9 ms·

I have a popular local restaurant as one of my clients (and have worked off-and-on for others), so here's the scoop: the more technically-savvy restaurants get

I have a popular local restaurant as one of my clients (and have worked off-and-on for others), so here's the scoop: the more technically-savvy restaurants get their websites from the same people that provide their point-of-sale systems, and those people are part of the "industry software" problem, such that they're constantly running about 5 years behind everyone else. This applies at least to Restaurant Manager ("RMPOS") and Digital Dining, both of which make life much, much harder for restaurant owners. Same deal for grocery stores, too. Have you ever wondered why you couldn't just load your local grocery store's website up and see what their current price for bread is, or if they're almost out of eggs? All business reasons aside, it's largely because their databases are (probably) running some hinky form of Pervasive SQL or Btrieve, which stores their information in a hard-to-decipher, proprietary format, which among other wonders "deletes" records by replacing them with nulls in the monstrously large flat file that represents their product database. (The registers communicate with the server over Samba with the help of a shared file...) For this, restaurants and grocery stores pay tons of cash. Wanna make a bunch of money? Go fix this. Restaurant owners might even name their children after you.

wallflower 17y ago
From my brief research into this area, it seems what restaurants want to do is increase profits. Information like who their best customers are and their preferences (like Casino VIP databases). Data on what is selling. The daily or even real-time profit/loss. Is this what the industry giant behind OpenTable does? Is OpenTable too big to fail (e.g. restaurants will continue to pay them annual maintenance because it is a legacy system)? Can OpenTable's database be cleanroom reverse engineered so that it can be migrated to a new competitor platform?
- thaumaturgy 17y ago
  There are a few different questions here; I haven't worked with an OpenTable system yet, so I can only guess at some of it. But: > ...it seems what restaurants want to do is increase profits. Yes, but -- based on my experience and dealing with restaurant owners in my area, anyway -- they're not interested in things like metrics or VIP programs. VIP programs don't convert as well as you might expect -- pretty poorly, in fact, given the effort that it takes to manage them -- and restaurants are typically run by chefs or foodies, not by numbers geeks. Restaurants are really hurting right now, so anything that doesn't immediately get butts in their seats is going to be a hard sell. Daily P&Ls are already built into the reporting functions of the POSs I've seen so far, and they have some functions for reporting on popular dishes, although that's usually more clunky. But, the line cooks are usually the ones that tell the boss what was popular that night anyway. > Is OpenTable too big to fail (e.g. restaurants will continue to pay them annual maintenance because it is a legacy system)? Yes and no. Once a restaurant has committed to a particular POS, it is very, very, very hard to get them away from it. First, it cost them -- at a safe ballpark estimate -- $10,000 to install it in the first place, so they're going to milk it for as long as they can before paying for anything else, even if it costs them more money that way in the long run. Second, change is really scary for these folks, because if they're trying a new POS, and it crashes on its first evening, they're going to have a lot of pissed off customers and that translates directly to empty seats. It doesn't help that every single system that they've tried -- or that their buddies in the industry have tried -- have been buggy, problematic, and prone to crankiness. It doesn't matter if you know that you're selling the best system in the world at the cheapest price; the owners are still going to be really gun-shy about converting. > Can OpenTable's database be cleanroom reverse engineered so that it can be migrated to a new competitor platform? In theory, anything's possible, but I would feel pretty safe guessing that OpenTable isn't going to make that easy. They're probably using something pretty funky. On the plus side, they're probably not encrypting it, they're just relying on security-through-obscurity. So, you should be able to figure out the basics, like item description, pretty easily. Figuring out sales or quantities might be a lot harder, not to mention figuring out links between items (like, add-ons for a particular dish). To be honest, you might be better off committing staff to manually re-entering the restaurant's menu in your POS, rather than trying to convert OpenTable's (or anyone else's), and then just roll that into your price. This is not as horrible as it sounds; most restaurants have only a couple of pages of items, with a variety of add-ons. I've done this by myself in under a day, so it's not too bad.
  - wallflower 17y ago
    Thank you for your responses. I have a couple friends who have worked at high-end restaurants and they admit to being addicted to the mix of chaos and orchestration and frenetic precision. > Daily P&Ls are already built into the reporting functions of the POSs I've seen so far I finally found the NYTimes article that made me wonder if there is better software to track costs. "“A few years ago, at Cafe Boulud, we couldn’t figure out why margins were so bad month after month,” says Lili Lynton, one of Mr. Boulud’s two partners. “Even the chef was baffled. We looked at everything and we finally realized — it was this reduction sauce, a really expensive reduction sauce, with truffles and mushrooms, which was in a bunch of dishes. Who would have thought? We figured it was the fish or the chicken or the meat. It was like a game of Clue, and the culprit was the gravy.” [1] If Daniel Boulud with his team of accountants has trouble uncovering hidden, multiplying costs - presuming they have best-in-class software, is the software good enough? Answer: Maybe Avero has found that niche. Or that niche is still open (despite Avero's elite client base - Danny Meyer) "Avero’s software gets you the 411 by downloading POS reports and morphing that raw data into valuable information you can use to improve and act on every aspect of the operation of a restaurant from menu planning—like how many lobster rolls did you sell for the last four years on Thursday nights when it was over 90 degrees out, what did Moms order most on Mother’s Day, or how many vegans are eating out on Friday nights in winter—to server performance and pinpointing the server who’s, say, strongest in cocktail sales but weakest in wine sales... 10,000 food and beverage professionals in 1100+ restaurants—guys like Danny Meyer and Steve Hanson—use his software. And thanks to Avero, his clients have realized an incremental $100 million in profits." [2] [1] http://www.nytimes.com/2009/05/17/business/17boulud.html?pagewanted=2&ref=dining http://www.nytimes.com/2009/05/17/business/17boulud.html?pag... [2] http://thestrongbuzz.com/potluck/details.php?item_id=15 http://thestrongbuzz.com/potluck/details.php?item_id=15
    - thaumaturgy 17y ago
      Oh! I misunderstood you. Yeah, I haven't had the pleasure of working with software that actually tracks individual costs on any kind of incremental schedule (like cost per dish per day). I don't really know how you would solve that particular problem in software; it's more of a process problem. Restaurants differentiate from fast food in part by not serving precisely the same thing every time, so that makes it hard to track costs per dish very precisely, not to mention things like shrink. It's still possible to get a rough idea though, which is better than they're doing now. Avero looks pretty sexy, I'll check into it a bit more. Thanks for the link.
stevenwei 17y ago
I know of at least one major restaurant point-of-sale system whose API requires you to: 1) make requests by dropping files on the filesystem in a particular directory (that they poll at an interval) and 2) wait for it to deliver the response, which it does by writing to a new file in another directory. You're spot on when you say they're running about 5 years behind everyone else (if not more).

wallflower 17y ago

From my brief research into this area, it seems what restaurants want to do is increase profits. Information like who their best customers are and their preferences (like Casino VIP databases). Data on what is selling. The daily or even real-time profit/loss. Is this what the industry giant behind OpenTable does? Is OpenTable too big to fail (e.g. restaurants will continue to pay them annual maintenance because it is a legacy system)? Can OpenTable's database be cleanroom reverse engineered so that it can be migrated to a new competitor platform?

thaumaturgy 17y ago
There are a few different questions here; I haven't worked with an OpenTable system yet, so I can only guess at some of it. But: > ...it seems what restaurants want to do is increase profits. Yes, but -- based on my experience and dealing with restaurant owners in my area, anyway -- they're not interested in things like metrics or VIP programs. VIP programs don't convert as well as you might expect -- pretty poorly, in fact, given the effort that it takes to manage them -- and restaurants are typically run by chefs or foodies, not by numbers geeks. Restaurants are really hurting right now, so anything that doesn't immediately get butts in their seats is going to be a hard sell. Daily P&Ls are already built into the reporting functions of the POSs I've seen so far, and they have some functions for reporting on popular dishes, although that's usually more clunky. But, the line cooks are usually the ones that tell the boss what was popular that night anyway. > Is OpenTable too big to fail (e.g. restaurants will continue to pay them annual maintenance because it is a legacy system)? Yes and no. Once a restaurant has committed to a particular POS, it is very, very, very hard to get them away from it. First, it cost them -- at a safe ballpark estimate -- $10,000 to install it in the first place, so they're going to milk it for as long as they can before paying for anything else, even if it costs them more money that way in the long run. Second, change is really scary for these folks, because if they're trying a new POS, and it crashes on its first evening, they're going to have a lot of pissed off customers and that translates directly to empty seats. It doesn't help that every single system that they've tried -- or that their buddies in the industry have tried -- have been buggy, problematic, and prone to crankiness. It doesn't matter if you know that you're selling the best system in the world at the cheapest price; the owners are still going to be really gun-shy about converting. > Can OpenTable's database be cleanroom reverse engineered so that it can be migrated to a new competitor platform? In theory, anything's possible, but I would feel pretty safe guessing that OpenTable isn't going to make that easy. They're probably using something pretty funky. On the plus side, they're probably not encrypting it, they're just relying on security-through-obscurity. So, you should be able to figure out the basics, like item description, pretty easily. Figuring out sales or quantities might be a lot harder, not to mention figuring out links between items (like, add-ons for a particular dish). To be honest, you might be better off committing staff to manually re-entering the restaurant's menu in your POS, rather than trying to convert OpenTable's (or anyone else's), and then just roll that into your price. This is not as horrible as it sounds; most restaurants have only a couple of pages of items, with a variety of add-ons. I've done this by myself in under a day, so it's not too bad.
- wallflower 17y ago
  Thank you for your responses. I have a couple friends who have worked at high-end restaurants and they admit to being addicted to the mix of chaos and orchestration and frenetic precision. > Daily P&Ls are already built into the reporting functions of the POSs I've seen so far I finally found the NYTimes article that made me wonder if there is better software to track costs. "“A few years ago, at Cafe Boulud, we couldn’t figure out why margins were so bad month after month,” says Lili Lynton, one of Mr. Boulud’s two partners. “Even the chef was baffled. We looked at everything and we finally realized — it was this reduction sauce, a really expensive reduction sauce, with truffles and mushrooms, which was in a bunch of dishes. Who would have thought? We figured it was the fish or the chicken or the meat. It was like a game of Clue, and the culprit was the gravy.” [1] If Daniel Boulud with his team of accountants has trouble uncovering hidden, multiplying costs - presuming they have best-in-class software, is the software good enough? Answer: Maybe Avero has found that niche. Or that niche is still open (despite Avero's elite client base - Danny Meyer) "Avero’s software gets you the 411 by downloading POS reports and morphing that raw data into valuable information you can use to improve and act on every aspect of the operation of a restaurant from menu planning—like how many lobster rolls did you sell for the last four years on Thursday nights when it was over 90 degrees out, what did Moms order most on Mother’s Day, or how many vegans are eating out on Friday nights in winter—to server performance and pinpointing the server who’s, say, strongest in cocktail sales but weakest in wine sales... 10,000 food and beverage professionals in 1100+ restaurants—guys like Danny Meyer and Steve Hanson—use his software. And thanks to Avero, his clients have realized an incremental $100 million in profits." [2] [1] http://www.nytimes.com/2009/05/17/business/17boulud.html?pagewanted=2&ref=dining http://www.nytimes.com/2009/05/17/business/17boulud.html?pag... [2] http://thestrongbuzz.com/potluck/details.php?item_id=15 http://thestrongbuzz.com/potluck/details.php?item_id=15
  - thaumaturgy 17y ago
    Oh! I misunderstood you. Yeah, I haven't had the pleasure of working with software that actually tracks individual costs on any kind of incremental schedule (like cost per dish per day). I don't really know how you would solve that particular problem in software; it's more of a process problem. Restaurants differentiate from fast food in part by not serving precisely the same thing every time, so that makes it hard to track costs per dish very precisely, not to mention things like shrink. It's still possible to get a rough idea though, which is better than they're doing now. Avero looks pretty sexy, I'll check into it a bit more. Thanks for the link.

stevenwei 17y ago

I know of at least one major restaurant point-of-sale system whose API requires you to: 1) make requests by dropping files on the filesystem in a particular directory (that they poll at an interval) and 2) wait for it to deliver the response, which it does by writing to a new file in another directory. You're spot on when you say they're running about 5 years behind everyone else (if not more).