I downloaded the app and it asked me for a forced login using Facebook, with no other choices.
I was expecting to at least be able to bypass the login screen, browse the app a little, and login once I have something to contribute to it or do something with it, but I never found the option.
I don't mind creating an account, but I do mind sharing my Facebook profile even if it's as little as the name and email. An additional option to just register using email would be welcome.
I did that, but facebook tends to remove those accounts after a while if they show no activity and i lost access to some accounts, because people who forget to have alternative login methods also usually dont have ways to login when your facebook is lost.
I will give the user the option to skip in the next release.
The app was originally designed to have more user interaction (voting on favorite dishes from a menu) but now it's core value is in the Hapibelly rating itself. The majority of users won't need an account.
That said, it's a free app. I'm not spamming or selling anybody's personal info.
It's very, very awkward to share with word of mouth. "Check out [hap * E] belly.com, friend." Friend goes to "happybelly.com." Friend is confused. Webmaster of happybelly.com is happy, yet confused. "With an I." Happibelly.com doesn't work, nor does "happybelli." Friend gives up and checks yelp.
Rotten tomatoes tells me if a movie is worth seeing, not if its worthy of a place in afis top 100. Yelp and others push answers to decor, service, lighting and mood above the most important question of "should I eat here?". I've seen restaurants with the shameful 3 star rating on Yelp with the top reviews complaining about a party of xx's experience in splitting a check while mentioning that the food is good. It turns out there are "3 star" restaurants that are just fine for one or two diners. I'm excited to see a binary opinion to my decision of where to eat.
The idea I had for a restaurant review site when I lived in SF was to allow only positive reviews, and it's basically a binary "did you enjoy yourself?" > "yes"/"yes, very much". And require a photo of receipt within X days of posting the review to guard against businesses buying votes. I figured the restaurants with consistent bad service/food would just not bubble up as much as ones with real fans.
While I can see your point in only capturing the positive reviews I wonder about something. Obviously consistently horrible shops just wouldn't show up but much of whats bad is much less consistently bad and this isn't just true of restaurants.
Bad reviews don't just help me find the contractors that would light my house on fire, rob me and piss on the ashes they help me find the ones who only deliver without difficulties 90% of the time so I can avoid being part of the 10%.
Also people would still be able to write negative things while giving what amounts to a 4 or 5 star review on your site. People after all love to complain. Being able to filter or sort for the bad reviews helps me obtain that information whereas it would be harder to do. In effect all the same negativity would likely be there just without the extra metadata.
My issue with places like Yelp is that people could post a negative review for the silliest of reasons and it can be hard to tell if the restaurant actually messed up or the person had different tastes than me, crazy expectations, or caused the problems themselves. I just don't trust them. If I had, I would never have discovered and repeatedly enjoyed some of my favourite restaurants.
For something like a contractor working on your house, yea it's important to see those reviews. For a meal? I'd rather just take a chance based on a lot of people saying they had a good time. It's what you do anyway when you check out a new restaurant based on a friend's recommendation or just on a whim.
And the whole point is not to have any star system, you either liked it or loved it and that's it. No opportunity for commentary so if you want to complain, go elsewhere. The tagline I dreamed up was "Haters gonna hate, eaters gonna eat."
I admit I never fleshed out the idea, but I thought of it as at least a higher bar against a random person on the internet being able to post a review. I could post a review to any restaurant on Yelp right now, but I'd have a much harder time producing a real-looking receipt for a specific restaurant. I guess I assumed when business owners pay for fake reviews they do it through some shady service that farms it out to other people indirectly.
I actually posted a question to quora, "Is it legal to use review information from websites to calculate an average score?"
Replied:
"In the US, so long as you have the legal right to access the data to begin with (i.e., you didn't crack into someone's system to get it), yes. Simple facts, and things like mathematical analysis of those facts, cannot be copyrighted.
Calculating an average is just a mathematical process. It doesn't require the type of creativity that would allow for copyright."
I'm not redistributing or displaying the Yelp ratings. If you look in the app, the Hapibelly rating is displayed, then all the links to the different review platforms are provided.
I think this might be a concern, depending on how the "hapibelly" score is generated:
>Don’t aggregate or blend our star ratings and review counts with other providers. You may show content from multiple providers, but Yelp data should stand on its own. For example, you may not average Yelp business ratings with other providers to show an overall score.
If you're combining ratings from different platforms, this seems to fall foul of Yelp's developer TOS (and a few others, I believe)
Thanks. I should probably contact Yelp if this app gets any traction.
Also, I think these are just the terms for using the API. I'm not sure if they can do anything if I don't use the API.
Even If Yelp and Foursquare are a no-go, that still leaves 6 other solid review platforms to use.
Yelp and foursquare are useful but not the best. The granularity of rating for Yelp is too big. Over 50% of restaurants are 4.0 stars. (That's why the app is useful) Foursquare is the least used platform with the oldest reviews.
What do you mean you're not using their API? If you're screen scraping the content, you're still violating their terms of service.
"You also agree not to, and will not assist, encourage, or enable others to:"
"Use any robot, spider, site search/retrieval application, or other automated device, process or means to access, retrieve, scrape, or index any portion of the Site or any Site Content;"
Whether one likes it or not, simply by visiting or looking at a site, you're agreeing to their TOS.
"By accessing or using the Site, you are agreeing to these Terms and concluding a legally binding contract with Yelp Inc., a Delaware corporation headquartered in San Francisco, California ("Yelp"). Do not access or use the Site if you are unwilling or unable to be bound by the Terms."
Perhaps we shouldn't be giving our review data to companies like yelp that lock it up then. The user certainly doesn't gain anything from their contributions being sequestered.
Because it sounds like a pretty good idea. Freely distributable attributed reviews under non exclusive licence is way cooler than a db full of yelp scrapings.
I don't really have any issue with Yelp's terms of service. After all, user reviews are their backbone so I don't blame them for wanting to make sure that any Yelp review information needs to be identified as coming from Yelp / under certain restrictions.
Wouldn't seem fair if Yelp did all the work and any app or platform can simply piggyback off of it and present it as their own, or obscure where it came from.
Relevant info: the lawsuit of craigslist vs padmapper and 3taps: "3taps claimed that Craigslist was a public website, so anyone, including 3Taps always had authorization. The court disagreed with this..." (Wikipedia)
I looked into this a while back, but bailed on the idea because it violates the TOS of a bunch of APIs. For example, the Google API terms specifically notes that you cannot aggregate their content[1]. There were a number of issues with other API terms as well, such as completely disallowing any caching or storing of data, I can't recall which ones, but if you read the terms, you'll find they're quite restrictive.
Ultimately I decided not to pursue the idea since there's no real opportunity for growth, since all your content comes from your competitors, who can easily shut you down.
If you ask me, it's hypocritical coming from a company that scrapes the entire web, runs it through an algorithm (PageRank) and puts the data thats in a form that easier to understand. They make 10s of billions per year scraping other websites. Isn't Google the king of aggregators?
But you can't scrape google when aggregating info? What ever happened to "Don't be evil?"
How do you plan to compete with or differentiate from any of the sites you are aggregating? My usual flow for seeing if restaurants are good generally goes like this:
1. google places near me
2. see google reviews for place
3. maybe read 1 review from whichever review company is highest on google's results
4. go back to step 1
Aside from the fact that this is iOS only (I have Android), requires Facebook, and violates the ToS of the services you are aggregating, I just don't see any compelling reason to change my habits from what they already are.
From my habits, I've found you really don't get the full picture using one platform. Google Users will review a restaurant in a completely different way than Yelp users. Zomato users are looking for different things than say a facebook user. That's why they all have different average scores.
1. I have a friend cloning the app right now for android. Many of the things brought up in this post (like requiring facebook ) will be changed
"In the US, so long as you have the legal right to access the data to begin with (i.e., you didn't crack into someone's system to get it), yes. Simple facts, and things like mathematical analysis of those facts, cannot be copyrighted."
Out of interest, why did you decide to build this as an app rather than a (mobile-friendly) website? I would expect the latter to be quicker to build and easier to distribute.
Automation is necessary when you're talking about finding reviews for over 500K restaurants.
The best solution I can think of right now is 1) collecting valid professional review links from around the web 2) associating them with specific restaurants and 3) performing sentiment analysis to determine a rotten or fresh type rating.
These steps are straightforward and would not yield 100% accuracy. Even a 95% accuracy would be unfair to the few restaurants who's rating somehow gets messed up.
That would be awesome! The main difficulty is crawling for the data, I suppose. Movies are very general, so you have a small set of top critics to pull from, whereas restaurants have reviewers in every city.
That's a hard question to answer since 1) it's been off and on for about a year and 2) I didn't start out with this idea.
The original idea for Hapibelly was to crowdsource popular menu items, using your mobile device as an interactive media-rich menu, voting on the menu items you like. It was done with just me as the designer. I created an entire backend that let's restaurants manage their menu, specials, and basic info.
My friend Sara helped me out with the updated look (https://www.behance.net/itssaramarshall). I have to give her a ton of credit for telling me letting me pick her brain and drafting a logo for me. I have another friend working on an Android clone.
The goal of this update was to keep all that in place and add more reason for people to download the app giving an aggregate score. The Hapibelly score became the main value add of this update. I spent more time gathering the data than on the app itself.
This update took me about 3-4 weeks. The first version about 3-4 months. Of course I have ideas for updates to make it more kick ass. With any luck I can pursue those ideas with a bit of funding!
Why don't you add a chart of the individuals rankings that go into your score for each restaurant, similar to the image you advertise on hapibelly.com? similar to here: http://www.cinesift.com/#/
I have legal concerns about displaying other platforms review information in my app. I link to review websites and their info is used in the calculation. Agreed that it'd be great visually for app.
I downloaded the app and it asked me for a forced login using Facebook, with no other choices.
I was expecting to at least be able to bypass the login screen, browse the app a little, and login once I have something to contribute to it or do something with it, but I never found the option.
I don't mind creating an account, but I do mind sharing my Facebook profile even if it's as little as the name and email. An additional option to just register using email would be welcome.
I just uninstalled it and moved along.