Hacker News new | more | comments | ask | show | jobs | submit login
Launch HN: Dinesafe (YC S18) – Crowdsourcing Food Poisoning Reports
76 points by paddoq 7 months ago | hide | past | web | favorite | 66 comments
My name is Patrick Quade, and I’m the founder of Dinesafe (https://dinesafe.org) and Iwaspoisoned.com (https://iwaspoisoned.com). We crowdsource food poisoning reports, and help detect and prevent outbreaks.

I launched Iwaspoisoned.com after experiencing a brutal bout of food poisoning from a deli in my hometown of Tribeca, NY. Out of concern for other consumers I called the deli to try and explain what happened and they hung up the phone. That inspired me to create a crowdsourcing platform to allow people to report. The idea was that if it was easy to discover if others were also sick after eating there, that would be useful information not just for other consumers but also for the deli owner.

Foodborne illness sickens 48 million consumers, and kills 3,000 every year in the US, according to US Center for Disease Control estimates. The financial burden is also significant. The total national cost of foodborne illness is estimated at $55 to $92 billion per year. and the impact on companies can also be significant, Chipotle lost $10 billion in market cap from its peak to its lowest point after it’s series of food safety missteps, and the founder and CEO was forced to step down.

We built a mobile responsive website with a simple form that allows consumers to report when they believe they have food poisoning. We moderate every report, with both a back end review, and front end/human review of each submission with the goal of eliminating malicious and inauthentic reports. We then geo analyze the data in real time, allowing us to pinpoint clusters of reporting associated with a single store, looking for multiple independent reports associated with the same store or location in the same time frame. We also store and analyze historial data. We developed an index and benchmarks that enable us to identify brands with negative and positive trends.

With our data, we predicted that Chipotle would have food safety issues ahead of their series of outbreaks, and we detected several of their outbreaks, and many others in real time.

Our business sells data intelligence to industry, focusing on enterprise clients in the restaurant industry. We offer daily alerts, intraday (outbreak) alerts, benchmarking, and secure messaging, plus an API service. We have also had interest in custom email branding/messaging, and "Dinesafe Certification" for those brands that rank consistently well in our analytics.

We now provide daily alerts to over 25,000 consumers, sending over a quarter million emails a month. We partner with and provide daily surveillance services to public health agencies in 6 countries. Within the US we partner with over 350 public health agencies, covering 90% of the US by population. We also provide services to the food Industry, primarily restaurants, but also producers, the grocery and convenience sectors and more. We have over a third of the top 50 restaurant chains in the US on our platform.

We are delighted to be on Hacker News, and look forward to hearing insights from the community. Thanks for your time.

Just because someone gets ill after eating at a restaurant doesn't mean that the restaurant food was the cause. There may be several causes including food the sickee prepared for himself/herself earlier in the day.

What confidence indicators are you presenting to the public so that a report can be flagged as possibly unreliable?

Agreed, this is why we do not pronounce a single report as a truth.

We do not present confidence indicators. We do explain the details of foodborne illness on our site, but we do want to do better on this, we will continue to refine messaging to ensure our subscribers, or other people who find us, understand the strengths, and considerations in looking at our data.

I agree; I had a friend who would regularly drink cases of beer on the weekends and though he said he never got hangovers, oddly he often reported experiencing food poisoning and would attribute his malaise to wherever he ate last.

In one of P.G. Wodehouse's stories, a man explains his hangover as charcoal poisoning, from the powdered charcoal used to color inferior caviar. I had never heard of such an explanation outside his fiction before.

I guess I have two sets of concerns.

The first is regarding DineSafe.org: I don't understand how the model works. I see that people can "report incidents", and I see that businesses can subscribe to get analytics on reports. What I don't see is the link: what incentive do people have to report incidents? What makes the particular types of people who report perceived foodborne illness to your website a trustworthy cohort to sample from?

Regarding "Iwaspoisoned.com" (which: I think get a better name?): it looks like what you're doing is scraping public food inspection reports (Everyblock used to do that too), and then allow people to submit their own reports that run alongside them. If that's what you're doing, it's deeply misleading: food inspections are (1) an authoritative source of information and (2) not usually reports of actual foodborne illness! The net effect is to blur the lines in both directions: to create the belief that restaurants in your area are routinely shut down due to foodborne illness problems, and to present the alimentary concerns of randos as equivalently authoritative.

For what it's worth: my concern wouldn't be that reports would be "malicious" or "inauthentic" (I'm sure there are reports like that, but I also agree that you'd probably be able to keep up with them). Rather: my assumption is that you'll have an even worse version of the Yelp problem, where people write simultaneously authentic and bad, poorly informed reports. Short of "our party of 4 ate at this restaurant and all 4 of us came down with severe GI symptoms" --- which is rare --- I don't understand how you could take any report at face value. Meanwhile: your "iwaspoisoned.com" site runs firsthand reports where people simply take photographs of their food and say in effect "this looks poisonous".

1. What incentivizes people to report ? Social responsibility/concerns for others is the most common reason given when we survey people. For many people a proper case of food poisoning is frightening and extremely motivating.

2. On trust - crowdsourcing is imperfect. We have a track record of detecting outbreaks. Trust is case by case, for instance if you as a consumer choose to distrust our users, you are welcome to. What we tell restaurants is that our data is imperfect but we strongly recommend against ignoring it.

2. On scraping - For food poisoning reports we do not scrape public inspection reports. We do pull in external data for related (but not food poisoning) information such as temporary restaurant closures due to food safety infractions (vermin etc), or critical violations and fail grades. We also allow consumers to report high risk food safety observations, (raw chicken burgers etc). However these are just additional data points that our subscribers and consumers want to see, they are not part of what we use for detection. Though they can be correlated. No intention to blur the lines. It is actually on our to do list to come up with an elegant way of allowing consumers to separate if they want, and we definitely want a checkbox on our map, and also in our app. We also have in our queue a way for consumers to customize their email alerts to choose which data they get.


Each time I went to "iwaspoisoned.com", I was greeted by a story and usually a photo of someone claiming they'd been served an undercooked burger. Here's what I just got when I loaded the site again:


That looks as gray and dreadful as any McDonalds burger. The impression I get from "iwaspoisoned.com" is that the median user of the site is, shall we say, of the "steak well done" persuasion. That is fine, those people are fine (those steaks are not fine though), but they are hardly a credible source of information about foodborne illness, and I think it's probably irresponsible to present them directly alongside authoritative (and often unrelated!) reports from inspections agencies.

What incentivizes people to report every possible detail of their lives on Twitter and Facebook? You know how I confirm if Instagram is indeed down? You go on Twitter, and everyone breaks out the "I go on Twitter to check if Instagram is down" meme. DSL Reports, and various others have come before this idea. People love yapping about things. ESPECIALLY when it's something negative to say.

The real question is if anyone really cares enough to support a niche site like this. The biggest problem is that food poisoning isn't something that happens daily for anyone.

Sorry, I wasn't clear: if you look at the DineSafe.org site, there's no feedback loop for reporters. There's an enterprise analytics application businesses can buy, and what looks like a blind report form for consumers, but no link between the two.

By contrast: if I report a hotel on the Bedbug Registry, I am participating in a resource that is open to everyone, not just people who pay for the enterprise app.

Iwaspoisoned.com is the site i think you meant to look at. It shows the reports etc, this is the consumer facing site.

Are "reports" on IWASPOISONED.COM part of the data set for DINESAFE.ORG?

Yes. Dinesafe.org sells data and data intelligence from iwaspoisoned.com

Dinesafe sounds way more serious. I'd be happier to report there (but with some of the characteristics of iwaspoisoned).

DSLReports is dying if you haven't noticed, with Karl Bode leaving, there are no more constant news updates. Really sad to watch this corner of the internet die.

I had the pleasure to attend an epidemiology doc's lecture about how difficult sourcing food poisoning actually is. Part of the problem according her-- most cases require some amount of incubation time (2-3 days and several meals). There are some exceptions, but most of the time it's not the last meal eaten, even if the symptoms began within hours. Could I suggest you encourage users to report their last few meals prior to illness? You could also try to codify user import-- things like symptoms and ingredients. "Temperature > 101.5*? Nightsweats? Abdominal pain? Diarrhea?"

There are three main types of contaminants that can occur: physical, chemical, and biological. Epidemiologists tend to focus on the last one. We are not biased toward any of them, we welcome reports from anything that can harm consumers. With chemical contamination the onset can be within minutes, and even with the bacteria Staphylococcus aureus (commonly just called Staph) onset can be as soon as 1 hour.

Further, in the spirit of our partnership with public health - attribution - while beneficial - is not the only benefit from our site. Simply knowing that you have a cluster of people in your county exhibiting norovirus like symptoms is important, especially when you have their contact information.

It gives public health the opportunity to follow up and conduct surveys including a food history, to try and determine attribution, but it also allows education, and prevention. Many people don't realize that if you have norovirus, you can be contagious for up to three days after recovery. So if you are a food worker, or work in a hospital, or daycare this point is critical to communicate to prevent further spreading, especially given how virulent norovirus is.

Besides all that - we have detected outbreaks of cases with longer incubation periods such as E. coli, so while some amount of misattribution is real it doesn't preclude us from achieving positive outcomes.


On the report box, each of the checkboxes on the "Send Report To" should have a mouseover that explains exactly what each option is.

I also don't understand the star rating. I don't think I have ever had "5 star food poisoning," but it sure does sound nice!

Great point about the hover, and information popups for mobile, thanks for the reminder. We had those in the past, they were not well used, but I still think it's great to have them, especially for the 'Send Report To' - that's a newer feature. Star ratings is something new we are trying out, some people want to rate, and as wonderfully suprising as some people are, some want to rate well. I think its' an offset to them 'feeling bad' about reporting a place, especially if it's a place they frequent, and usually like. Not sure.

> I launched Iwaspoisoned.com after experiencing a brutal bout of food poisoning from a deli in my hometown of Tribeca, NY. Out of concern for other consumers I called the deli to try and explain what happened and they hung up the phone.

Wow, sorry this happened to you. When I was at university, they served Thai food with nuts and didn't have a "Contains nuts" sign even though it's protocol to have one. I ended up puffing up and the manager took absolutely no responsibility and said "tough luck", I don't think he even cared whether or not I was sick. I had to run back to my dorm room and ruffle through my roommate's medicine to find some Benadryl -- overall one of the worst experiences I've ever had.

Crowdsourcing is very tough because of quality control, but this seems possible and I really hope this works out for you guys. Best of luck!

As someone with celiac, I have given up trying to eat at any restaurants other than places that specialize in serving customers with allergy/intolerance/gluten issues. I have had incidents and many have stories of negotiating a safe gluten free burrito or whatever, and then they accidentally served a normal one. Like you say, when they screw up we have no recourse, but unlike an allergy, the effects may not be apparent for hours, but lasts for days. The restaurant and the individuals responsible are generally unaware of the severity and insulated from the effects on the victim. I could be out of work for a week or more due to one slip up.

Now that we mention special needs, there are a couple websites that collect celiac-oriented reviews for restaurants, such as https://www.findmeglutenfree.com/search

Sorry to hear it, yes there is a lot that can go wrong. Thanks for the support : )

I'm skeptical that this will be effective.

There are obvious cases of food being unsafe for consumption (ie. A food truck got 150+ people sick at a fair https://www.cbc.ca/news/canada/toronto/contaminated-cronut-b...). In these cases, your website is kinda useless; everybody knows if a restaurant serves bad food, it will make the news.

But one time I got sick after eating out at a restaurant. My colleagues eat there all the time and have never gotten sick. But one day I did, much to the surprise of my coworkers who joined me for lunch and were completely fine. Was it the food? Did I wash my hands? Maybe I touched a door handle and put my hand in my mouth after. Maybe it was something at my desk when I got back to work? Maybe it was something in my breakfast?

I think you're just giving a soapbox to people who have no ability to determine whether or not their illness was food poisoning. And better yet, what's the purpose of just putting these claims out there for everyone to see? It makes much more sense for me to spend my time with the local health regulator, explaining my concerns to them, rather than complaining openly. Perhaps your service would be better if it was billed as a way to contact food regulators rather than a yelp-but-food-saftey.

Earlier this week I was WRECKED by something. I was genuinely praying for a higher power to take me from this planet. I called my hotel to see if anyone had reported an illness from breakfast... they were surprised by my call. It got me thinking “something like this has got to exist” ... what a coincidence! Glad to give this a trial run based on my recent experience and hope it grows to become more popular.

The bar is so incredibly low for food safety. Restaurants, especially really popular ones, get away with a lot of really bad behavior. At some point data like this will be as paramount as their yelp score.

FYI: the name of this project collides with the City of Toronto's own food safety program:


> DineSafe is Toronto Public Health's food safety program that inspects all establishments serving and preparing food. Each inspection results in a pass, a conditional pass or a closed notice. Find out more about the inspection system.

Yes, Their consumer facing platform does not have the URL , but we do share the name. I have since met the guy who built and runs that program, we are all ok with it, since our Dinesafe is strictly a B2B platform. We actually partner with them, and send them data.

How do you weed out false positives?

Bingo. It seems pretty easy for one restaurant to file reports about a rival.

I'd be happier with a site that aggregated food poisoning reports from various government health departments, but I'd worry about Joe Random telling me he died eating someplace.

But then it wouldn't be crowdsourcing, which is the point.

We do a pretty deep dive on every report in the moderating process. Someone with malicious intent could get one report past us. Getting multiple reports past us would be do-able but harder. Getting them actioned by public health would be close to impossible. Rolling reports off after 30 days hopefully de-motivates trolls, if their hard work is gone so fast.

This wasn't much of an answer to "how", just restating that "we have a way to do it, we promise!". Could you give some explanation as to why it would be close to impossible to create enough reports to have them actioned? What markers do false reports have? It seems you only collect some free form text, and I can't imagine any way you could tell a good fake from a real report. So to a business, driving out a competitor out may very well be worth the cost of a few unscrupulous internet warriors.

From us the data to goes to the governing public health agency. Each spammer would need to individually re-present their case to that agency including a 72 hour food history, and potentially bio testing. The next steps can include epidemiolist, site visit, lab testing and more. The likelihood that everything could be falsified including lab tests with all these government officials is low, and it is only when a case passes through this process that a health agency will consider making an announcement such as "Restaurant X is the source of X pathogen causing X people to be sick"

We collect more than free form text.


So if I went out to eat and got sick, I might think I'm performing a public-service by reporting it.

Then you come back to me and say "Hey your report was weird, go get some lab-tests done"? That seems like a situation which would deter most reporters.

Once we receive and moderate your report. It will go to public health. There are many paths it can take from there. We don't require lab tests, that is just a potential scenario, most common where there is either an important (deadly) pathogen potentially involved, or if it appears to be potentially a larger outbreak, and it is a pathogen that can't be detected in the facility. It's not necessarily ever a requirement. What we offer to consumers is that if you speak up there is potential that your report can make a difference.

Thank you for the clarification.

Also very few goverment agencies release complaint data, and for those few that do, it is not done in a timely actionable accessible way.

Depends on what you mean by weed out. Each report is a signal. A single report on our site is a weak signal but still may result in a) an inspection and identification of critical violations b) discovery of an outbreak c) a false positive. One hundred reports relating to the same place same timeframe, would be considered a strong signal, and more typically results in a) and b). Our approach to handling the false positive aspect is that we roll data off our website after 30 days.

So protection racket as business model. As a food lover,I personally hate this, it will potentially add unnecessary burden and stress to food seller.

A useful trick for reducing the chance of food poisoning with take out is to re-heat the food in a microwave.

You do have to get the internal temp of the food pretty high to do a good job, like 165+, but for many foods it works great.

Won't eliminate the risk from all pathogens but it's not a bad precautionary measure.

I like the idea of Dinesafe. There's not nearly enough effort on enforcing health codes.

Thanks. Yes is is true, many state and city health agencies are under difficutl budget constraints, in some cases having as few as a third of the people needed to meet inspection targets

This is an interesting tool. I can probably name a few that are probably already on your platform. I imagine chick-fil-a & dominoes are already on that list, they are IMO always one step ahead in technology over their competitors.

One of the fastest ways for a restaurant to die off is social media and foodborne illnesses. Only big corporate clients, naming those in the 300+ chain stores would be looking into this technology. My question to you is - do you intend to eventually work down to growing chains and smaller independent restaurants as wel?

While these smaller chains do not have as much money, we are seeing an overall trend towards SaaS solutions in this market. Things like yelp, google, etc are already being heavily used. Do you intend to penetrate this larger but lower revenue market? If so, how would the offering be different at all, in terms of pricing? It seems like right now you are using a quote by quote basis which 100% makes sense.

Regulation. How can I, as a normal end user, trust your reports though? I see you mentioned how you are validating this information, but how can I really trust it? Also, do you intend to notify users based on a specific mile radius of their home/work address (e.g. 25 or 50 miles out). I wouldn't want to get spammed with places I don't go eat out a lot

Do you intend to do any partnerships with Tripadvisor, and or yelp, or google(for reviews) to subadvertise in the same medium? Because that is generally what most people look up when eating at a restaurant if they are picky about what they eat.

Next, what's stopping your company from pulling a yelp? E.g. predatory business 2 business practices, by hiding negative reviews etc. How do you intend to combat this, if at all, or big corporations trying to buy off bad social media if this scales upward?

In addition - do you at any point intend to remove outbreak data past a certain point? E.g. an outbreak 10 years ago at a restaurant might not be relevant today - they might have changed ownership, etc. How do you intend on combating this?

I wonder as well, if you intend on targeting smaller clients eventually (restaurants) would you also be targeting foodtrucks too & or caterers?

What big data analytics did your model use to "predict" chipotle's outbreak? I am curious about case studies.

Did you also at advertise yourself at the chicago NRA show as well. If so, what was the general reception of both small, mid, and potential big clients

1. Yes we intend to offer intermediate levels of service, eventually moving all the way down to a self serve option for a single restaurant.

2. We are still working on pricing for the smaller participants. The offering will likely be slightly different, but this scales very well, so we still think they can get access to quite a few of the features, just needs to be low /zero touch on our end, and the pricing needs to be right for them to ge involved.

3. We have geo-fenced notifications for users, we eventually will allow users to set a custom radius, and frequency and sensitivity, all pretty easy stuff to do, just need to get to it.

4. We are open to partnerships and have been approached about this. Yelp's business model is somewhat predicated upon 'Eating out is great' so hosting disparaging information may be less interesting to them. I think as our service continues to become more normalized and understood and expected, there will be a lot more great opportunities for partnerships

5. While I am sitting in this seat, we will not allow companies to buy away reports. Trust is everything, and I think consumers can figure this out really fast, and once you lose trust, you don't have much, especially with what we are doing, there is much more at stake than a normal review site. I thnk the companies don't want this either. As what we are doing continues to become more normalized, if our service started to become corrupted, then consumers lose trust, and another medium will pop up, and it may be a more adversarial organization. So in the end I think the large brands understand that, and should want us to continue to be the leading platform in this space.


6. We roll data off our site after 30 days, to combat harming restaurants where the circumstances may have changed. We think this is one reason restaurants should actually prefer people report on our site vs twitter, or yelp, or facebook. Our site is not intended as a permanent public record.

7. Yes we have had interest from theme parks, music festivals, caterers, airlines, gas stations, big retail that also sells food. We would go all the way down to food trucks, we get data for all of these, and welcome everyone taking advantage of it.

8. What big data analytics did your model use to "predict" chipotle's outbreak? The relative volume of data we received from Chipotle consumers was staggering - and this was prior to their first outbreak when they were still considered 'high quality' we did not need quants or data scientists for this one.

7. I speak at a bunch of events, but not yet at the National Restaurant Association, they are still fighting transparency on restaurant ratings! It will take them a second to warm up to us. We have started and had some great conversations with them, but it may take a little longer for them to get involved. When I have spoken at restaruant associations or other more broad events (like the Food Safety Summit, or International Association for Food Protection, which has a lot of industry participants) there are two camps. The tech forward companies who embrace data, even where it's imperfect, and are really interested. The remainder is a mixture, when I hear from the critics among these I typically point out 1) people are already reporting on social media 2) What I'm doing was bound to happen 3) Given how responsibly we are running this service, they ought to be glad it is me who is running it. They usually get these points.

didn't expect you to answer all my questions but I would feel more confident using your product now. Personally I don't eat at sketchy joints anyhow, I'm a picky eater myself although I tend to eat anything in front of me

7. I have not been to the food safety summit or internatiionalist association of food protection, but those are tailored more towards corporate clientelle anyhow. The NRA show is just a slog of everything so you would have a response more skewed towards the smaller but larger number of potential clients I mentioned. You are correct in that the adoption process would be much more difficult on this level, but it is also a better way to get your name known just in general among the growing franchisee/franchisor community.

On another note, its important to think of restaurants not as venues of just places where they serve food. Really these big chain stores are in reality real estate companies putting investments into their properties. Wawa and McDonalds are both great examples. Smaller companies generally do not have this mindset so bear that in mind.


question - whats your plan for marketing to end user?

For larger marketing efforts towards the general public, what is your plan here anyhow? The success of any software largely depends on its users sending in reports (e.g. people eating). Are these people going to find your site purely based on google searches?

For instance, if I imagined myself in the shoes of people reporting. If I got sick from a restaurant, I would potentially as well go on a ranting spree and google things like "how to report restaurant for ecoli/sick/__insert term__/etc". And find your site


question - How do you plan on using social media instagram / twitter / facebook / email / etc to end user?

I'm curious how this will work. For instance, there's the end users social media(s). The restaurant businesses social media(s) and regional socia media(s). Your company's social media and possibly regional levels of it too.

You said you will be using SMS to alert end user, but have you tailored a strategy laying out all edge case scenarios?


question - if I search "X restaurant name" would I see a domain name link to dinesafe.com/restaurant_name?

I have a question regarding this. Obviously corporations are going to have a big social media presence, but smaller restaurants (should you most likely be targeting this avenue) have a weaker social media / google presence.

I can see if your service were to really take off, it might inadventely affect businesses negatively. For instance I might search "XYZ restaurant name" and out comes a list on the 7th or 10th listing about a food safety illness report on google.

I know you mentioned earlier you have no intention for businesses to tamper with data entered from their clients. But, what about giving them a 2nd chance? Unlisting the report directly on google, but leaving the report intact on the site. Have you considered recourse on this angle as well?


question - mobile users - webapp or native iOS / android app?

Reporting seems to be a one off thing I can't imagine someone frequently reporting items. Do you, or do you not intend to make a native app? Possibly for searching through reported areas, etc


question - how do you gauge how much a corporate client will pay for your software?

I'm just curious how this works. You don't need to answer this or be very specific I am not entirely familiar on how marketing analysis is done to give an accurate quotation estimate for how you quantity the "risk" they avert and therefore potential revenue savings.

From a salespoint, it sounds interesting and sometimes hard to convey. Its like selling just general IT services - some people will question a year down the road "why do we need IT if they sit around doing nothing and not make us money" mentality. Either through a change of leadership down the road nearing the start of the next fiscal year, etc.


I probably don't have any more questions just curious if you had thought of the above cases

Good! glad to hear it :)

Thanks for the tip re the real esate perspective, I had not thought of it that way. The publicly listed companies definitely have a larger group of participants who care about the outcomes, in their investors. -----------------------------------------------------------------

question - whats your plan for marketing to end user?

Earned media has been helpful to us here, search, and social media have also been helpful. Google search "food poisoning kfc" (or any brand) and you can see we do reasonably well at that. With more resources we can do a lot better at all of these, and there is a virtuous cycle for us of identifying more outbreaks = more press = more awareness = more reports = identifying more outbreaks.

We are also pretty close to roll out languages - we have active users and reports in most countries, and public health as well, but improving languages will help that.


question - How do you plan on using social media instagram / twitter / facebook / email / etc to end user?

There is a ton of work to do here, and a lot of upside for us when we get to it. In terms of the best strategy - this is a work in progress, and as we pull more brainpower in to help us with that we have a high degree of confidence we will get it right.

question - You said you will be using SMS to alert end user, but have you tailored a strategy laying out all edge case scenarios?

We have not ironed out all the edge cases as yet. Our email subscription is the most popular, there is demand for our app as well, but that is down the road. There are some interesting opportunities with SMS and messaging, they need more work.


question - if I search "X restaurant name" would I see a domain name link to dinesafe.com/restaurant_name?

I have a question regarding this. Obviously corporations are going to have a big social media presence, but smaller restaurants (should you most likely be targeting this avenue) have a weaker social media / google presence.

I can see if your service were to really take off, it might inadventely affect businesses negatively. For instance I might search "XYZ restaurant name" and out comes a list on the 7th or 10th listing about a food safety illness report on google.

>> Yes this does happen, there are some cases where we do rank higher than the companies own website. Not our intent, we are thinking through this.

question - I know you mentioned earlier you have no intention for businesses to tamper with data entered from their clients. But, what about giving them a 2nd chance? Unlisting the report directly on google, but leaving the report intact on the site. Have you considered recourse on this angle as well?

>> We offer the restaurant the opportunity to secure message with the complainant (anonymized complainant information). Fast, and positive response to customers can often result in review retractions, and even stronger brand loyalty. If a customer wants to take a review down we do it. We encourage this route. Given we roll data off our site after 30 days we feel like we are doing a lot already, but we are open to refine our solutions on this point.

By the way Patrick if you ever need input or feedback I am very well acquainted with the industry in many different angles. Feel free to reach out to me, I am interested in seeing how your company progresses

I have my contacts on my username

You can always search through posts comments I've made here on hackernews to get an idea of what my background knowledge is in.

Thanks! I will do that. I appreciate it.

>We moderate every report, with both a back end review, and front end/human review of each submission with the goal of eliminating malicious and inauthentic reports

Humans don't scale. How do you plan on reliably doing this when you get larger?

>With our data, we predicted that Chipotle would have food safety issues ahead of their series of outbreaks, and we detected several of their outbreaks, and many others in real time.

Not to sound pessimistic, but I'd also like to know if you had any false flags? Where you predicted an outbreak and it never happened? As you can imagine, falsely publishing something like this could be hugely detrimental to a company's reputation.

Cool idea, just curious about certain aspects of your operation :)


Scale > Great question. We think we can do it. We have a great offshore team with affordable scaling, also we just implemented our first machine learning module, and plan to build on this to help speed up and improve accuracy.

False flags on detecting outbreaks > No we have not. Mainly becasue we don't make pronouncements such as "Restaurant X made people sick and they have a problem". Mainly because foodborne illness can be complicated. It may have been a supplier problem, it may be something related to that restaurant but not that restaurants fault. What our site does make clear is that 120 people reported sick after eating at Restaurant X, and at that point both the restaurant and public health ought to examine the situation immediately. It's up to public health agencies, the epidemiologists etc. to make the official judgement.

False flags on predicting food safety issues (potential precursor to outbreaks) > So far our goal is to partner with companies that exhibit high risk and the health agencies that monitor them. So we have not made any public predictions on that, except I think once on Chipotle, which was correct, but I'm not certain we would do that often. This will probably evolve, and it's something we need to be careful on.

More importantly, what's stopping a competitor from submitting a bunch of fake reports? The authenticity isn't inherent from the text of the report. Adding friction won't help here.

Agreed. The text of the report is only a small part of it. We are looking at a range of other things on the back end. Also a scenario like that is a red flag to us, and it triggers further due diligence. It'will also trigger public health, who will then require people to pick up the phone, it will require a 72 hour food history so on, physical site investigation. It does not work like 50 reports come in, and an outbreak is immediately declared - far from it. That is just the start of further work. Not to say someday the system won't be gamed, but we are doing a lot of hard work to ensure it does not happen.

You throw around high report numbers, but most listed locations have one report.

I would probably like it to receive a warning about a recent report a place has while I'm there.

How about user verification, perhaps even with a verified doctor visit getting a higher trustworthiness rating?

I would consider DineSafe to be a nicer, more memorable, internationally better-faring name than iwaspoisoned.

We roll data off the site after 30 days, so you are only looking at recent history, and also not every report goes live on the site. I think the largest we had reported in this period are maybe a couple of cases citing between 20 and 40 people sick.

Thanks for the sentiment re Dinesafe. Iwaspoisoned.com is not for everyone, but it does fare well on search and being memorable.

A single human doesn't scale, a work force of multiple humans can.

With something this sensitive I would be very wary of machine classification.

Agreed, I expect we would use it to supplement but not supplant human reviews.

Do you plan to offer an API or a app which show venues where food poisoning was detected?

Hi, yes we have an API service. We have had interest in an app, and do plan to release one, we are beta testing right now for IOS and Android.

The cynic in me is screaming that this site is simply a lawyer referral intake front. There are a number of websites that try to brand themselves as public service tool but it's basically attorney advertising except more clever.

That's their monetization. They don't have to sell advertising because the submissions are worth real money alone. I do a fair amount of work in the legal space and these deceptive tactics are really disappointing to see. I'd name some of the top offenders of this but I rather not give them anymore traffic and signals for google. If you're funded by lawyers or are a lawyer yourself please just say you are without the fog of a non-profit community driven platform.

If that's not the case with these sites (no way to really know what sort of handshake deals were made but I'd be surprised) I wish you the best of luck.

Ha. ex-finance guy, nothing to do with law. Thanks for the well wishes. As fyi - I have been approached by law firms for lead generation, but I think only twice ever. As I understand it, succesfully litigating a foodborne illness case can be relatively difficult.

I can't see any way this will be abused.

I'm curious if it will have a commercialization model similar to Yelp! - a la "extortion as a service", wherein the business owners have to pony up to defend themselves against potentially false allegations.

We are not attacking them, nor are the public health agencies, so hopefully they do not feel the need to defend. Also given we roll reports off the site after 30 days we believe there is less of a concern from that perspective. We are not a permanent record like Yelp or other platforms. Naturally they do have to pay if they want to get any of the features and services we offer to help strengthen their food safety program, but we are presenting ourselves as a partner not an advesary.

Yes it's not extortion. It's protection.

Exactly. This is one topic where everyone is on the same page, people don't want to get sick, restaurants don't want to make them sick. We are helping restaurants to do better on their most important core objective.


On the one hand, alerting potential customers to genuine cases of food poisoning is a good thing. You can reduce the instances of people getting sick. Great!

On the other hand, you've now created an incredibly easy way for competitors, disgruntled former employees, or anyone who doesn't like a particular restaurant to mount a libellous campaign of false reports. I understand you review reports and look for multiple instances to see if it's more than just one. But it seems it would be incredibly easy to fool any such system with a bit of coordination between a handful of people. And it's their word against the restaurant's; how do you know who is really telling the truth? (and the restaurant may not even know about the accusation until they find out from your app).

Crowdsourcing can sometimes be effective, but there's been numerous examples of it going wrong, and businesses or individuals being unfairly harmed by accusations that turned out to be incorrect. I hope you've got a pretty good team of lawyers to defend yourself in cases where you get the moderation process wrong, bankrupting a restaurant, and leading them to sue you.

We have not and will not accuse anyone of anything. From a legal perspective, nothing new to see here, people have been posting reports of food poisoning on social media for years, Yelp, Twitter, Facebook so on. We are just enabling it in a way that more easily facilitates positive outcomes.

As for a truly negative outcome e.g. a pronouncement by a public health agency that Restaurant X caused a foodborne illness outbreak. This is not possible with a handful of people, as it requires microbiologists, epidemiologists, environmental health experts so on. Unless that handful of people can credibly fool the science, and all these expert participants.

We do have a great law firm, and we are advised as long as we remain a platform for people to report there is zero risk, and this has been tested many times by companies with deep pockets.

ps. Not tested on us. We have never been approached once by a large organization, or their law firm

Applications are open for YC Summer 2019

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact