You are a product manager at a start-up whose technology can detect emotions. Evaluate potential applications.

Let's start with some clarifications: 

Is the technology video, text, or sound based as in does it leverage video, text, or sound to detect emotions?

Interviewer: We are talking about video and sound technology.

Does the technology require any time of maintenance schedule or cost for up-keep that should be considered? 

Interviewer: Nope, no maintenance. Everything is wired and hooked up. 

How much does it cost us to create a shippable instance of this video hardware + software product? 

I'm going to assume 5,000 for the base module and 200 per camera

What types of emotions is this technology capable of detecting? 

I'm assuming all different kinds of emotions. 

What size are the cameras?

Interviewer: Um it's about the size of my coffee mug

Is the technology on the edge, or is it processed on our servers in the cloud

Interviewer: We process the data and video/sound feeds in our servers, and store in a accessible location for customers.

To sum up, the assumptions and clarifying points are: 

1. It's a video based technology that can view and listen to a user and determine that users emotion
2. The technology works indefinitely without a need for maintenance
3. The cost for the main basestation is $5,000 to develop with $200 for each camera and microphone.
4. The technology can detect all kinds of emotions. 
5. The cameras are about the size of a coffee mug.
6. Data is processed on our servers, and customers are allowed to access.

Using a mindmap, I'm going to brainstorm some potential solutions (takes 1-2 minutes):

We can evaluate the potential applications in terms of customer segments with relations to 

At a high level, I don't think that the consumer marketplace would be a direct customer of this technology, but could easily be secondary consumers of this technology as businesses host the technologies. 

So then, let's dive into commercial applications of this technology. Does that work for you? 

Interviewer: Yup, sounds good

Within the commercial application space, there are multiple types of businesses, but I'm going to particularly focus on the ones that have a stake in the customer service game as emotions play a big role on how a customer perceives a business or product, and how businesses respond to customers. 

Let's focus on entertainment, hospitality, and retail. 

For Entertainment: 

Gauging consumer emotions to musical products and releases

Understanding the emotions as it relates to cinematic products and releases

Reading the room for a live stage act or performance

For Hospitality

Gauging the customer emotions as they receive service from staff members

Predicting when to take corrective action and how to ensure satisfaction

Understanding when guests are exhibiting frustrations in public spaces

For Retail: 

Understanding when customers make certain buying decisions

Measuring emotional response to a design or product category

Determining customer service response to potential escalations

Bonus: Discuss how to price and launch

Interviewer: How would you price and launch this product then? 

To launch this product, there are many pricing models that we could attempt to sell it to market

The first is pay once, and own forever (office 2019 model)

The second is pay once, but then an all-you-can-eat montly fee

The third is a subscription based model based on processed data

I'm going to pick the third pricing model as that will allow us to grow as our customers grow, and ensure a steady stream of income based on volume. 

Assuming the video feed is HD and the sound quality is high, we are expecting some potential cloud costs as well which we should consider,so let's jump in and determine that baseline cost: 

Let's assume for now that a customer pruchases a module and 5 cameras from us, and those cameras are running 15 hours a day. 

At 2gbs an hour, we have 30gbs * 5 = 150 gbs of video per day, or 4.5 tbs of video and sound data per month

Knowing storage costs on the cloud, let's say that we incure a cost of about $100 per month. We may have some nodes that are scaled to process the image and sound, so let's assume for now that that costs about $200 a month for a medium size node with about 8 CPUs and 32gbs of RAM. 

Therefore, if we are looking for a 50% margin to cover labor and other operational costs, and make a profit, we're looking at charging an average customer about $600 a month for the processing service. 

I'm going to assume a linear scale here, so if the customer only has one camera, then we would charge them $120 a month per camera. 

So year 1 for a single business location with 5 cameras could bring in $6000 for hardware/software, and $7,200 in subscriptions for a total of $13,200 for year 1 and $7,200 for year 2 assuming no additional cameras. 

At 1,000 business locations, we would bring in 13.2 million in year 1, and $7.2 million in revenue every year at 5 cameras per location and a single base station per location.

If the goal of the startup to increase user adoption of the technology, then a strategy may be to give out free hardware up to 5 cameras. What this means is then for year 1: 

The company loses out on it's cost to manufacture which would be: $6 million dollars for the 1,000 locations, but would bring in $7.2 million dollars in subscriptions netting $1.2 million in revenue for the first year, and $7.2 million in revenue thereafter.