3 Ridiculously Slow Systems for Startup Disruption – Idea dump Monday

The world is too damn slow. Product Delivery. Buying a Home. Education. Product Delivery Why does it take 5 days for Amazon to deliver my product from their warehouse less than an hour away?  How could a startup step up … Continue reading

Leave a Reply

Your email address will not be published.


3 + = 9

Custom Search

15 Responses to 3 Ridiculously Slow Systems for Startup Disruption – Idea dump Monday

  1. Re: Buying a Home. If you pay cash, you can already close very quickly. I’ve closed on houses less than a week after looking at them.

    When financing is involved, it’s a different story. Many lawyers are involved, which slows things down. You know the dozens and dozens of forms that you have to sign at closing? (the last time I closed with financing, I signed over 150 forms) Each one of those forms represents some past lawsuit that burned *somebody*, and when that happens, an additional form is required to indemnify *somebody*.

    The only way to really streamline home buying is to either pay cash, or get rid of all the lawyers.

    A startup that eliminates lawyers? I’d support that.

  2. mike THE KING bradshaw

    The house buying thing is so slow because lawyers are involved.
    Disruption in law offices is using something other than Wordperfect5.1 to create documents!
    Also there are so many parties involved in a house sale and purchase, it is just going to take time (ever tried to get 4 people to agree on a time for something?).

  3. Re the delivery question: I just ordered like 8 items from Amazon and they were shipped in 5 shipments: One from California, one from Kentucky, one from Connecticut, one from Texas, and the last one I cannot remember off hand. Only two of those took more then 2 days to get out of the warehouse, and those two were not filled by Amazon. So some of your initial assumptions are faulty here. Robotic pickers in the Amazon warehouse could speed things up somewhat, I don’t doubt, but then you still have to ship things long distances. Self driving trucks could help out here. Then, after all that, you get to a local distribution point, are your copters better? Maybe, but this local distribution point would probably look a lot like the post office, with a lot of mail to go out, a little to each address. Now your copters would probably not like carrying heavy loads, so they would have to make many trips back and forth to the center. Not to mention that, even if you have some sort of standard top opening mailbox, every house is different. You would have some that had overhangs, hard to see tree branches, windy conditions, etc. Robotics would have to be much better then it is now for air delivery. And then you still have to deliver big/heavy packages.

    • That’s the best thing about it. There is no limit on drone size. We could have huge airship drones carrying containers between warehouses, and tiny drones carrying xbox games, books and DVDs from the warehouse to your dropbox. It’s entirely scalable and flexible. If more people are ordering smaller items, then Amazon can build more of the smaller drones, if more orders are coming in for larger items, then they can build up larger drones. It’s completely flexible.

      Let’s remove the middlemen, the delivery guys walking that last mile for the product (the most expensive mile of a delivery). From what I’ve been told, Amazon already makes use of robotic pickers in their warehouse in some areas with fixed pack sizes (DVDs).

      Ultimately, Standardization of package types/sizes is the key.

      With the drone ships, why bother with a local post office? You could have them all go straight from the warehouse to the homes, without the local. So long as it’s within their battery lifetime.

      Maybe you can print out and place a QR code in your garden or rooftop or mailbox and the drone will know where to drop it!.

      • First you can’t scale physical objects like you do computer software. You have to have enough on hand to handle the bigest demand and you can’t just dismiss them when you don’t need them. You have to store, power and maintain them in the slow times. Even if you can assemble them quickly to meet rushes, you have to have the parts on hand and the storage and infstructure for all that.

        But my main complant is it is ineffcent to make single trips for long distances. Actualy it is ineffecent to make many single trips at all, but it becomes more so as distance increases. So what you end up with is many little distro centers that your remote locations can stock. And you just rebuilt the post office, or fedex, or ups. The final mile IS the most expensive, but your plan makes ALL you miles the most expensive. Air is the fastest way to ship things, so why are trucks still used? Air is the most expensive way to go too. And copters are less efficient then airplanes. One reason that you are now seeing all these mini copters around is battery weight has gotten better with Li-Ion bats ; they cannot handle weight. Now then, I could imagine a local delevery truck like thing that cruised the streets with minimal stops while copters droped off mail and small packages from that but they would be traveling yards not miles.

        • It comes down to a detailed cost analysis.

          Is it more cheaper to build and maintain large warehouses at key locations together with that expensive last, inneficient mile, or is it cheaper to have a single location per country with all your stock and maintain small, ultralight robots to fly to all locations DIRECTLY (this is important, as anyone who has studied path finding and the travelling salesman problem can tell you).

          I say the robotics method is better. It does scale, just not in the ssame way as computer programs do. Demand is flexible, and the robotic delivery units can handle real-time demand better than storing and moving products between warehouses based on “expected” demand, as we do now.

          Just look at the massive profit increase Walmart saw when moving certain products to stores near where hurricane Katerina struck.

          In terms of scalability again, once you have enough of the units, individual breakdowns have a smaller effect on deliverability overall. Look at hard drive failures at a Google Datacenter. Efficiencies of scale come into effect and change the game. This is real world scalability in action.

          The more you consider it, the more inevitable the robotic delivery drone concept becomes.

          • Wow, I don’t understand some of your points at all here.

            “It comes down to a detailed cost analysis.” Neither of us has that, do we?

            Is your proposed solution to the traveling salesman problem to go to one site and come back to base, go to one site and come back to base, etc? Seems like that is among worst possible cases.

            Then first you say, “the robotic delivery units can handle real-time demand better than storing and moving products between warehouses based on “expected” demand, as we do now.” Then, “Just look at the massive profit increase Walmart saw when moving certain products to stores near where hurricane Katerina struck.” I those seem like contradictory statements to me.

            I don’t disagree that you can make your system some what flexible but I think you are overestimating the scalability here. Sure when you have a lot of copters the price per copter goes down to an extent and you are less impacted by individual losses but, you have all the storage and maintenance issues. You just don’t have to deal with this sort of thing when you use a postal service.

            As to the expense of having one central warehouse vs many smaller ones, you don’t have to build the many smaller ones, the postal services have done that for you. I assert that the current system is more efficient then having a fleet of aircraft that must carry items of varying weight long distances at a moments notice. Ultralight copters are ultralight because lifting weight through the air is expensive, as I pointed out before. Now there maybe merit to the idea that people would be willing to pay more for very fast delivery, I don’t have market studies as to that, but amazon does offer next day delivery for most items if you are willing to pay more for it. Would your system really be able to do that much better?

            Once more, your system solves the last mile problem by making EVERY mile like last mile. The last mile is cheaper because for the non-last miles you use bulk transport. A semi truck is not cheaper then shipping your packages by motorcycle because it gets more miles per gallon, the semi is worse by a long shot. It is cheaper because it gets many more packages per mile per gallon.

          • I proposed in the sense of the robotic drones flying direct, without taking the “long way round”, hence a significant fuel saving over the use of a land based vehicle, which needs to avoid objects in it’s path. Pythagoras’ theorem should tell you just how much of a saving this can be, more than enough to cover the cost of maintenance of the drones.

            Later, as the technology develops, multi-drop deliveries could easily be put in place. Imagine the robotic drones picking up containers, which can be programmed to drop a single item on command (like a mini, single item sized vending machine).

            There’s also the mothership concept which could be used. A large drone, with smaller drones taking delivery containers from it as required.

            The Walmart comment showed just how much profit their is in meeting demand as it happens. If Walmart made $X million extra profit from using expected demand for a region, imagine what the profit margins would be like if they could exploit real-time demand. Huge.

            See here’s the thing. You don’t just have the copters. You have the USP ON THE COPTERS. You OWN THE TECH. This means no one else can copy you. You’d be so far ahead of the game that no one can catch up. Online, people forget just what USP is. It’s protection for your business and it’s a significant barrier to entry for anyone looking to compete. These copters would be just that. A significant USP for whoever develops the best.

            Storage, is negligable, the copters could be racked up and made to come in to the repair facility as needed. Since every copter could be made out of the same basic parts, if a copter dies out we just reuse that parts.

            The postal service is excellent for sending letters, low physical volume, stackable, each “delivery human” can handle 100’s at a time. The postal service is not so great at parcels and packets of goods. It’s just not built to handle them at modern day volumes, and it’s crumbling on costs. Your considerations are based on hypothetical assumptions of the current system being better than this proposed system on cost, which have no data backing (and nor does the drone concept). Hence, hand-waving on both sides.

            The drone system would deliver in minutes-hours, depending on distance and drone availability. Just imagine for a second, you place an order, go make a coffee, and when you could back, pop, your item has arrived. That’s the vision. We’re not talking days here. Days is just too long for a system that has the capabilities of this.

            Drone V2: Floating helium filled warehouses. We’ll have them hovering over major cities like huge blimps (we can place our site logo on the side)…with mini drones delivering out, solar charging when needed. Start out with just DVD’s and games (easy, standardized packages). Key selling point – fast deliveries. Deliveries in minutes, order by smartphone any time, day or night. Drones locate via GPS, cameras for facial recognition, and the drop the pack right at your feet. In minuntes. Imagine that.

  4. I think education takes the time it takes because that’s how long it takes to learn something, and that something isn’t always the very core of what you’re attempting to learn. Would I hire a programmer who “learned” programming in a two-week course? No. Would I(and have I) hired programmers with no formal education but with 10 years on the job education? Yes. Life is learning and learning is life and both take time. While I’m certain there are places that need work here and there, your weekend long pottery course will almost invariably end with an ash tray full of holes.

    • True, for some things. With other things, I just want to learn the basic skill set to “try before I buy”. I want to see if I’d like it. For that, I still need to buy in @ a minimum of a 10 week course.

      I guess ultimately, what I want is flexibility of choice, and availability of choice. I want it my way and not the way someone else decides is best for me because that’s the only way they can do it. Computer based education gives this freedom. But can it be used to teach hard skills? Carpentry for example?.

  5. Some more detail on what you’d have to streamline to buy a home (any or all of these could probably be targeted for profit):

    1. Title search. So you can have *some* idea that the people selling it actually own it free and clear and/or all liens will be satisfied at closing.
    2. When you sign the mortgage papers, you have a 3 day right of recission. So, in the current legal framework, unless you’re buying with cash, the money can’t be released to the seller any sooner than 3 days. I think there’s also a 3 day rule around disclosures, so you can’t close until 3 days after you’ve signed the disclosures. That’s a 7 (business?) day minimum.
    3. The county clerk/registrar’s office. So that the transfer can be legally recorded. I don’t know for sure, but from what I’ve seen of these places, the systems are from the dark ages.
    4. Appraisal. Again, unless you’re paying cash, the mortgagor is going to want to know that the property offers sufficient collateral for the loan they’re writing. From what I’ve seen, there’s a lot of *paper* here; i.E. This could be streamlined.
    5. Home inspection. Same idea as the appraisal, but to protect the buyer: hire a pro so that you know there aren’t any hidden problems. Other than scheduling (which is definitely a possibility) and reporting, I don’t see much room for decreasing latency: you need a guy who knows what he’s doing to physically come out and inspect stuff (look, poke, sniff, etc). Also, many home inspections (at least in my experience) actually find some defect. The seller needs an opportunity to remedy the defect. I suppose you could have some kind of “pre inspected” listing, but I for one wouldn’t trust it — I’m going to hire my own guy to do the inspection, not someone chosen by the seller or his agent.
    6. Application process. There’s probably room for improvement here, but the approver does have to check some things that aren’t necessarily streamlined: employment, credit, tax returns, etc.
    7. It’s been a while, but last time I looked at houses, not all of the listings I visited were unoccupied. So the seller is going to need a bit of time to coordinate the move-out: rent a van, rent storage if needed, take a day off work if needed, find temporary housing if needed, change address, buy a couple of 12-packs to entice friends to help you load crap into van, cancel utilities, etc. “Instant home purchase” might make things a bit abrupt for the seller who has not yet moved out. Some of the same things are true for the buyer: you’ve got to “tear down” things at your old residence, and much of this is not instantaneous. Even when I lived in apartments, I didn’t move in the same day I signed the lease — for all of the reasons above.

    With all that said, I bought land once for cash, and closing only took about two weeks. The biggest delays were some last-minute haggling with the seller about terms, a title search that didn’t come back clean, and scheduling a time convenient to both parties. If the county offered an electronic title search that came back instantly, and scheduling didn’t take a couple of days, we could have probably closed in under a week.

    • Seems to be partly a manpower issue. We need to replce the “men” with machines. It’s the only way to drive more efficiency out of the home buying system.

      1. Can be computerized.
      2. Can be removed.
      3. Can be replaced with computerized system.
      4. Human required for appraisals. Could be computerized partly, but current condition of property needs ot be inspected.
      5. As 4.
      6. Ripe for further computerization.
      7. Move-out time, no movement possible here.

      Great list B, food for thought.

  6. Learning is not necessarily something you can speed up. University classes are spaced like that, though, because there is the assumption that you have a lot on your plate. So you have two hours of pottery … followed by 3 of statistics, 3 of organic chemistry, etc. But you can’t just learn a discipline in a weekend simply because becoming comfortable with a field is a function of experience and experimentation. Allowing certain students to say “this is my only class, give me lectures ahead of time” could speed up the formalities, though.

    Buying a house could be sped up, though.

  7. The first idea (faster delivery) was actually the basis of the infamous Kozmo. Wonder if they would be able to make it today….

  8. You had me then you lost me. You can’t rush education.