Ask HN: Which problems can we build an MVP for in a week?

If you're going to tackle something in the area of real estate for your MVP, I'd advise being careful not to create something that is a two sided market place.

So if the app requires people to list property, but as well as users to view the properties etc then it's something that will be hard to gauge any traction on quickly as you'll need both sides to be rapidly signing up.

Developing a product that is focused on either the seller or the buyer would improves your chances considerably for validating an MVP.

I was thinking about a 'Yelp for rental apartments', where people can rate their experience with various landlords. Hopefully it could make bad landlords provide a better service. But as you point out, it's such a huge task to reach critical mass on something like that.

I know this market really really well. There have been attempts at this, but unsuccessful.

http://consumerist.com/2006/01/26/trembickycom-the-bad-landl...

Yeah, Swapt too tried something similar, but seems like they are doing a pivot, or at least re-launching.

http://www.swapt.com/

If you know this market well, I'd be very interested in getting in touch with you, for possible input on future project I might be involved in within this space.

hah, those guys just hired people to write reviews. Their "users" had hundreds of reviews each, and almost all were freelance writers. I wonder if they told investors they were "growth hacking."

The reality is that it's pretty unlikely to get traction with a review site unless you have a very cheap source of traffic (like an existing rental site).

A review site is just a two-sided marketplace, as the first commenter pointed out.

Hey there - I'm the founder of Swapt. As a5seo pointed out, we did pay certain users for reviews (much like Yelp in its early days) and offer other incentives (e.g. donations to local charities). Our average cost per review was around $1-3 when you factored in all the incentives (note that +50% of our reviews came from unpaid reviewers). The average number of reviews per user was 8 - much less than the "hundreds" claimed above. There were a few users that wrote +100 reviews, but they all worked in real estate (e.g. agents/brokers).

The major problem with "apartment reviews" is that the average person has only lived in 3-4 places and doesn't typically want to write about places they've visited, which significantly limits the # of potential reviews per user. As a rule of them, only ~1% of users contribute content, so you really need that 1% to contribute a lot of content. This works for Yelp because that 1% can contribute hundreds or thousands of reviews (e.g. they may try a new restaurant every week). But this isn't the case for housing given the low turnover.

Brokers can write significantly more reviews but some users found their reviews to be biased (e.g. only giving positive reviews to apartment communities that they represent). The brokers were also very motivated by money, so unless you have an established audience that will help them secure new business, it's a tough sell.

Overall, I learned quite a bit from the experience and definitely think there are numerous ways that startups could bring greater transparency to the rental market. But if you're trying to do "apartment reviews," just understand that this will likely take you many years and many VCs are unwilling to be that patient.

If you want to know anything else about my experience with Swapt, just ask or hit me up on LinkedIn. Best of luck.

Yeah and unlike restaurants, people aren't potentially renting a new place 3 times a day. So I'd guess it would be a struggle to even start collecting content.

Plus, something like that has varying use depending on where you live and the rental market. In Melbourne, Australia the supply is low so there is a lot of competition for rentals. Most people wouldn't care too much about the experience of the last guy, nor do they really have the luxury of choosing, they just want a place quickly.

unfortunately, Tenants are disincentivized to complain about their landlords.

That's pretty good advice and not how I'd considered that sort of thing before. Thanks for posting.

Never really looked at it like that before. That applies to an unrelated project I'm working on that is built like that. It would make sense to focus on one side to gain traction then open up the other side to feed it once it has a user base.

You won't get real answers here. Go and talk to people in the real estate market and listen to their pain points. Forget about finding actionable data on the web. The real world has it available right now. Just pick up the phone.

This should be a lesson at all "coding bootcamps" - learn to pick up the phone and talk to people.

Not just to all coding bootcamps students, but to all coders/founders/creators. I've met very few who actually do it, spanning from amateurs to experienced developers.

It's weird right?

The rest of the world, the people who will be paying you money or helping you solve your problem, don't always operate in the same closed off world of email-only communication.

In my work experience : Phone calls or in-person meetings = $$. Email only = lots of emails.

So, in this case - I'd compile a list of realtors and/or letting agencies. Then, call them up and ask them about their pain points.

Not only will you get REAL and valuable info, but you'll be making potential connections. (call them up next week and say, "hey, remember that product I was making to help you? well - I did it, you should check it out")

A room full of programmers will likely never have the answers that you are looking for (in this case).

It is! I will take your advice and do it. Almost every time I talk to potential customers I'm surprised over the amount of valuable info I get from the chat.

Might I recommend a quick perusal of the following resources for some ideas?

http://paulgraham.com/startupideas.html https://github.com/samsquire/ideas https://www.reddit.com/r/codeprojects/

Lots of chaff, but there's a few pretty awesome ideas to be found which could probably be tackled in a week by a focused team.

Here's a simple process to follow:

1) Identify a pain-point you are personally experiencing in some X-industry

2) Search through to the deepest darkest hole of the internet to find an existing solution

3) Identify whether current players are poorly marketing themselves OR if the market is too small (only 500 people need this??)

4) If after analyzing 3), the idea is viable, go for it, else begin from 1) again.

That's a good process! My personal pain-point in this is paying way too much rent for way too little. (However, that's just the supply/demand mechanisms doing its work). But also, the landlord aren't delivering what it promised before I moved in...

It might be interesting to look into creating a service for form-letters for various real-estate complaints. I know that previously, I've had to deal with landlords who were withholding my security deposit, or refusing to make repairs, and a way to quickly generate official, legally-correct letters for requesting what I am legally entitled to could be useful.

Definitely! I'll look into this one.

Problem: I have a 30+ variations of the same photo, and I only need the "best" one. Help me find it with the help of Turk.

I bootstrapped a real estate related web app and sold it for a bit over $10M. My advice is that unless you enjoy creating stuff for technology laggards (i.e. people who adopt things only begrudgingly), avoid that industry.

Mind sending me an email or connecting on LinkedIn? I'd be very interested in getting your input in the future: per@forlagetpropell.no https://www.linkedin.com/profile/view?id=92593053

I'm curious what is your web app?

Here's an idea: aggregate available APIs (Google, Zoopla, TFL) to help people make decisions which neighbourhood(s) might be right for them. Make users answer questions (e.g. What newspapers you read? Where do you work or regularly commute to? Do you have children? etc) and narrow the options based on the answers.

Thanks for the tip! I like it. This would make a great project for using D3.js, to visualize the data in an exciting way, if people are to browse through neighbourhoods for example.

Do you know of any services that does this already btw?

Check out www.illustreets.co.uk (built in Meteor)

Problems in Real Estate listings:

Accuracy (many expired listings out there)- a good MVP would be a third party solution to feed inaccurate listing reports back to Rightmove / Zoopla etc, along with information on what part of the listing was inaccurate.

Shared Flat - finding the perfect room mate. Kinda like a dating app, only for room shares.

As someone else mentioned, Neighbourhoods aggregation of data from Gov websites on Crime, that sort of thing made into a good pluggable in solution using a postcode. .e.g. Pass a Postcode into your MVP API, spit back aggregated neighbourhood info.

gov info london - http://data.london.gov.uk/

and there's glasgow too http://open.glasgow.gov.uk

Probably some more if you searcha bout!

Awesome data sets, thanks!

Also: http://data.gov.uk/

The real problem is finding the useful needle in the data haystack, because there is an incredible amount of open data out there.

I've had exactly this on my ideas list for a long time, it's something that interests me.

Thanks! I know that SpareRoom does flat mate matching, but only for their own platform. But it works quite well, I believe.

The question is wether it would be better to have is as an independent service or on top of a platform (like SpareRooms). Not quite sure about that..

If you want to help inequality in the UK then figure out a way to get money, people and jobs out of the black hole that is London.

The biggest pain I found living in London (that can be solved in a week) is bus lines visualization. I find almost impossible to visualize the route of a bus, much less to combine two routes to get to a point. The official site gives line number information from point to point, but no visualization. I don't know if the London transports has an API with real time information, but it could be cool to have all the buses diplayed in a map. This could be expanded to see delays, or other line issues.

Curious to understand: how are the official ones not good enough or useful?

http://www.tfl.gov.uk/maps/bus

They are usable but they can be improved quite a lot, they are too general. The night map is specially bad.

There is a visualisation on the website its just super hard to find! http://www.tfl.gov.uk/bus/route/14/?direction=inbound

It would be cool to be able to click somewhere on that map and then somewhere else and have all the routes in x radius display. (I dislike the tube so this would be amazing!)

When I lived in Boston, I had this problem. When I visited London, I had this problem. Now that I live in Austin, I have this problem.

Doesn't Google Maps work well for this?

I find useful for tube conections, although it's really(really) slow and more confusing lately. I've been unable to find a useful bus route with google. But maybe it's me.

You might talk to some realtors and investors in commercial real estate. I know one guy who complains extensively about all the spreadsheets they mess with to do investment return projections. He wishes it were a web-based app he could access. He says this is an endemic problem in the industry where there are some very high-priced tools for modeling, but not much at the lower end. A SaaS might have a place taking over a lot of work in the low-mid end of the market.

Sounds like an interesting problem to tackle. I'd be interested in getting in touch with your friend. Mind dropping me a mail? My email address is per@forlagetpropell.no.

A week is a pretty short time, especially if you haven't previously integrated all the parts you need for the app (maps, geolocation, Facebook, push, etc).

But if you have, you could do a mashup that does this:

As you walk around London, have a look at the local area. If there's a flat to be shared nearby, send a notification. If you want to share a flat, make a screen that tells people about your flat.

IMHO that's more than enough work for a week.

I've been looking for rentals in London lately, here's my frustration. Perhaps this already exists and I can't find it (please tell me): a good, map-based UI for refining search for rentals.

When I'm looking for a house to rent in London, it's always lists. I never lived there, so I see area codes and addresses, and they mean nothing to me. W3? Somedisuch park?

So now, I first go to all those sites, sort by price, some obvious filters, write down the addresses in an excel sheet and pass that to a maps API. Then I have to check is there a bus close? Subway? Nightbus? &c. It's a huge chore.

I'd like big map with all the listings, and filter controls on the left. Like for the tech compare & review sites. Max price, min price. Smoking y/n. Pets y/n. Iterate and refine the searches. Hover over a listing: info. Buttons on each info box to remove individual listings, or favorite / star. Perhaps some of my own comments. Slowly iterate iterate until I get a list of 4 or 5 that look most promising. If possible, allow overlaying with custom maps, or at least have info about transit (nightbus can be a requirement for some, subway for others).

I wouldn't even mind seeding some of the data, if I had to. At least it'd give me a good overview overlapped with transit.

Does that already exist?

EDIT: Worst part about my current workflow is updates. Last time I checked was last month. Okay, now what changed? I can't know efficiently.

EDIT2: Well whaddaya know, searched again and here's what comes up https://www.padmapper.com/ . Okay, I guess that's a solved problem! :) great

If you're looking for whole flats zoopla has a really powerful map-based search where you can manually draw areas on the map that you want to search in.

If you're looking for a flatshare then you can search on a map on Spareroom, or there's http://kangaroom.net/ which I found recently and is very slick.

The biggest real estate websites (Rightmove, Zoopla) do have maps as an option in the results. I also like to only use maps, whether I like the area or not.

idea: spareroom with personal weighting.

I want to optimize for travel time to work, travel time to grocery, travel time to gym, price.

Everyone is optimizing for travel time by deciding to live in the same country where they work, for example.

Yet the concept of showing properties on the map is flawed as it assumes distances matter, while that's not true.

Maybe just forget the idea of spareroom and create a time-map. You tell it where you are, and it colors you the world by travel-time. That may be feasible in a week.

What I've learned from multiple startup weekend and service design "bootcamps", is that understanding and using the process is the most valuable thing you can get out of your time. You're learning how to cook -- whatever everyone is making is secondary and just the output. Both the service design process and an understanding of BART group dynamics would be very helpful to you.

somebody told me just yesterday it was really painful to apply to non-profits. their applicant websites are terrible and timeout often, also the international red cross has dozens of different career websites and you have to apply to all of them separately...