How would you improve Amazon Photos?

I don't know about Amazon Photos can you please explain the product?

Interviewer: Few points: a)CAn upload pictures b)btw, it's free for pictures with no size limits and for vidoes it's upto 5GB limit c)no compressing so you get full quality d)App as well as on desktop/laptop (based on interviewees assumption) e)Interviewee is correct: can share pictures with others

Me: Thanks. as a Lead PM I tend to want to think about what our business direction is from leadership coz that helps me focus on specific solutions to help move the needle in that direction. I then need to think of who my user is so I can step in their shoes and try and think what frustrations they may have. This helps me come to some solutions to evaluate and execute on. Can we take similar approach?

Since Amazon Photos is likely new and there are possibly competitor products, I think it's fair to state Amazon would want to focus on acquiring customers.

Interviewer: Sure, you can also consider this adoption.

Me: Correct. To me adoption / Acuqisiton merely measn the same.

Interviewer: BTW, ther is no limit to how many photos you can upload.

Me (thinking): He is hinting at something but I can't put my finger on it.

Clarifying questions: I think safe to assume any georgraphy and safe to assume that the solution isn't specifically to be solved for app vs. desk/laptop. Also, do we have specific data on what customers have been wanting? (No). Ok I can't think of anything else to ask right now to narrow and define the scope but if I think of something else later on I will ask.

Let me think of some user segments for us to improve experience for. This could be a 2 sided market place in the sense that we could have vendors posting offers for framing pictures of consumers / uploaders choice, etc. But let's think of consumer for a second. a)Amateures b)Professionals like journalists c)Hobbyists d)Students of photography

I would like to choose Students of Photography. While the user pool isn't at scale here my thought is there competitive solutions in the marketplace but to my knowledge none are addressing this niche segment. (later on I did think maybe Hobbyists and Students aren't that different here).

Pain-points:

1)Enhancing pictures (they don't come out right)

2)Learn/provide guidance aspect is missing

3)I miss opportune moment pictures

I woud like to focus on #1 and #2. I do think #2 and #3 could be somewhat similar.  I thnk if we address the entire journey, taking the picture and enhacne is post having taken it, it would be unique to competitors that just focus on post journey. Yes, post journey phase we still can't and don't wnat to try to compete with the likes of Photoshop. However, addressing both #1 and #2, I think can ave major impact to customer experience and will help us bring them on-board to amazon Photos (my way of reminding the interviewer of BO indirectly).

Solutions:

Solutions here consider that we have integrated Amazon Photos with cameras itself beyond just havign a on-premise or cloud based solution for laptop/desktop.

2a) Suggest what settings to choose while taking a picture or automatically change to the right setting. It may be so that we provide an opt-in/out feature for this in setting. This auto changing settings can happen by integrating with APIs from Wheather.com and other wheather solutions and looking at meta data of other really well taken shots and metadata of currrent environment (camera lens is evaluating light source, histogram, amount of light hitting the sensor, etc.). There is significant learning and improvement involved here which may mean we would have to rely on AI's reinformcement logic.

Trade-off: Additionally, there is a challenge of how to determine best taken shots (user reviews may have to be relied upon and what if the users inaccurately rated a picture as well done + to me a picture may be well done but to another photographer it's not and if my rating of well done pic is used to make settign recommendations for him, well he may not be happy with decisions camera / amazon photo software took).

Eval:

C=5, V=5, R=5. I rate Value to be 5 just because to my user who cares about learning, this is a like a live freely available photography coach using community at large! Complexity is definitely high here since you have to integrate with many APIs and rely on refining AI Algorithm.

2b) Look at other pictures that are taken at same location as I am in to suggest me what frames to take. "Framing" is 50% of photography. This can be attained by using GPS data and storing that in metadata so you could do a metadata lookup / comparison and then on the screen you can prob click a tab that shows me other highly rated pictures in that location.

Trade-off: Downside here is that camera screens are at max 5" so too difficult to clikc on these things and while it's really still possible I doubt user will actually wnt to go through so much detail to take a snap just because there usually isn't that much time. However, on theflip side, our user is not a traveller but a student so that person is likely not going to care about spending 10 min per shot (I know I spent 20 min for 1 shot at Norte Dame cathederal in Montreal).

Eval: V=4, F=5, R=3. Complexity is high becasue you will have to aagain identify best UX and also need to implement appropriate AI functionality and relay on evaluating user ratings before using it for the feature.

This is a lot but let me quickly think through solutions for other pain point. I will be short considering time constraints.

1a) Inegrate w/ Photoshop to offer specific editing options most widely used. This would be focused not for the camera but more for the app (desktop or mobile). These may be providing help in terms of editing WB, ISO, Aparture depth, etc. (C=3, V=5, R=5). complexity is low considering we only need to figure out what features to offer and not offer and integrate. Value is high sicne without having to buy expensive licnese for Photoshop I have limited capability to enhance my pictures and make them very different than just using filters on Instagram. It helps me stnad out too!

1b)Give me inputs based on my picture (considering location and type of picture I possibly attempted to take such as a zoomed in shallow depth of field or picture in front of the sun, etc.) and metadata of my picture, suggest me what editing options others with best rated pictures have used and wwhat those pictures looked like before and after. This may require user permissions, much larger datacenter usage and this one is also potential for errors deteriorating user experience. C=5, V=5, R=5.  

As I think of prioritizing and figuring out an MVP, are there any constraints such as time, cost, resources or anything else that I shoud know about?

Alright, it's best to prioritize feature that impacts the start of user jjourney and that to me is pain point #2. Reason being, our focus is customer acquisition and that's best achieved by enhaning user experience on beginnig parts of journey. Long term goal we can address end of journey which is now that I have taken a pic how can i make it better. Out of the 2, I would MVP 2a because 2b assumes that user will actually shift through all the pictuers and decide which frame to take which I think is unreasonable to expect. A user typically wnats to take what he/she feels attracted to. While option 2a will auotmatically be able to adjust settings which my speifically chosen customer base will find more valuable in addition to any other users. Option 2b is only more applicable to my specific user segment some of the times.

I would like to evaluate the success of this in a QTR (3 mo should potentially provide enough data for fair evaluation). I may have to be concious of when to launch the feature (maybe right before summer months so that more people actually go out and take pictures and use this feature).

Success measures:
# of net new subscriptions of photos (increase suggests this feature is the dirver but can't be sure. It could be other things such as photographer reviews of this feature, marketing efforts, bad press from competing products, etc.)

% of users who use this feature (increase of this seems to be potentially an ideal indicator).

We can also measure NPS and possibly other ways.

Overall, I wanted to improve Amazon Photos for students whose focus is being attentive to details that drive a better picture than just going for most amount of pictures and speed at which they can take a picture. It's a smaller segment but focusing on niche segment can help us gain their attention easy and be hard for competitors to take them away in future. Also, we can later on get them involved in other ancellary businesses to strengthen an ecosystem. I choose option 2a since that seemed to be the most valuable to my chosen user and it would I think do the most in terms of acquiring those customers. There are trade-offs like I had mentioned and we would want to figure out solutions for those.