Building Luxury through voice
In preparing for battle I have always found that plans are useless, but planning is indispensable — Dwight D. Eisenhower
Disclaimer: The views, thoughts, and opinions expressed in the text belong solely to the author, and not necessarily to the author’s employer, or other group or individual.
I started AlexKill in June 2019, I was heading product sole for the whole end to end product. At that time of idea validation, Alexa was accepted in market and multiple skills were already built on top of it in different domains. It was showing some sort of product-market fit for voice assistant and solving the pain point of the targeted audience.
This post is a summarization of my experience building AlexKill, as the only team member in the product. I hope this post will help others PMs and product companies of a similar stage(i.e., those before Seed funding and product-market fit, aiming to scale) in :
1- Building value-added solutions for real problems faced by users
2- Understating the sole of your product
3- Learning important lessons that will help while you take up this challenge.
Before we go further, It is important to understand the following points for readers who are not aware of AlexKill.
Note: If you are already aware of AlexKill, You can skip this part and move directly to the solution section.
Industry background -
When the first generation of Alexa devices was launched in 2014, beyond the original consumer rush to buying the latest fad, much of the business world completely ignored them. Or, worse, some laughed at them. Five years later, no one’s laughing. And it looks like everyone wants a slice of the Alexa market. Take the state of California, for instance. It now has a skill for tourism which allows visitors to get tips on what local towns, attractions, and cities to visit within the state. You can now use Alexa skill to learn whether a flight is on schedule. In Europe, over 20 percent of doctors use Alexa for patient care. Plus, banks let users check their balance and transaction history via Alexa, as well as make purchases on the platform. Even thermostats have Alexa skills built-in.
In April of 2019, Amazon announced Alexa has become HIPAA compliant. What’s more, at the latest Consumer Electronics Show (CES) event in Vegas back in January, there were over 300 different devices that announced they’ve incorporated Alexa in their latest design.
The smart speaker industry was worth around $2.68 billion in 2018 and is forecast to be valued at $11.79 billion by 2023.
Today, voice-activated digital assistants are now a dime a dozen, with Amazon Alexa, Google Assistant, Apple Siri, and Microsoft Cortana dominating the market.
In fact, 90.1 million adult Americans say they have used a smartphone voice assistant monthly, compared with 41.5 million for smart speaker assistants. An additional 77.1 million US drivers say they use digital voice assistants in their cars at least once per month.
Product Positioning -
Between June 2019 to December 2019, Alexkill offered solutions to premium hotels/hotel chains to automate support queries and room automation through voice. The whole solution was divided into two parts.
1- Support team automation-
Before this solution, Guests of any hotel have to interact with reception or help desk for any help/query. Ex- If any guest wants to book a cab then guest has to call to support team/raise a ticket by hotel app(if any), After that support team will manually book booking into system according to availability. It needs manual efforts and needs a big pool of support team. Alexkill helped to automate the whole process. Here Guest-only has to give one single command while relaxing in the room and WORK DONE.
2- Room Automation -
In this part, We automated all the smart devices placed in any hotel room, After this guest only has to give instructions to Alexa and BOOM. Alexa automatically performs all the further tasks.
In general, flow is — User gives instruction to Alexa, Alexa backend passes the query to our system, Our backend system segregates the queries according to devices and performs the task accordingly.
Solutions
1- Building value-added solutions for real problems faced by users -
To make a successful product, the way to approach things is by first understanding the target user persona, then understanding the core problems of that persona, then understanding the available alternatives to the user. Only then you can propose a solution that addresses the main problems of the user and also differentiates your offerings from others in the market.
User’s core problem -
1- Amit wants to enjoy his vacations in one of the premium hotels without doing anything. Amit also wants a pleasant time during vacations without doing anything.
2- Amit doesn’t want to interact with anyone apart from his family. He went to the hotel’s dining hall with his family where he has to wait for long for food. Also, Amit wants to understand more about facilities like — spa, casino. Oh, and he also wants to visit all the beautiful places in the city.
My personal understanding
Within a few days of product building, I come to know what acceptance was not a big problem for this product, which was good news as a PM for me. Since guests don’t want to do any activities, they want everything on the bed without interacting with anyone.
Also, we are only targeting premium hotels(five stars) in starting and there were a long list of services that we can cater through Alexkill. Basically we were thinking to provide every service through voice where guests have to interact with the support team, It may need one-way communication or two way. Ex- One-way communication may be cab booking, Spa booking, or AC temperature changes and in two ways it may be customization in food.
There were only two problems
1- How guests will understand our product?
2- Once they are familiar with Alexkill, How the product will give them customized feelings?
Between this time period, I was fortunate enough to work on both the part of the solution(ie: Support team automation and query automation).
1- Make a voice-based customized solution for Amit through which he can do all the activities while enjoying his time in bed. He only has to give a single command to Alexa. ie: Alexa Hotel, I want to book a cab for tomorrow morning. And boom his cab has been booked in seconds.
2- Allow Amit to control all the smart devices of his room through voice assistance. These devices can be of two types, Either device is smart so it can directly catch commands from Alexa or we have to add smart connector between devices and Alexa so It can understand voice commands from Alexa, converts them into devices understandable language and manager the activities.
I am not going into technicalities of how these solutions were built because Alexkill is still in the market with new identification and it would be a disservice to them if the nitty-gritty that gives the startup a competitive advantage is announced to the public. However, I can mention that both the product was months long and needed third party hardware support for non-smart devices. Apart from that the whole product was built internally.
2- Understanding the sole of your product
To achieve product-market fit, it is very important for us to understand as early as possible “what are the core benefits that a user expects from the solution”. As a young startup with a limited budget, there are only so many things one can prioritize. Hence it becomes even more important to prioritize the right stuff only.
I personally focussed Alexkill product to just serve these 3 main purposes:
1- Help users to understand the product
2- Help users to do all activities which he usually does via phone before this product
3- Help users to do all activities which he usually does manually i.e.: switching on the TV, Increasing the temperature of AC.
As a Product Manager, my main goal was to consolidate all feature releases under at least one of these buckets and keep deprioritizing anything that doesn’t impact any of these buckets. If 2 features are impacting the same bucket, prioritise the one with a higher score of “impact multiplied by the number of users/clients who will be impacted”.
iii) Learnings in the process of managing product at an early stage tech startup
- Design overhauls are tempting but should be avoided at all costs to maintain speed in the initial years. While making a new feature, make sure that the designs are flexible so that any addition or removal doesn’t require a design architecture change. Also, while improving existing features (even those with old and crappy designs), it is better to use existing design architecture to save on development time and move to market quickly. It is important for users to test the value proposition of the feature and give feedback.
- Always ask for more explanation on why something is not possible. I made this mistake dozens of times where I assumed that something was not possible because a developer or a vendor or another team member said so. Sometimes people don’t even know if something is possible, but won’t look up for an answer either because of status quo or laziness. Sometimes people already know the answer but won’t tell you either because they didn’t understand the use case and hence didn’t understand the question properly, or because they had an incentive of hiding the question (eg commercials negotiation with one of our vendors). Often enough, people will come up with solutions if you keep prodding them why something is not possible. The disadvantage I faced of not asking for more explanation earlier was that — a) I was often restricted in my imagination of what was possible, and b) I had no answer to the stakeholders when someone in the meeting asked me why something was not possible and the only answer I had was “Because XYZ said so…”. Understanding the main reason behind everything is probably the most important quality to cultivate for a PM.
- In case of stand-offs with the engineering team, find the non-negotiable aspect of the feature and let them figure out the best possible solution. This is related to the point above, but more in line with dealing with engineering. Sometimes engineering team will upfront tell you that what you want to build is not feasible or not scalable later or not efficient to build as a feature. Most of these arise when you are trying to solve a really important business or user need, but it is heavily complicating the product from a technical perspective. Use their creativity to your advantage by explaining just 1 non-negotiable truth, and then trust them with the solution. Example — “I don’t want guests will initialize service via service name expect that everything is negotiable. Now go figure”. You will be surprised with the solutions the engineering team can come up with.
- Think beyond the apps and website while thinking about products. Think about the product as a solution to a problem and all the related cases that need to be served by the solution. When we think of building a product, we usually think of wireframes, competitor analysis, analysis of growth curves on data tools, and, most importantly, the actual development. Those are indeed the more exciting aspects of product and also product management. But I learned that every product is a solution, and the app or website is just a part of the solution. It is the process around that product that makes it a solution.
- Measure every little change in the product and note down the impact on the metrics. Keep tweaking things for maximum efficiency. Once the product is in market then you have lot of user data . Aggressively from start, we have to do proper analysis to tweak the features according to the need. Here the work of product Analyst comes into the picture.
If you would like to get in touch to talk more, please reach out to me on LinkedIn. Alternatively, you can also email your queries to ankit.malhotra2506@gmail.com. Here is my Medium if you’d like to follow my stories, wherein I share my learnings and experiences! :)
Happy Reading!!
