. So, so we need to define very different metrics on LTV like say payback periods and uh retention smile curves and so on to be able to run that well. So there's a lot going on there. [[6:26]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=386s) >> But the last one that uh I want to talk about is um companies sometimes not only knowing the objectives of the membership program that they want to do uh and hence either choosing the completely wrong model for you. Should you go for the subscription? Should you go with a membership earned versus paid? Should you go with coins? Should you go tiered or not? All of these are variables. >> Interesting. So that leads me to my next question. And for the benefit of the audience, if you can just list out different kind of membership/ loyalty program that you have seen >> with maybe objectives which kind of you know what kind of businesses do they fit in well versus not if you can just lay that out structure that will be useful >> definitely. So uh at the at the highest level I've seen three kind of uh of programs and and they drive different things while we'll come back to like what kind of companies what kind of businesses should choose what kind of uh programs like the the most obvious one is subscription >> like every B2B company has subscription like at the highest level it is essentially what it's a predictable um you know promise that you make to the uh to to the other side of the uh of the audience. It's a predictable promise at a certain fee but it does not give you any pride of becoming like a member of of something right like it's just a transaction but it still has to follow the same things what am I giving you what are you giving me uh and in different tiers and all of those things etc right like that's lion and there's membership u membership programs are known uh for example Amazon prime people know that I'm a prime member I want to ideally that's what Amazon wants me to do I'm a prime member and here are vets associated with that or uh >> I am um uh you know Hari bond Y gold member. [[8:05]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=485s) >> So that's kind of a a membership program and so on. So so membership programs come in two ways either it's paid membership or earned. So those are two variables. The third one are loyalty programs. Loyalty programs again can be the the core currency they are driving is actually coins or points or things like that or miles things like that uh rather than money that you need to predictably pay. So those would be like the the top three uh kind of programs right now coming to what kind of businesses choose or should choose what kind of programs there's a lot of literature out there but if you are fundamentally a low frequency category itself low frequency and high AOV category like say airlines >> ideally you should go with loyalty point programs because you do not want to drive high amount of uh frequency there but I'm not going to change the number of times that I'm going to fly just because I'm an Emirates uh member But whenever I think flying uh to the choices of >> my first preference should be uh Emirates. So so even when I'm searching on make my trip or Xedia or where >> if an Emirates comes it's just driving preference because I'm never Emirates is not the destination. So it's not even trying to be that >> uh or a Marriott for example if I'm a business traveler. So I'm not maybe I'm traveling 12 times a month a year in uh for business. It is not trying to make me travel more. Um so and and of course like if if my entire use case is about say 25 times Mar is trying to move preference towards that. Okay. So that's one um but if I'm a high frequency category like u e-commerce >> mhm >> Amazon or I say swiggy like that's where I build swiggy one >> here I'm trying to basically like a increase frequency uh and get all of your category use cases to me. Wallet share. [[9:51]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=591s) >> Wallet share. Yeah. And um so all your frequency means like there are competitive categories. For example, say Ola, Uber, >> Vicki Zomato. >> Uh in those cases u most people are dual apps. >> Yeah. >> Uh like they will do this and that. They will have their favorite app and so on. But you want to dramatically move uh the preference to you and that's where a membership program makes like a lot more sense and especially to your cohorts. [[10:17]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=617s) So, so those are the are the core constructs. On top of that, then then you get into like uh other point is should you have it in a tiered way or not? >> For example, Swiggy one has a Swiggy one light and then there's a Swiggy one or for example um your u Maron it has some seven levels of of programs and so on. So those are some other ways in which you can look at that level. But if you're an extremely low frequency category like real estate or or something so you should obviously not have anything like that. [[10:47]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=647s) >> And then you have you know these outlier examples you talked about Costco right? Uh when we were chatting earlier >> but the model itself >> I was I was going to going to come to that there are there are companies which are built ground up uh thinking membership uh program that's the that's the way that they that they feel that in a very crowded category. Okay let me take one step. Most membership programs are built after you have found a PMF, after you have gotten the growth funnel going, acquisition is sorted, activation is sorted, retention is kind of a challenge. CEO will go back and say how do I get that? Some companies they start with this is going to be my acquisition strategy as well as activation as well as retention referral revenue strategy all tied to the membership program. So two examples right like one is Costco like of course the entire company's P&L is based on like us running a membership program and that's the only me the membership fee is the only thing that earns us profits. [[11:44]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=704s) >> Um the other one notably is uh uh Door Dash. >> Mhm. >> Uh so Door Dash started with so in a crowded marketplace where everybody is doing all sorts of marketplace businesses but they would tie up with restaurants in a in a different way. So there are some restaurants that will say you are on Nash pass zero, you are on dash pass one and uh two but if you are going to give us 30% commissions uh then we are going to like make sure that your uh listings show up higher we will give you like 3x more frequency of orders compared to the rest. [[12:15]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=735s) >> So they have a supply side program as well. >> They have both side program. >> So dash pass is the unifying theme between um supply and demand >> and and that is fantastic like most companies don't think about like building two-way programs. At Swiki also we try doing that. M >> it's hard to pull off but if you are able to pull off those companies then have inherently core modes of retention at at both sides of the marketplace >> another third one where you didn't start that way but you very quickly became that kind of company Amazon is like that kind of company like prime uh if you go back to 2003 or something when they had launched prime I don't remember the exact year but from the that day onwards it became one of the three strategic pillars of Amazon itself and it's not like some program slabed on top so your entire company works the membership program backwards. [[13:01]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=781s) >> Interesting. I hope we get some of those kind of models in India as well. I don't see enough of them today where membership is the the default. >> No, no, but you do uh you know if if you look at uh quick commerce today. >> Uh what percentage of orders do you think that Zumato or Swiggy get through their members >> members? Uh it's it's north of 75%. >> Uh Amazon India like gets uh possibly 55 60% orders through Prime members. So they are hidden in plain sight but they are absolutely there and they uh completely govern the business models of these companies. [[13:34]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=814s) >> So let's go deeper into specifics now starting with your journey of Swiggy >> and Swiggy one that you built out there. Do you mind taking us back to what era was that like what was happening why did this discussion even come around what was the objective of this program that you launched at that time? So let me refer back to what I was saying. The the first thing is the ability for the of the company to package benefits that can be uh given to users that are desirable, predictable and profitable. [[14:04]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=844s) >> Um so Sri started experimenting with free delivery coupons and that's something that that we had understood from at least I had understood since 2010 when we were pricing at Flipkart. Uh >> there's like discounting on delivery fees. >> Yes. >> Okay. So Indians generally by and large hate paying delivery fees like that's a fact of the matter right and maybe some of that is changing. Um so if you had to run coupons uh say for example 100 rupees off 200 rupees off at the time we didn't have like a sophisticated couponing system but you run these coupons and give it to users cohorted on the hope that they will buy. [[14:39]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=879s) uh instead of that you could actually run a free delivery coupon and what Swiggy started observing like um was that users even though uh 100 rupee coupon is like a more valuable coupon uh would still choose the option of free delivery coupon which would cost us like the maybe like a third of like a a coupon and that's where we understood that the intangibility of that like but still users feel that I'm getting like a a good um this thing plus it is kind of a pain point a 100 rupee coupon is a delighter a free delivery coupon is like a almost like a painkiller almost like I hate paying delivery fee. I'm removing this thing. I I want to see my cart page clean like that. That's kind of the uh the way people think. So So Swiggy started experimenting with Swiggy Super >> and again like uh uh unlike what I'm going to suggest to everybody else but it started uh doing it as an experiment. [[15:32]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=932s) We launched it very fast in 3 months on top of free delivery like that was the only value proposition. This is what 201819 when was this? >> This is 1718. >> 1718. >> Yes. Uh yeah. So Swiggy Super became the program. >> Um so free delivery anybody who can come in 99 rupees per month. Extremely simple pricing free delivery for all orders. Now this is where now now the whatever battles cars are speaking. uh but but then you drive ideally this program should have been only for for customers that are high AOV high frequency already and so on and and so that like you are attracting those but then you you started seeing okay what are the kind of customers that getting attracted and even what is the good behaviors and bad behaviors that the program is driving so I still remember like when we had like no minimum order value >> uh many users started ordering papad and samosa and like 90 rupees things for free delivery and so on and completely unprofitable. [[16:26]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=986s) >> So, so that then a bunch of like these things kind of started um we started tweaking the program a little bit to for it to make sense across cohorts and so on. Um but in 2020 uh covid hit so so there was no delivery itself to be done the free delivery didn't make sense so it shut down for a while and then then we started looking at the program with a lot more holistic approach so soyg one is where the more interesting I would say part where like it was done in a very structured thoughtful way um now when you're thinking about membership programs either you will look at like for example objectives I want to drive these four strategic objectives and my program has to do them >> or you can start from the other way here are the benefits that I have what are the objectives it can drive >> the third variable is which kind of customers so that's why these are very hard to construct like just like in a product market fit you can move either the product or the market >> I have this product which market can it activate the best or I have this uh market what should I do to serve this so it's it is a zigzag between those and then you have to align with with one point which says key all right I'm going to build this for this product this market it seems like sufficiently good objectives design principles of the program and this market that we'll serve. So in Swiggy's case uh for example uh one key objective compared to for example increased frequency that was one of the core objectives ring fencing was like an important objective as well and again you can take a frequency objective to this also >> uh instead of increased frequency we didn't want to see churn uh so so if a lot of users are of high frequency users or medium frequency users of say four five orders a month are going to zero uh potentially because they have other competitors that they can look at and so on. So that program had has to address those. Um we also said that we were launching food commerce at that time. [[18:13]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=1093s) >> Uh and and this program's another objective was to kickstart this category into existence. Uh for example, that was completely different objective compared to uh for example increasing frequency. Um it also wanted to be a program that kickstarts city number for example 30 to 60 like some cities it is where we wanted to like get much better at. So we had to write like all of these objectives properly. you had same program servicing multiple of the because these are all standalone big objectives to be served right and >> yeah and now you had to create um a set of benefits >> in a very sharp way because now we are digital right like we can still tweak um very sharply define how if we run this in a certain way uh model it out how customers across different AO and frequency buckets will actually move so you have to do all that modeling so had done 32 different options of which programs Cohort what? [[19:11]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=1151s) >> 32 >> not cohorts cohorts were like more than 300. >> Okay. >> Uh >> 32 programs program options. >> Yes. Program options. And if we choose program option number one which objectives will be will actually be met which will will be not and then there will be trade-offs. For example, this option number one looks fantastic to drive category goodness uh high frequency ring fencing. All of that is are great but it's highly unprofitable and we can never get to a payback period which is good. This option two looks acceptable. uh but it's very hard to communicate so we don't think that it'll it'll fly and so on. So, so my job there was to actually help choose three divergent programs and help drive the consensus within the company to like essentially choose one program and then but once we lock and load then we actually lock and load because then it becomes we'll need to move money from marketing to this one or ops to the program and so on because it becomes a serious commitment for the company rather than just a cute program that you'll run experiment and fail. you know generally Indian consumers as we were just discussing they are bit value conscious how do you like the all these programs even Swiggy and you know zonato and others they're all paid >> programs versus why like this is a high frequency but maybe it's a lower AOV category is that the reason why it was a paid program or or how do they even go about defining pricing as in what is attractive enough >> for a customer if it's a paid program >> so now we are we are getting into uh another rabbit hole Um for any do you remember the price of Amazon Prime >> around 1,000 rupees a year. [[20:47]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=1247s) >> All right. So but do you remember the the price of uh Swiggy one or Zumato Gold or >> like I think now they're all in this quarterly 3 to 400 rupees a year a quarter >> in that range. >> It will be different for for different people. Uh so for example it it MRP can be 3,000 per year or say 1,000 a quarter >> or 300 a month but you will always find like some sort of coupon automatically applied or whatever to drive it down all the way to one rupee if if they want Rahul as a customer. [[21:21]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=1281s) >> Interesting. >> So so how are they priced? Uh they will be differentially I would not surprise promoted >> uh to to different customer segments. Initially many companies don't even have that capability. uh but but they out to do that especially so it has pros and cons uh the pros are of course then then you can manage your P&L much better and you are able to attract the right customers and and so on but it also has a con that uh customers may feel distinguished differentiated or whatever that unfair like hey why did you get it cheaper while I'm like a better customer and and and stuff like that uh it also has an negative that that the pricing is not memorable >> like you for example 1,000 rupees a month uh or 1,500 rupees a month for prime. [[22:06]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=1326s) >> Once you decided on the program for Swiggy1 in 21 and launched it, can you point out like few important consumer behavior changes you saw for people who either bought the program or did not buy? So one the first program was achieving its business objectives uh but the NPS was not great. So so people would buy it because the ROI seemed great. uh for them. However, uh the communication of the program was around that there's a select set of restaurants around which you'll get free delivery >> and c consumers didn't like that at all. [[22:43]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=1363s) Uh so for them the communic just because of the communications had to launch another V2 of the program which says all right around an 8 km radius every restaurants will have free delivery. It may be a dashini or it may be a fivestar it's fine because initially like we had tried to be a little too cute on which restaurant should be there and so on. Um anyway the behavior changes that that we saw I think we uh some fantastic behavior changes actually across that so there there was cohorts of AOE cross frequency and we had predicted people who are in four to five NU cohort then who had like taken 3 months now they're sitting here with this frequency how much will they move so we had modeled all of those things out but the the numbers if I could kind of uh say at a median level Um so there was a a cohort which said uh which was predisposed to buying something like a Swiggy one. [[23:38]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=1418s) >> Okay. >> So so we had that cohort their frequency has essentially doubled. Uh so >> this would be your like more valuable cohorts I'm assuming >> they are the ones who are already uh buying at high frequency and and all of those things etc etc. Uh so for example eight times >> food delivery uh segment that would move to 16 >> or for example people who were doing two times in smart they will move to four >> uh who had never tried quick commerce they would start doing that. So very sharp movements across those cohorts and and and essentially became like a almost like a too good to be true thing that happened. However uh if you are profitable company you'll enjoy a lot of those. If your if your if your unit economics does not permit that that also becomes a cause of concern because >> there is another completely different point of view to be made here that those people who are doing eight transactions uh anyway with you they were anyway doing it without any incentives. [[24:35]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=1475s) >> Why are you like now shoving money >> and to get to 16 all of them unprofitably. So how does that work? How you solve like what do you do like you change pricing you? >> No I think I think what needs to happen this is where the P&L question comes in Rahul which is we cannot think of of pricing uh we cannot think of P&L in snapshots we have to think of them in horizons. So for that membership program you do have to define a payback period like if this is a cost of acquisition of a customer into the program this is a membership fee that I'm going to earn today. So you also have to say I'm acquiring them at 1 rupee right now but in 3 months I'm going to ask them to pay 300 and then maybe a yearly fees of 1,000. Uh and so that's that's kind of one access the core P&L is UEIE is getting better. So so that needs to be factored in uh when we're looking at the payback period and I think first one it was like 18 months or I don't know the exact numbers that payback period is is the most important. If you do not have a payback period in mind before you started the program, do not start that program. Otherwise, that's the coming back to the first question which is saying 95% programs fail is because of this. So you start with a blockbuster damakar program >> uh and then after 6 months oh my god my piano can't support it. So now I need to devalue the program a little bit and then it becomes a mess. That's what happens with most credit cards also by the way. But slightly tangential question lot of time founders they're also debating whether let's say both my new user and retain user cohorts are unit economics negative so I'm overall >> negative on unit economics >> I have to still grow >> instead of spending a c of th000 rupees on a new customer >> why don't I create a membership program that maybe between the two the repeat customers lesser lesser negative on unit economics versus acquiring new customer right >> how how do you s how would you suggest >> so it is it is indeed a multivariate uh equation the way at least I I look at this uh in I look at like what what is my my P my P&L look like from a customer lens >> uh we just put like different kind of cohorts and we basically say which kind of cohorts is at what level of acquisition what level of retention and what level of uh economics >> and and essentially try to define is there in the future do I want like a um a flattened out P&L with respect to customers or do I want to see a fatty middle that is always there and it's it's more about like the the founder and the executives to basically discuss key do we want to go after like a high retention high repeat strategy high customer strategy or do we want to play the more acquisition kind of game and usually the answer comes high acquisition game because a it's easier doesn't require as much amount of brain sales or a strategic uh I would say money movement went from acquisition to retention because these things will play out over 18 months versus acquisitions like I can show it right away and my marketing really starts working um but but I have seen P&Ls where like it just made sense to to go membership first and not go the acquisition route first. It's just it's not one unique answer for every company by the way. M >> um but it's the strategic clarity of the founders or CEOs. This is the shape of P&L customer cohort wise that I want to see over the next one year. You would also like you talked about prime and swiggy they both have single program that span across different categories. [[28:03]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=1683s) >> That's right. Like Amazon will do commerce, video, music, I don't know multiple other things. Swiggy would do instamart swiggy food delivery and few other things. Whereas there are other examples like if you look at eternal right Zo they have a different program for each app of theirs >> like what are pros and cons? I'm sure there is there's good and bad in both cases. >> Yeah. Again I think you touched upon a uh very interesting I don't I don't know rabbit hole. So so first before programs uh did you also notice that they have different apps? [[28:36]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=1716s) >> Yes. Uh so they yeah for a consumer there like multiple apps that they're downloading and all right. So which one is the right strategy? And that again is should you have one app with multiple categories? Should you have multiple apps for example uh and should you have a unifying thread across those? For example, Flipkart. >> Uh Menra has a different membership program and Flipkart has a different one. But they have super coins which is kind of like common between them. Um so there are variety of ways in which you can do that. Uh RA and and and there is no right answer here. uh except the ability for the uh for the leader to understand does how different are these categories. That's point number one. B is is there is a is there a halo effect that I want of my program to fall on the other one >> and how hard is it going to be for me able to uh for me to be able to pull that off. For example, in Swiggy's case like Instamat was like the category genesis was the program in a way like like Swiggy one users were basically the the initial 90% of Instamat was come only from Swiggy one user which is very small cohort of the overall base of like the MTUs that were there. That's what the objective of the program itself was. [[29:47]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=1787s) Uh versus Blinket is independent app has its own identity and so on. it actually works out better for them if they stay separate from from Zamato because irrespective they work um and because otherwise your promises kind of become global >> uh so the benefits of having a unified uh kind of thing is okay x number of deliveries across ABCD things is that an easy communication so >> which one the next question is in if you do an program which spans across so many different categories then the program has to be very simple you can make it specific to that particular category, right? Video and I don't know food delivery are three different categories altogether. [[30:29]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=1829s) >> That's right. That I agree with you. Um so the simplicity of the program takes takes a big hit uh when you're trying to pack in too many things to one one simple program. But it's anybody's guess to be honest. There's no one right answer. Uh you have to like essentially think about it from long-term go back to like is this a strategic priority and is the objective of the program uh to drive category goodness into the other one. [[30:51]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=1851s) I've seen like some of these things happen um in um uh in travel as well. Uh so so a bunch of global even Indico for example is also kind of tinkering right now. They have flights they want to drive hotels they will drive it only with uh blue chips and and and so on. M and one last question on the Swiggy program like if you have to look back what do you think are the top three reasons why it succeeded as in if you have to just summarize >> I think the first one all the three that I talked about in the beginning um its ability to actually understand the and package the benefits that would really work for the cohort of users uh that it cared about and did not care about. Uh so I think that was definitely one B is Swiggy's ability to to kind of have a long-term view and a and a capital allocation logic plus uh strategic investment in the program. Uh if that if that already is not too high um then you are always caught in a catmos uh every two months you'll look at it like this doesn't make sense let's just shut it down while it is driving growth and so on. So that was uh I think the the the second one. Um third thing I would say is like just the flexibility like a it seems very rigid. Uh but just the flexibility and constant learning from the market and the ability the levers that you give to the program >> uh while changing uh I think that was very important and and very good. Let me take a contrast for example. So like I mean when when Amazon trend launched in the beginning free delivery for all products for all customers >> very simple. Now when your communication is simple you lock yourself into such a P&L. Now that cannot be moved. Now users will do what users will do. If it is too good to be true then >> maybe after so it took up res then to run a company along with that and and to still make it profitable eventually. It is a more powerful program definitely >> uh but it didn't have many levers of of movement. Um so so I think in Swiggy's case there was levels of moment for example pricing or for example the cities that it chooses to launch in and the cohorts with which it will like uh do like a lot more communication versus not things like that >> and like while globally you would see these programs to be more annual in nature in India we have this monthly quarterly like does that make a big difference? Oh, huge, huge. Now, let me give an example. Now, Rahul, for example, you are the Okay, actually, let me take one step back. [[33:19]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=1999s) If your program is too good, then everybody in your cohort of of users uh will come buy your program, milk it to the maximum benefit. Do you want everybody to milk your program with with the full benefits because that has operational cost and so on. So there always needs to be a spread of people who are coming and using it and they will be unprofitable. It's fine. They will be and some who are coming it using it giving you their loyalty but not milking milking the program. So that kind of for each cohort you have to define. [[33:48]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=2028s) >> But how does quarterly matter versus >> I'm going to come to that. So >> for example um if I if I sell you a monthly program >> then the number of times that that a user is going to come and do that ROI calculation in their minds is monthly. >> Okay. >> So they are going to milk it a lot more. M >> uh where if I get you in for a annual program >> then you're going to milk it less >> but at the same time it's a higher capital that you that as a user you're asking to pull uh >> so every company needs to basically understand their users should I get in get them in for a one month program and then upsell uh 1,000 bucks or should I get them in for uh for a yearly program and and never worry about it. So these will come only after like bunch of experiments as well as your understanding of the category and the user behavior in the category but all of them are extremely important variables Rahul that you have to consider plus there is churn also right as in if you do a quarterly program you don't know whether I'm coming back to you after 3 months paying again or not >> whereas I paid for four quarters you have me for four quarters as a loyal customer right >> so like what churn have you like generally historically what is a okay >> or a go a benchmark for a good Chun no okay okay so there is no benchmark for good chun so I think I don't think we should be like talking about global benchmarks because even in categories it is very very different if you look at even as a matter like the the medians as well as the customer behavior across those companies itself is very different so there's not even a category answer to this uh however let me talk about this imagine like a full x cross y of a cross frequency buckets uh say for example four to five uh a bucket in frequency and 300 350 to 400 rupees a now within that also there different customer segments and if you do a one month program to that you actually have to run this to say within this only 30% of the users actually moved to the membership program this is what the behavior that they showed and this is how they churned and a one-year program this is their behavior shift this is how they churned and this is how they were um uh for example their frequencies were increasing or a were decreasing and so on so the answer will come per cohort-wise that's the discipline that you need to have as if you're building a growth operating system so I wish I had like an easier answer to this. [[36:02]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=2162s) >> That's okay. That's okay. So, let's not talk about a regular startup founder, right? >> They are thinking that I want to start a program. And let's take an example. It's a high frequency category. Mhm. >> Uh you know it could be a like you buy chronic medicines if you are a chronic patient very regularly or you have a you know for a certain age group you are a parent so you have to buy a lot of stuff for your baby when they are zero to maybe 2 to four years and then after that you know the frequency comes down kids also become more >> autonomous if I can call it uh >> like what should they how should they think about whether a a good time to start a program. Um I think the number one thing that they have to think about and always in that sequence do you have or can you create a set of benefits that make people turn >> and say I love this uh in and the easiest answer usually is free delivery and so on but >> or discount some 2% 5% discount if you're a membersh member >> so so I'll come to that as well uh one is uh can you actually does your category lend itself towards that >> that the that the benefit is uh valuable, appreciable, unique to you, non-copyable things like that and can be packaged into something which is easily consumable by the user. [[37:26]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=2246s) >> So that's that's literally step number one. If you are able to do that, then you have to look at like what are the what is the shape of P&L that I can imagine. Can you think in snapshots? Uh and the the point that we were talking about do you think of yourself as this is going to be a year of acquisition of like X more customers or this is going to be year of high frequency >> their focus is to be figured out. [[37:46]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=2266s) >> Yes. Yes. So sometimes founders are too early to start this and and maybe they should not uh maybe they should start thinking about it but launch in year three uh when they have a good mix of acquisition and and high repeats and so on. Um so that that's kind of the second one. And the third one is they should enter only when they have high conviction about their strategic uh P&L shape >> that this is what I want to do because it will take a lot of decisions across capital movement the way you organize your company itself to run this where it's not only going to be a programmer slabed on top like it will go down to does your supply chain run that way for example Amazon primes deliveries are faster. [[38:30]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=2310s) >> Yeah. uh and then it can just work uh that way. In cases when when people have tried slapping it on top but not putting it to supply chain let me give an example say Ola select >> uh like some years back they had launched >> could you even promise a 2-minut driver to you could not so then you promise things like hey five times more pings to driver around you not does a customer care about that >> so your ability to deliver on that promise consistently also is important so here's my suggestions to the former um either that or if If you are a super smart founder who's who's just starting out, start out with a uh membership program itself. If you're that kind of founder, which is a few and far in between like a Costco or Amazon and so on. Apart from that, I would say think about it very deeply. [[39:18]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=2358s) >> Very interesting. Very interesting. And I'm sure a lot of people come to you for advice on this. In in reality, like what are the most common pitfalls you've seen founders fall into as far as membership >> concern? I think uh membership loyalty subscription I'm taking all of those three in the in the common umbrella. I think the most common pitfall is thinking of them independently rather than the full growth operating system of a company. >> Um and and this is one of the exercises that I do with uh with leaders also let's forget about your overall snapshot at P&L. Let's look at like customers flow through towards acquisition. How many of them are coming? How many of them are going to become dominant in like D7, D30, D90? How many of them have reached when do we call a habit has been created activated? So user activation >> how do you know that from this activation they will go to loyalty? Uh what is your insurance against against this if if some loyal people move out? [[40:19]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=2419s) Uh how do you know that or so the diagnosis itself and the tools that we create for that are not complete? Um and from here if people churn out what is a good churn, what is a bad churn? Uh so so all of those things need to be thought of together. In my opinion that is the most common pitfall like most people have just no idea because they're looking at it in snapshots. My BNL looks this way on a unit level on a customer level broadly MTU DTU things like that >> but not and how important is it to have a differentiated membership program from competition? [[40:53]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=2453s) >> Oh it indeed is very very very important. uh so so >> like I can't figure out between three quick commerce or five quick commerce companies you know what is the difference they all have something delivery >> and that's why they will all uh retain their own set of customers >> uh and and and not uh have anything differentiated for for one customer to like aha so usually what happens is people always start with value based programs and never experiential programs >> value based programs are the easiest to pull off because the entire uh supply chain doesn't need to be disrupted nothing needs to be done >> rich price >> rich price or discount. Those are numbers that you can just give >> uh and then take back also if you don't like but they are not memorable. [[41:32]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=2492s) >> Uh they cannot >> and and because you are able to launch it so fast anybody else is able to launch it so fast >> but if you are able to give something which is like actually deep into your supply chain a unique promise that nobody else can beat stuff like that that's when it's very hard to replicate um and it is also something that makes the customers go wow yeah this is special. M so that would be like you're saying next day delivery versus 3-day delivery. Yeah, I think I think I think look at which program comes to mind like say kupang their rocket delivery is is only a gupang wow for example and now everybody in in in in Korea is like I have to be here you know they're just so differentiated so it it needs to be differentiated enough from your rest of your product or okay how about it your core product a is it differentiated enough with your core product itself b is it differentiated within the category and the competitors so you have to think from both point of And Amj if I were to ask you I'm sure you have seen this you know a lot of global membership program if there is one that you really admire on how it has been built >> which one would that be? [[42:39]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=2559s) >> Uh it's a membership program but a tier program but a loyalty program. It's it's all the three together. I think Marriott Bono is is built very very well. M >> um if you look at the tiers uh just the way that they've understand travelers the way that they've understand different segments of them and just how integrated it is to hotels to their app everything is bonvoy now >> uh that's one um and you do get value benefits definitely that you can buy through points and all of those and then people kind of understand their points they've integrated with like all AMX this that everything etc. But even more importantly, experientially it is differentiated. So, so if you are a gold member, you will get early check-in. And again, the beauty is they've not promised it, but they will give you >> uh and if you are like at a higher tier uh in in in this one, they will actually at the highest tiers what I've uh what I've realized if you do 200 room nights per year something like that, they'll give you a private concage as well like in whichever city that you are in uh to be able to sort that out. So I have I have thought and given that Marriott has I think 40 plus brands. [[43:43]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=2623s) >> Mhm. >> Uh just how integrated it is. I think >> interesting >> pretty fantastic. >> I hope I don't get to spend 200 nights in a hotel in a year in a hotel. But this is this is amazing. Thank thanks for being here. I think we learned a lot of stuff. I never realized life is so complex building a membership program. But I think the fact that you need an objective, >> you need to figure out a cohort. you need to be really committed to launching it and then also measure the profitability. [[44:12]](https://www.youtube.com/watch?v=XzNN96cMdpg&t=2652s) >> I think that is the most important thing. I know you are obviously you should be saying that uh but even on the other side uh it's it's super important. No, >> thank you for being here. I'm sure a lot of people would reach out to you after this episode is released asking for your advice. So be ready for that. >> Certainly. Thank you so much.