Implementing a product-centric model for digital transformation
Reed Hoffman who’s a co-founder of LinkedIn used to talk about creating digital products and he used to say “if you’re not embarrassed about the first version of your product you’re not doing something right” and I think that is fundamentally changing now because people are building complete simple products. And my favourite example is Uber. if you can’t order through the app, there is no other way of ordering a cab. If Uber was an MVP there would be a tele-calling facility available, right? So, you need to now build simple very complete products.
The second thing which you do today as part of the digital augmentation which is that you plan for AI and IoT from day one. Most of us do classic web application development and we don’t provision for these things. We don’t port upon validation, so we have data quality issues, we run into not having designed for data platform engineering on the data side, we don’t do any of those issues and the result that happens is that certainly the application becomes monolithic and big and then you are struggling to launch it.
The third thing is that you can’t really build a modern-day product leveraging Microsoft Excel, Word, as a collaboration platform. There are much better platforms in the market today, purely from a team collaboration perspective, and these collaborations are not just sharing files, or exchanging messages over slack, these are actual collaboration tools which go much beyond just your engineering collaboration capabilities. The fourth point is the growing impact of open source. Today if you’re building any web-scale product, it’s almost impossible to build it on commercial off the shelf technologies. I’m reminded of a quote by a very prominent retailer in the UK, he said “if I have to become a digital-first company, I fundamentally have to stop being a retailer. I fundamentally have to become a technology company” and as you know that’s very hard to do if you’ve been a retailer or a bank all your life, and that’s how digital is profoundly impacted.
The last point is that a lot of consulting firms, you look at Gartner Mckinsey, even Accenture, they’re talking about this whole post-digital world. I think IT likes to reinvent new terms because that’s how you increase sales year on year but the point really is that new sets of capabilities are getting added, so if you were doing AI and natural language processing especially, that’s no longer a competitive edge. Neither is IOT for that matter. The new capabilities that are getting added now are voice commerce, the ability to order through Alexa and Google, your Microsoft Cortana's, those kinds of interfaces. AR, VR is becoming more mature and obviously quantum computing in certain areas, like for example drug discovery or cybersecurity.
New capabilities in the digital world
These are the kind of capabilities that are getting added. But again, the UI UX is getting complex because of that. Now how do you interact with the bot? There’s no app anymore, or how do you create an experience through a voice interface when you’re able to order through an Alexa device?
The UI UX complexity has increased tremendously. Basically, hyper-personalization which is now possible whether it is doing loan processing, Look at traditional loan processing, how it used to happen in banks, they would do a social-economic criteria for a certain category of people etc, but now big data allows you to do that, you can now easily do it at a microlevel scale. The same thing is happening in eCommerce with re-targeting engines, they can convert traffic into basically actual clients. Now virtual reality, It's a typical thing, you call the dealer and you say “I'm looking for a test drive” and he’ll say “when are you looking at buying?” and the moment you say “maybe next month” he’s not going to get you that test drive, right? The other thing is that you want to buy an off-roader but you don’t have an off-roading capability near your house, so how are you going to test drive that car? Now Volvo is actually doing this already, where through virtual reality you can actually go and visualize how the off-roading on the XC-60 might be. It's all possible to do today.
So, people are actually taking the digital experiences a different level altogether. In IoT, industrial IoT is getting very hard to do. There's a very interesting article by Jeff Melton in the latest MIT technology review where he talks about the cultural challenges of transforming an industrial company, and that has nothing to do with IoT. It’s more cultural. OYO is a Softbank investment and is the world's largest hotelier now. They have seventeen thousand rooms. They’re ahead of everybody else, and imagine if they were to make an impact in the US markets? They haven't done that yet but they’re getting everywhere, they’re in China, Japan, they just made inroads into the UK and Europe, and obviously they will arrive in America too. This is a company which came out of nowhere in three years with a billion dollars from OYO and is now the world's largest hotelier. This is what digital does to you. Imagine if you were Starwood or a Marriot or a Mercure, what are you trying to do? You’re still trying to build a good website, you’re still trying to do effective campaign management, your loyalty programs took two years to integrate if you look Marriot and Starwood; and you actually upset more customers that you actually owned before that. But this is the impact of digital - you're still integrating your legacy and somebody has come and just taken away your file completely.
Another example of digital, the service industry is dead, it's all about programmatic SEO now. How can I take traffic coming out of Google and converting that to an actual customer? That's what the hospitality industry is all about. That's what these large web scaled platforms have done. So, the old traditional service industry providers like hotels chains where we are, that is dying a slow death, nobody wants to pay a premium on the service side anymore unless you're doing like a very classy boutique hotel, but this is where the industry is going, it’s all about programmatic SEO. It’s all digital. Here's an interesting example of booking.com and Expedia. These are an example of an online travel agent OTA which has done massive conversion of customers, and if you look at Airbnb, which just bought HotelTonight, right? What have they done? What did Airbnb do the first instance, they actually digitized trust? How many of you would have been now comfortable letting out a room in your house? It was practically alien as a concept. They’ve digitized trust. And now what’re they doing? Look at the playbook which they're following. With the HotelTonight buy, they’re getting into the online space, because they have customers, so what is the difference in their strategy compared to Expedia and Booking.com? There’s a classic difference between the two. The question is, what is the classic difference between Airbnb and any of these OTA, Booking.com or Expedia.
They have earned trust, so they actually have a user base, spot on. These guys are just paying Google and converting traffic. Airbnb has a community, which they can now sell additional services through HotelTonight. If they were going to add an airline activity there, it's just literally a feature engineering from them. All they need to do is just buy a metasearch engine, do some contracts with large carriers, Delta, United, in some of the key markets they have and that’s the end of it. But now, the interesting thing you also need to understand is that most of the money which Airbnb hosts are making Is not from those one or two rooms which they are lending out in their houses, but more from the entire properties, and if you look at this Data from Merrill Lynch, if you look at the supply of hotel’s compared to Airbnb in most of the major cities in the overall US, they actually have almost double the penetration. This is not a casual operation anymore. They're actually ahead of, more standardized conventional hotel properties. This is all an impact of digital. All they have is actually a website and an app, nothing more. Booking.com and Expedia are actually converting customers. They’re paying Google for AdWords, SEO, sophisticated data scientists are running algorithms to convert. What does an OTA do now? All an OTA does is take any traffic which Google is generating and convert it into their portal. That's what they're doing. No inventory of their own. No wonder Airbnb has a higher valuation than most of the hotel properties combined actually.
There's no digital without IoT but standalone IoT markets are struggling, go anywhere. The reason why IoT struggles very simply is because of the up-front capital that's required for doing the IoT. The entire networks, the sensors, the cost of it. Unlike RFID and some of these other technologies like barcoding, there is no RIO which is justified at the moment. So you spend more, but you’re not getting any return. That's the big challenge in IoT.
It really hasn't taken off and I think it's true for most of the IoT, unfortunately, in spite of all the hype, IoT hasn't really taken off. Simple, Loving and Complete. The MVPs are dying in the digital world. You need to have one small capability and they need to demonstrate value from day one. You can’t ship un big products anymore. The second thing, again plans for IoT from day one. You look at the electronic medical record example. What is the data? There’s a lot of patient data there. You could build population health, you could build insurance claims, you can even use that data to say where the facilities should be located. Where there are people here, telemedicine facilities should be located. You could not do that with a non-digital world. So now you can run a series of analytics and even based on certain health parameters, even predict that this is the kind of claim or issues he's going to get into. What we can also do is take the classical EHR and make it into a sort of a practice manager/clinical decision support system. Both for the patient, as well as for the doctor. One of the things which we see, unlike the classic product management we had on the legacy site, is that architectures in the digital side, they change with the change of traction of the customers, so you don't need to build for scale from day one. You need to build for scale and revisit the scale architectures in eighteen to twenty-four months. So, faster prototyping, always start with Python, dropbox is a great example of this. They started with Python and now have gone to Golang, when they have hit the scale. But if you were building desktop products you didn't have to worry about that. You’ll always build from a scale from day one. But that's fundamentally what digital has changed on the architecture side. And I encourage you all to explore some of these new technologies.
Voice commerce, it's already becoming very big, Amazon Alexa has taken the commercial lead although the accuracy of Google voice is much better. Natural language processing is becoming more and more sophisticated, when you look at bots, the NLP schematics are changing, now if you look at the problem with bots is that as long as you train them well it's good but they can’t have off the cuff conversation. So, NLP now there's a lot of research going on in that area also, again quantum computing, I've been exploring that space a bit but I think it's still a little further out and when it comes to blockchain, I think blockchain is still struggling with the usability issues or adoption issues in my opinion. There are some interesting use cases where it has done well, again when you look at the medical record example, so MIT has an open blockchain program because if you look at the US healthcare system, you pretty much have a duopoly between Cerner and Epic when it comes to health records and there is no conversation or integration between the two providers that can happen. So as long as you’re tied to a certain pair you remain with them. There are interesting companies like Patronchology which has actually built a blockchain-based electronic health record and it's all designed for collaboration between various polyclinics or various specialities, so people are doing some interesting things around the blockchain side also.