Operating Like a Start-up inside the World’s Biggest Company — Ramit Arora, Microsoft

0:00Hello Vegas. Good morning. It's nice to see your lovely faces so early in Vegas. I hope you all weren't out drinking too late last night. This is a live taping, the first stop on the Sub Club podcast world tour. So, we'll be in London, New York, San Francisco, Berlin later this year. So, if you haven't heard of the Sub Club podcast, I talk to industry experts like Ramit every 2 weeks on the podcast. And

0:25so, looking forward to this chat today. Think there's a lot we can learn from one of the leaders of some of the biggest apps in the app stores. So, yeah, my guest is Ramit. He's a lead PM on monetization at Microsoft covering Office 365 and Copilot. Yeah, and thanks for joining me on the podcast today. Thank you. Thank you so much. So, yeah, I essentially am the growth lead for Microsoft 365 apps for macOS and I support growth for the mobile apps. So, so essentially the core apps, Word, Excel, PowerPoint. And also supporting the Copilot app and the Microsoft 365

1:00app as well. So, yeah. All of them big apps. So, so yeah, excited to talk about them and all views are my own and do not represent Microsoft's official stance. So, I'm supposed to say that. So. Okay. Of course. One of the things when you and I were first starting talking about having this conversation today that really struck me about our early conversations is that it really sounds like you and your team and even the kind of app team inside Microsoft is like started up inside a startup inside the like biggest company in the world. I always thought of Microsoft as enterprise and people go to the web to

1:39download Word and it's a native app on Mac and I kind of always thought the apps were almost an afterthought. But the more we talked, it's like I mean, Microsoft is very serious about building fantastic apps and monetizing them on the app stores. Yes. Yeah, so tell me a little bit about that kind of approach and how you think about growing an app business inside the

2:00biggest company in the world. Yeah. In the very, say, like 2013, 2014 era, these apps started as companion apps to the desktop apps. So, there is always this concept of desktop class productivity, which is that you have a screen, which is, say, an iPad or a desktop, and that's where work gets done. And then there's mobile productivity, where you have light edits. And then, over time, we started to see that mobile being the primary device a lot of customers. And this is more true in Asia and other regions than US, but mobile being the primary device and not a shared device. So, people have, say, in one family, they would be

2:36like two like everybody will have a phone and maybe they will have one laptop. So, we started to see people doing very long-form documents, like sometimes up to 50 pages of documents from their mobile phones. So, that's when we started to get very serious about this. And then the other trend that we saw was that because App Store has a very, very frictionless experience for buying subscription, we started to see that Microsoft 365 subscription, especially in consumer, like our personal and family subscriptions, and now Copilot Pro, keeps seeing this trend that people just want to make the purchase from their phone. So, it's kind of like how a lot of people buy from the

3:16Amazon app rather than from the website because it's just a more handy device. There's when your credit card is attached. And we see almost a 5x higher trial to paid conversion rate on mobile than on other direct channels sometimes because of very obvious reasons like this, credit card is already attached, and there is a frictionless experience, people trust providers like Apple, Google, Amazon to essentially manage their subscription. So, while we have a very large direct business, we started to see that App Store has this real potential in terms of very high trial to paid credit card already attached, people having this as their primary device. That's where

3:57they're managing a lot of subscriptions from. Yeah. So, those were some of the things and then we started to get very serious and App Store just gave us so much organic traffic that sure we pay some percentage to the App Stores, but anywhere you acquire you have to pay some percentage. So, so so basically the organic traffic just blew our mind and so we just invested

4:18and and here we are. Uh-huh. It's so funny cuz Apple's so protective of their percentage and I think people discount just how much it being so convenient is actually increasing conversion rates and leading to new subscribers who just wouldn't subscribe on the web. So, you hadn't told me that before sitting here on the stage that you had seen a 5x increase in conversion

4:38from trial. That that's incredible. Yeah, yeah. In some platform I mean obviously we we sell across multiple platforms. I think only on consumer we have like 10 different channels that we are selling on. So, we have other channels that also perform very well, but App Store is one of the highest performing in terms of trial to paid

4:54conversion specifically. Yeah, that's amazing. And how do you view competing on the App Store though as far as there are so many productivity apps, you know, so many notes apps, so many places that you can create documents and everything like that. How does a massive brand like Microsoft approach competing in this marketplace that is very different from the web? Yeah, I mean it has been a bit of a learning for sure, but one of the things I will say so competing on App Store you just have to operate like a startup because if you see any other distribution channel like a web or say even retail partnerships, those kind of

5:32things, we have some advantage there that okay, we have an exclusive deal with a retailer or we have a website that has say millions of users that are coming on on App Store it's very democratic. App Store I like to think about it like YouTube is, right? Like Paramount Pictures and a YouTube influencer, they're pretty much competing for the same eyeballs. So, we don't get a lot of advantage. Of course, there is some brand value that we do get. So, when we do our ASO, for example, we have branded keywords, so that's part of our strategy, and then unbranded keywords. But, App Store is unlike anything else in the sense that a

6:06startup can start in Bangalore and start taking the spot on some keyword and start showing up above word, right? We have to be like at our best game, and we try to retain at least the top three spots in most of the categories. And to your other question, uh being a massive brand, so think of this as like startups are kind of like department stores, and we are like the

6:30Costco. So, we have all the scenarios. We have I mean, if you look at our Microsoft 365 apps specifically, like it has Word, Excel, PowerPoint, notes, PDF, signing, and there's a scanner, and so there is so many scenarios that are there. So, from a value perspective, the value density is very high for our app. So, for like $9.99 for the family and $6.99 for personal, just the value that we can pack into one app, and those prices are obviously US prices. Some countries may be different, but like just the value that we can pack into one app, it can be up to 50 to 60% cheaper if you go and buy all

7:10those scenarios individually versus you buy like a M365 subscription. I know I'm giving a very department store type of methodology, but like you just see how much value we can pack into one app, it just blows, you know? I mean, if you try to buy, say, a scanning subscription, etc., etc., it will cost you a lot more than buying one from Microsoft. And the other thing that it also helps us is that a lot of productivity scenarios are very one-time use. So, like for example, in 1 month, you will have one scanning use case. In 1 month you will have one say two document use case. But when you

7:47combine everything together in one app, then your retention is maintained because, you know, you may have a scanning use case today, but you may have a document use case, then you may have a So, overall on a suite level, we see good retention because on a suite level there is that retention which which is maintained because of different different use cases that people have a

8:05use for. Yeah. So, with the App Store being so competitive, and I think this applies to any app these days. It's it's so commoditized. There's so many apps in every category. Like I have a weather app side project, and there's like 100 weather apps. And so, I'm constantly thinking like how do I differentiate, and how do I kind of stay ahead? As Microsoft, you've got way more resources to be able to think about those kind of things, but how do you approach that differentiation and adding enough value,

8:36and kind of staying ahead of the curve? There's no one good answer to something like this, but in general, if I have to talk about from a framework perspective, if you're trying to maintain a leadership position, and this applies to like broader than App Store, then you have to think about neutralization. So, this is a framework that one of our VPs used to talk about a lot. Like So, you have to keep neutralizing any competitor move, for example. Then you have to keep differentiating. Then you have to keep incubating, and you have to keep maintaining on So, for example, Apple releases dark mode, Apple release Vision Pro. So, we always maintain a

9:12leadership position in the sense that our apps are already on Vision Pro day one. We were day one dark mode. We were day one on most of the new innovations that come out. Like we are day one there. So, as soon as something gets launched in say WWDC, we are thinking, "Okay, how does it apply?" And then, depending on the stage of your app, like if you're a mature app versus a startup,

9:35you can like do different percentages. like now say we do 30% incubation because we are a more mature app. In In the past we were more bothered with like neutralization and differentiation, but now we can just do a lot more incubation because we have a base. And now we are focusing on, okay, how can we bring AI? How can we bring some of the more advanced features that can pretty much help people do complete document creation on their mobile, like they don't If you think about what AI can do

10:06these days, like you don't need 10 tabs. You can just pull all the information like one place and create any artifact like a document, video, etc. So, we are very focused on that. to talk more about AI, but before we move on, I want to re-kind of state that framework cuz I think a lot of apps can use this. Like when you're competing, thinking about neutralizing, so, you know, when you see another app come out with a very differentiated feature, you might not like copy that. I mean, especially you know, Microsoft, but like as an app developer, finding a way to neutralize that in that, like find your

10:36own way to differentiate or or have some form of functionality that handles aspects of those use cases and then differentiate, make sure that you create new things that are differentiated, maintain that leadership position, and then incubate, like constantly be like talking to users, like doing user research, and trying new things. Right. And it's cool to hear that like, you know, again, the biggest company in the world thinks a lot like a startup and thinks through these kind of frameworks about neutralize, differentiate,

11:06maintain, and incubate. if you have an app on the App Store, it's not like a bond you bought and it'll keep paying dividends, you know. Yeah. You have to keep working on it. Like you have to keep at it and you have to keep thinking about the next thing and the next thing and the next thing because it's a pretty competitive marketplace. And the other thing that I would also say is that when you are thinking about these four things, I think it's very important to think about them from the lens of jobs to be done rather than like say this competition brought something and then

11:39we copy. We never copy. We genuinely first see, okay, is this job even important for the customer? So, I think that's super important to have a very clear even the largest apps. So, like even if you take a an app such as Word, there's a very defined set of jobs that Word needs to do for a user. Like people should be able to create a resume. They should be able to create a note. They should be able to write a book. Like I can list out 10 things that Word should be able to do or Excel. Like you know, you used to be able to create a

12:07spreadsheet, a budget, or invoice. So, if the app is able to do all of those things on any form factor, be it mobile, iPad, or desktop, it's super important that we focus on that aspect, that what are we trying to do? And then there will always be more So, we are a more horizontal like from product standpoint.

12:25We have a more horizontal solution. There will always be more vertical solutions. Like for example, somebody would have a very, very tight invoicing app which will have many, many things like bank connections, etc. I mean, Excel will not go that deep into something like this. But then of course, there could be add-ins that can support something like that. So, there is that horizontal versus vertical focus that you have to think about. You can't go too focused on one vertical if you're trying to build a more horizontal presence. So, All right. So, now I want to do want to get to AI. I think you know, so many

12:58people I mean, it's such a buzzword. There's so many AI apps doing really well and so many people thinking like what should our AI strategy be as an app? And I think too many people are throwing it in as a gimmick. And not like deeply thinking through like what is this new paradigm of AI mean for our app and how can we introduce it into the app in a way that actually increases the value and like people actually use it instead of just being like some kind of gimmick you throw in so that your title can say whatever whatever AI. And then Microsoft, it's like you get working on

13:37Copilot and then being at Microsoft, you've got like a front-row seat. But not only that, but you work with a lot of teams across the company, user research, and other folks to really understand how to take the best of AI and implement them in a thoughtful way that actually brings value. So, what's that approach? Like how do you take this new hot technology and actually make

14:01something meaningful from it? Yeah. So, so I mean, first of all, I am myself kind of new to AI. And secondly, whatever I will say will be irrelevant in 1 month because it's going so fast right now and things are changing so fast. But I mean, let's just hone down particularly into, say, generative AI. Like AI is a very broad term. Like

14:20there's machine learning, there's that. But like generative AI has a few things and that's my opinion. Like it has a few things that are interesting. I think one of them is summarization. It can really summarize. So, like meetings can get summarized, for example, or any document can get summarized. It can do an outline-based creation. So, like you can give an outline and and some of these things are already sort of available in Microsoft 365 now. So, outline to document is something. And it's also a change in habit that has to happen. So, like I think the features and the capabilities are there. Like people also need to now start to learn how to use

14:58them. There's a big emerging prompt engineering field out there. One more thing which I find is very interesting is RAG, the retrieval-augmented generation. So, at least in the context of, say, a document or a note type of scenario, you can potentially pull a bunch of information on a smaller device now from different parts of the web. So, like in the past, if you were trying to say author something, you're going to Wikipedia, you're going to like 10 websites. Now, you can ask Copilot to bring some of that information to you, and also reference it, and so on. So, I think retrieval-augmented generation is pretty powerful, and people are using

15:36I've seen many startups that are using it very successfully, and that could be a pretty big thing. And then, there are some of the other scenarios like translation, which has been around for a while now. So, I think it's very, very important that in the context of whatever your app you're building, how do these new capabilities help? Because there is a problem map, like how I talked about jobs to be done, and then there's a solution map. So, in the past, we used to have create, read, update, delete. Like, if you think about it, most of the apps were kind of like CRUD apps, and then databases came, Kafka,

16:12like big data came, and so capabilities increased, you know, you could process a lot of data. And now, it's AI, so I think it's an increase in solution map, but ultimately, it's on the developers on how they make it useful for the jobs to be done. Like, you can throw in a bunch of AI, like most apps just go and integrate with OpenAI's API, but like, there has to be some meaning. It's almost like saying that oh, I have an app, and there's this new database, like 10 years back, 20 years back, they were apps that would just do some kind of a read operation or a write

16:45operation on database, because databases were cool at the time. Right. And so, similarly, AI is cool at this time, but ultimately, the companies that win will have to find a very meaningful use case that uses these. There will be some companies that will not find a use case, because for some companies, summarization or or RAG is not relevant, and and that's okay. It's not the end of the world. I think there's a lot of hubris, right? Like, if I don't do anything with AI, my investor will not fund me, my career will go nowhere. So,

17:14I think that's a very scared approach. Yeah. The problem map and solution have to align. One of the things we talked about a few weeks ago was I just love the way you framed it. And this is the way I think a lot of people could think about how to bring AI into an app is that whatever somebody might want to go to their computer to do, how can AI help do that in an app? Like how can we bring more desktop class experience? A great example I've seen in the market is Photo Room. They do background removal. It's like historically that was like you got to go

17:47into Photoshop. You're not going to do that on your phone. You're going to do that on a desktop with a big monitor and you're going to have the tool and you're going to like do the outline. It was like a ton of work. And Photo Room came along and just makes it like seconds inside an app. And so I like that framework of thinking of like how do we bring these desktop class productivity into mobile on the small screen by using these new tools and these new capabilities. And again, like you said, it's not just about ChatGPT style large language models. It's also about using other algorithms and other technologies

18:19that are springing up around the kind of ChatGPT. So what are some other examples at Microsoft of kind of bringing those desktop class experiences into mobile that have, you know, shown to be really useful to customers? I mean, so far what I talked about like summarization, like meeting summarization, RAG like retrieval augmented generation, those are interesting scenarios and I mean, in consumer there are a lot of things that people are trying and some with more success, some with less success. Some funny example is like you can now create a song for your loved one and it will be like in any language. You can go to Bing

18:55and just tell it to create a song for you and it's a bit gimmicky, but it's a real thing. I I'm sure some developers will probably find a use case for something like that algorithm because that algorithm can do things. I don't think that right now any real songs are getting created by AI, but that day will happen. So, I think that that's a more consumery thing. On the more businessy side, I think yeah, meeting transcriptions, recapping meetings, those are some very interesting scenarios. Like, you enter a meeting a little bit late, you can get a recap of what has happened, your meeting notes are sent to you, task items are sent to

19:30you. So, I think those are like good examples of things that are actually useful because now you don't need that. And I think one more which is very interesting which I tried is writing status updates. So, essentially, I do my work in Word files, Excel files, etc. And then Copilot can actually pull a lot of that information and write a status on a given topic. So, if I can prompt it properly, it will write a whole status that I can then send out to the team. Or at least write 80% of it, which is pretty good. Like, it's a big time-saver. I think it just becomes

20:02super useful when it gets connected with all the apps that you are working on, and then it can pull data from your context because that's live data that it's pulling. And then it can really help you write content. A lot of people are writing books these days with AI. I think you may have seen like author plus

20:19AI. Those specific examples aren't going to apply to a lot of apps, but I think the thinking behind it is how you have to think about it. And a couple other examples you had shared with me were slide generation in PowerPoint. Like, being able to on your phone create a more sophisticated presentation quickly using AI. And then getting insights from Excel. Like, if you're on a desktop

20:43computer, what are you going to do? You're going to sit there with your keyboard and mouse and create complex equations, and you're going to reference web to figure out, you know, how do I do this exact equation in Excel? And now in Excel, there's capabilities to use AI to get some insights from that data. And so, the way to think about it is that like jobs to be done. It's like, what are the things that people would do on desktop that they can do in the app? And I thought those were some really good examples. The other thing I wanted to talk about is your approach to growth

21:12and the way you think about it. And again, I was like surprised when we started talking that your team just in the past few months implemented an MMP for the first time. And so you think, you know, Microsoft's got everything together and like the biggest company on there, they've got all the resources, they're super sophisticated, but like you're on the learning curve like a lot of the people here in the audience and who are going to be listening on the podcast of like, "Okay, we need to find new channels to explore." So, how do you think about growth and, you know, starting to really push as you have seen

21:41more and more success in selling these solutions to consumers on the App Store. How do you approach that growth? I think one of the things which is and I'll specifically keep it to sort of mobile and Mac and App Stores for this talk, but I mean, we see the App Stores as the channel which can bring us a lot of organic traffic. And it can enable frictionless in-app purchase, for example, right? And growth that we are doing pretty much what all of the other people are doing. And you're right, like we're not as advanced as one would think. Like, so we have a pretty good ASO story in general. We have user

22:17research that kind of informs us what kind of scenarios are important. We also get a lot of like Gen Z trend and those kind of things. And then we're able to take that and put it in our keyword optimization. And then we also have a jobs-to-be-done approach. Like we know, okay, we think that, okay, these are the 10 things that we really want our apps to rank for. And because it's a very broad app, we have over time also become good at, okay, what are the things that people are willing to pay for. So, I think it's important to know that and I'm just going a little bit deeper into

22:50keywords. Like, what kind of customers do you want to acquire? The ones that will actually find value enough that they will want to pay for the premium subscription. So, we are getting better there and then we are integrating with an MMP. I mean, I think one of the things that happened in the 2015 to 2018 era, we got MMP, we got an

23:11engagement platform, etc. And we kind of misimplemented some of that because one of the things I mean, it wasn't it was a combination of a lot of things. So, what happened in that time like privacy laws changed a lot like GDPR came along, and we are under more scrutiny than any other app. And we have enterprise customers, so we sell one app. It's one app for all customers, enterprise and consumers. So, I think the implementations they were not done in such a way that it could handle GDPR and handle and and also like a lot of the software companies, most of them in this conference like they were also

23:46pretty young back then, and the solutions weren't very advanced. So, at that time we had an MMP, we had an engagement platform, and one of the things that we kept facing was actually two things we kept facing. One was privacy issues because privacy laws were changing really, really fast. And the second thing that we also faced at that time was just like value that we were getting from these software because we were paying these big bills because we had a very big monthly active user base even at that time. So, we were paying these big bills, but we were not really using it because on one side we

24:20were just figuring out the products because at that time like the products were not at a place where you could start doing optimization. So, optimization comes much after your product market fit is there. So, we obviously had product market fit in a big way on desktop, but it took some years to kind of get that same stickiness on mobile where we saw, "Okay, our mobile apps, our iPad apps are now like at a point where we can now start doing optimization." So, I think now the pendulum is turning again. Now, we are okay. We are at a point where the apps have good product market fit, and

24:55now we are optimizing. So, MMP is there, ASO strategy is there. We try to use all of Apple's promotional offers, promo codes, and contingent pricing is a new thing. We're looking into that. So, we we try to be as up-to-date with whatever we can do with these apps. One of the problems that we always run into is just justifying the business case for some of these things. So, when you have an app that has double or triple-digit million MAU, then most of the time the magnitude of opportunity has to be very high. So, it's so it's like the team size doesn't grow the same way how your MAU is

25:34growing, right? Like because also team size cannot grow. You cannot like because like to add one engineer or say to add 10 engineers, you also need a manager. Your team size doesn't exactly grow, and all of you know that. So, our thinking has to keep growing. As our MAU is growing, we have to think bigger and bigger, and and our prioritization has to be more and more strict because like if you have double-digit million MAU, you can't be spending time on an optimization that will impact 1 million users. You have to be spending time on something that will

26:04impact say 50 or 100, whatever, right? It's a big number. So, our framework, our VP, is very very strict about this, like estimated impact. Any kind of vendor we have to bring in or anything that we have to do, any kind of optimization, he's like, "Okay, give me the estimated impact." And then we have this huge list of ideas that come from all sorts of places, and then we are like, "Okay, let's triage." And a lot of times you'd be surprised that the

26:29biggest ideas are the simplest ones. Like, "Okay, let's do something on purchase flow or let's do something on the first-time experience." Because if you have an app that big, then sometimes the simplest things, like performance, reliability, file open speed, size of the app, those things matter start to matter a lot. Yeah. Are we able to 100% transaction completion? Because with that kind of numbers, you know, you we at one point we were seeing 5, 6% transaction

27:01failure. So, we first have to fix that. So, I think that is where it becomes different from startups. Because It's not all that different though. And that's what I was thinking the whole time we were talking is that I think too many small apps don't have that prioritization framework in place. I hear apps all the time. They're like $100,000 a month in revenue. You know, like orders of magnitude smaller than what you're working on. But then they go try and like oh, we got to have win-back

27:28campaigns. Mhm. It's like well, if you're only at $100,000 a month and you only have thousands or tens of thousands of users, like that's not low-hanging fruit. And so, from the smallest apps in the App Store to the biggest apps in the App Store, you really should be thinking about these prioritization frameworks. And you should every feature that you're working on, every tool that you're going to implement, every strategy that you're going to put in place, you should be thinking and have a hypothesis. Like what's the real impact going to be to our business? And again, if you're like oh yeah, we got to do win-backs, like

28:02that's a huge thing. Well, okay, if a thousand people are going to churn from your app over the next 6 months and you can hope to win back 5 to 10%, like how meaningful of an impact your business is that compared to like so many other opportunities that you can have much higher impact in your business. I think that's actually a great way to think about it. So, as we wrap up, we're going to link to your LinkedIn, but anything else you wanted to share? Like is your your team hiring or like anything you

28:30wanted to share from your end? I'm actually not sure. My team might be hiring. Yeah, no. I I actually don't have that information top of my head. But yeah, I mean, please feel free to reach out for any questions or discuss any app growth strategy. I'm pretty active on LinkedIn, so. Awesome. And then for those of you here in the audience, as I said earlier, you know, the Sub Club podcast, we talked to folks like Ramit every 2 weeks and we actually talked everybody from like indie apps who are scaling up, mid-market apps who are having the same struggles. And then of course, I do work

29:02for RevenueCat. In these kind of conversations, I don't talk a ton about it, but we're a subscription monetization platform that helps with all this kind of stuff, kind of that base layer of infrastructure that does push data to WinBacks, you know. And again, I'm I'm not here on stage saying "Don't focus on WinBacks if you're early." But as you grow, that's a really important thing and like our data at RevenueCat can really help push that. So yeah, thank every everybody for joining and hope you enjoyed a conversation today and check out the Sub Club podcast and RevenueCat and everything else we're

29:32working on. Thanks so much for listening. If you have a minute, please leave a review in your favorite podcast player. You can also stop by chat.subclub.com to join our private community.