My mother is a real estate agent and worked in both NY (westchester county) and NC (raleigh) and I grew up helping my mother with myriad activities necessary to list and sell a house

You'd be surprised how much work is required to keep the real estate market liquid. I'd say half the work is necessary, some automatable, some not so much. However, a lot of the work doesn't even have much to do with work necessary everytime a home goes on the market, but rework that is repeated every time a listing goes on the market. A listing that was on the market last year that goes on the market again this year needs all the same work done to get it sold.

Here's just a sampling of the types of things that need to be done on every sale:

-- finding inspectors and getting inspections done

-- getting radon tests done (difficult to schedule since house needs to be vacated if I recall correctly

-- showings (how to show a house, how to talk about it. how to sell a lifestyle, especially in a home with no furnishings)

-- fixing things, which is often dependent on contract negotiation. Sometimes things that need improvement will be deducted from the price, other times fixing them will be a requirement to close the deal.

-- home staging. for homes with no furnishings, companies exist to stage the home, especially for open-houses.

-- disclosures (near an airport? did someone die in the house? carcinogenic chemicals on site before construction?)

-- proper photographing, which is hard for most people because photographing homes that will generate interest isn't as trivial as you may think. One of the biggest issues with photographing was lighting both inside and out. Most agents don't carry lighting necessary to photograph homes. Some homes have built in lighting, some have lamps. Those that use lamps may not have any because furniture has been removed. Most agents aren't technically competent enough to handle a camera more complex than a simple point and shoot. Outside lighting for homes requires planning to make sure the weather was nice and that at least the front of the house was lit, which itself depended on which direction the house faced (north, south, east, west). Getting the photos from the camera to MLS and creating brochures with them is complex for many people.

-- MLS management

-- Searching of the myriad criteria that people evaluate a house on (schools, commute, etc.) Everyone has different lifestyles and needs. The criteria for one client will differ dramatically from the criteria for another client.

-- centralized showing

-- lock box management so other listing agents can show a house.

-- scheduling

-- Knowing whether a house is a good deal or not. Your agent is essentially your negotiator. A good agent will not only find you a home that fits your lifestyle, but they will help you negotiate with the seller/buyer on your behalf. This is probably where the fees are most valuable.

-- Negotiating with a bank. You have no idea how many deals these days get blocked by a bank, especially if the home is underwater.

-- floorplans (bonus points if any startup makes it easy to mash up floorplans to room photos, the same way that people mash up photos with latlng data and maps).

-- ability to simulate your current belongings in the new home? Do you have too much stuff? Not enough stuff? What fits? What doesn't?

-- Architectural vocabulary enlightenment. What's the difference between a victorian and an edwardian? What are dorian columns?

-- Energy efficiency of a house? What are the bills and utilities the previous tenant paid?

-- Hooking up utilities in a new place. Can it be streamlined?

-- etc. etc. etc.

This is all just the tip of the iceberg. I don't think there is a one size fits all solution that can automate all of the above, but there is definitely room for a few dozen startups and that you could essentially create an entire VC fund with real estate as your thesis. TBH, I think that is the only way "real estate" will ever be solved. There are way too many moving parts to have everything solved by one company. APIs are necessary for solving the problems above.

Finally, NYC and SF are very very very weird markets by real estate standards. Most markets are fairly balanced, but both NYC and SF are heavily skewed in favored of sellers. This makes them very unusual and I can't imagine a solution that works for those two markets working in other markets.

Its horrible finding an apartment in NYC. Especially if you're a working professional. Most people don't have cars, and apartments aren't helpfully near each other, so you're spending a lot of time hopping on trains and walking streets. If its the weekend, you're spending a whole day just to see a few places if you need to wait for trains. As you can imagine, any place where giving the super a six pack gets you a set of master keys, people aren't very reliable. Half the time you show up and the apartment is gone, the listing is wrong, or the place is a heap. If you work during the week, you're either running around at night looking for apartments, or by the time you get to the weekend, all your research is out of date. This is why brokers exist in NYC.

Funny, that's exactly what I thought 3 years ago...

I think the root of the problem is there are too many available apartments but most of them are a waste of your time to view. A great broker previews every listing and only shows the best deals.

Similarly for the landlord, most renters who come see the apartment will not take it, so having brokers do the showings saves the landlord a lot of time and effort.

I suspect a major factor is that the fee, when paid by the tenant, acts as a filter to shut out less desirable (i.e. poor) tenants and those who might not stay long. This is possible because demand vastly outstrips supply.

At least it's like that here in Germany - the broker fees for good apartments are almost always paid by tenants. Any apartment where this is not the case has probably major downsides. In other cities, where the supply/demand situation is different, broker fees are paid by the owners and the broker's added value is marketing (i.e. finding good tenants at all)