Show HN: Online Reputation Management for Restaurants | Hacker News Reader

Show HN: Online Reputation Management for Restaurants

bistro.is

24 pointszachflower39 comments

Wow, those colourful tiles are overpowering. Lost my focus immediately. You should consider simplifying them, focus on key features and not too wacky of colors, and more whitespace.

Seconded on the tiles. Maybe flip the current scheme? i.e. go white/light gray on the tile backgrounds, and use color on the icons themselves?

Looks good otherwise, and seems like a really useful service.

Not good:

> Unable to connect to MongoDB: connecting to mongodb://web:###plaintextpassword###@dawson.mongohq.com:10036/bistro_production failed: Operation now in progress

Hmm, as the site is down I don't even know how to contact them. Hope they'll see it here very fast.

Thanks for the notice, Sujan. I didn't realize the MongoDB library we are using wasn't suppressing those types of errors out in production. Fortunately, all the data stored there is just crawl data, as we are careful to silo sensitive data.

Not a restaurateur, but I'd imagine they would be most interested in how you handle negative reviews. Couldn't find info on how you guys handle that.

Is this more of an alert system on reviews being placed?

While there are several review site management services out there that offer services that sweep bad reviews under the rug, we believe that excellent customer service is what makes a good business great. We run our business with that mindset, and encourage our users to do the same.

Currently, when a negative review comes in, we notify the user and provide advice on how to properly handle it. In the future, we will be incorporating some more advanced sentiment analysis to notify users of negative mentions on sites that don't offer a star-rating system (Twitter, Facebook, etc).

Right now, we notify the restaurant owner that a negative review has come in and we encourage the owners to respond to any negative reviews (in a friendly manner).

Having a family in the restaurant industry, I can say that most negative reviews and rants can be diffused with a little discussion and an offer to make it right. With Bistro, we just want to find every mention (from review sites or even personal blogs) and just start that conversation.

http://bistro.is/img/bistro-review.jpg <-- "this is a pretty good of your restaurant"

Great catch! We fixed that early on in the live reviews, but clearly didn't update the screenshot.

Thanks!

Out of interest, how easy would it be for you to adapt the system to manage the reputation for hotels, for example?

Doesn't sound like too much work given that you've said the system works on either web scraping or with API's of other review sites? Obviously you wouldn't want to jeopardize Bistro in it's current format through generalizing the platform for other "industries", but if done right, looks like a good opportunity to me!

It is surprisingly easy for us to rebrand into different industries. We did a test with the golf course industry a little while ago and it only took a few hours to go from idea to working product.

In the top area, under "Location" I just see a comma. Also in the placeholder ("i.e., Monk's Cafe"), "i.e." should be "e.g."

What is happening there is that we were able to get some geolocation data back from your IP, but a standard "City, State" format isn't available (for whatever reason).

That being said, you are absolutely right. It is very confusing. In the event that happens in the future, we will leave that field blank so users (such as yourself) can manually enter a location.

As for the "i.e." versus "e.g.," I was honestly not aware of the difference. You've prompted me to Google it, and the results are super interesting! Thanks for pointing it out! I'll make sure to get that updated.

It detected correctly that I was in Toronto and then the search for restaurants names would all return homonyms all across the world. =(

We recently opted to open up the restaurant searching outside of the target area to ensure we found the restaurant we were looking for (Google wasn't always returning accurate matches within the city selected), however it sounds like it is providing a confusing user experience.

Removing the city selection is something we've talked about, since it is less necessary, but it sounds like something we should move forward with.

Like the design :-) Not sure if other seeing this, but there is a tiny comma near the closed button, screenshot here http://www.tiikoni.com/tis/view/?id=3590fd8

I think what is happening there is that we were able to get some geolocation data back from your IP, but a standard City, State format isn't available (for whatever reason).

You are right, though, that is very confusing. In the event that happens in the future, we will need to leave that field blank so users (such as yourself) can manually enter a location.

Really great idea. Attacking any segment of brick&mortar businesses with this kind of product is a big opportunity.

I tried doing this before with SquidCube but it was a bit too broad, and lacked the required focus to be compelling.

Thanks!

The focus was one of the main things we found to be a differentiator in this business model. Down the road, we'd love to look in to a variety of customers, but we know for a fact that restaurant owners alone could use this. Thanks for the comments.

I see squidcube is no more. Any links to old blog posts or something? I'm running a customer service platform and like to compare implementations.

Thanks

You can share positive reviews, favorite the best ones or even reply to reviews on the original source.

For the reply, do you just link to the original site? Or do you let users reply through your interface?

For the current version, we link directly to the review on the original site, however we are currently working on integrations with all of our review sites to allow our users to reply directly through our interface.

"Bistro was designed with Proxima Nova, Adelle & Coquette typefaces and developed using Sublime Text 2 on a Mac."

What's the benefit of telling people that it's built with ST2?

Telling our users the tools we use doesn't provide any benefit to them. That being said, we are huge fans of ST2, and this seemed like a better alternative to screaming about it from the rooftops :)

Combing over the content on your website, I noticed this as well. I find it to be quite charming and most people won’t read the footer text anyway. I’d say: leave it in.

I design (dead tree) books for a living. In the colophon, I always include information about the typefaces I’ve used (font families for print are usually quite costly — $1-3k — so I mention it to gloat about my purchase as well as wanting to give credit where it’s due), the paper it’s printed on (usually FSC certified), and the eco-inks (if used.) I also mention in which country the book is printed.

I don't mention it’s laid out in InDesign on a Mac and that the copy is styled in InCopy — that’s standard. For programming, there are more options. I use BBEdit and Coda, but I respect your choice and I appreciate you letting us know what you use.

I've always been a big fan of colophons in books and on websites—basically telling people how the thing that you're looking at was created.

It isn't necessary to the end user, but it is fun to show to designers and developers that we might want to work with.

If it isn't necessary for the end user, it shouldn't be on the site.

Put it in http://humanstxt.org/

I like the idea. The fact that your Twitter account has a ton of fake followers is a bit shady, though.

Very nice. Do you have a background in restaurants or did you identify the market?

My co-founder's parents own a restaurant. Being the technical person of the family, he has always been responsible for managing the restaurant's online presence.

Checking and updating the large list of review sites and social networks took more time than he had, so Bistro was originally created as a way for him to keep a pulse on his family's restaurant's online reputation.

were you able to get any other early market or customer validation beyond your co-founder's personal problem? Awesome concept!

I'm the co-founder with the family restaurant…

We showed it off to a number of marketing managers of local restaurants in Boulder, CO and have since sold subscriptions to a number of restaurants in the area. Since the launch, we've gone through a number of tweaks and actually end up removing "features" fairly often. We know that restaurant owners are busy, but we continue to understand what they care about and what information we should drop to make their days easier.

How are you collecting the reviews? Web scraping?

We aggregate all of our reviews and mentions through a combination of web scraping, API integrations, and manual scraping.

Thanks for the response.

What was the thinking behind current pricing structure? I think they are underpriced, but then again I know nothing about restaurant industry.

[Edit] Are you also doing sentimental analysis of mentions and reviews?

Our pricing structure is the biggest point of debate among us, however we are testing different models across a few different marketing platforms and will stick with what performs the best in the long run.

As for sentiment analysis, we were performing a rudimentary analysis on all non-rated mentions (Twitter, Facebook, etc.), however the algorithm we were using was providing too many wrong sentiments, so we removed it for the time being. We are currently evaluating new algorithms and APIs, and may bring it back in if we can find one that can comfortably handle sarcasm (a favorite of the review sites) and mentions with no sentiment attached to it (the phrase "I am eating at <restaurant name>" is neither positive, nor negative).

Nice catch. Definitely a typo, we're only a few months old, but I'll have that set dynamically from now on.

Reply on news.ycombinator.com