RecommendBox - Home

Having launched RecommendBox at the end of March we then received a lot of feedback and got to observe how people used the site. We are fortunately enough to know a lot of smart internet savy people who are very generous with their time. It quickly became obvious that the launch version of RecommendBox wasn't right so Scott and I got out a blank sheet of paper and asked ourselves the question - 'Given all the feedback that we've received, and the lessons that we've learned, what would we build now?' Our conclusion was that we would build something very different! So we are now in the process of building a new service, under a new brand, which we expect to launch in alpha form in July.

I'm not a great fan of flogging dying horses and I've certainly been guilty of doing that in the past, so I really pleased that we've been prepared to objectively accept our failings and move quickly. Internet startups are a game of iteration and in this case extreme iteration was definitely required.

So, the next release went up a couple of hours ago. The main feature changes are around the finding of friends on the service and the sending of invites. For both of these you can now just login to your webmail (hotmail, aol, gmail, yahoo) and the system will (as per Facebook or Yelp) search the whole of RecommendBox for you to see who you already know. We do NOT save any of your contact info by the way!!!!

There are a few other cosmetic changes including a fledgling go at Twitter integration (this is live, but hidden and may or may not be of any use…... but, was fun and quick!!). Now just going back over any emails we have been sent as feedback (thank you to anyone that has taken the time to do this) to see if there are any quick wins there. I Iwill put up any feature suggestions on the UserVoice page.

I also added the version to the front page – I decided it was best to share with people where I (we) think the application is at…..... (here is a good article on what all this alpha, beta malarkey means)

Just released some new functionality on the site. Two main things (as the title gives away!!)

• There are now very basic profile pages, you can edit what appears on these pages from the link below your picture on the right hand side.
• The ability to browse recommendations made to your friends. This is through the ‘To Friends’ link under the Recommendations tab.

More to come….....

When Robert and Scott invited me to contribute to the Recommend Box blog, they spun me some yarn about wanting me to write something because I’m a big fan of the site. In reality, I suspect it has more to do with the fact that, in addition to being a Recommend Box fan, I’m also a ‘writer’ and so therefore a fan of the sound of my own voice.

They get content for the blog while they’re busy developing the actual product, without having to hire a professional ‘copy writer’ to churn out nonsense about ‘vision’ and ‘caring about your feedback’ – and I get to ramble for a few hundred words for no money at all. FTW, as they say on the Internet.

But no matter. As a proud Early Adopter, I’m happy to do whatever I can to help. And at this early stage I think the most constructive thing I can offer is a handy five-step user guide for new users who want to get the most out of the ‘Box.

So here goes. You’ll want to take notes…

PAUL CARR’S TOP TIPS FOR RECOMMEND BOX N00BS…

Tip number one: Don’t call it the ‘Box. Making up nicknames for things that don’t require nicknames (see also: TomKat, J-Lo, So-Co) makes you sound like a dick, and no one will pay any attention to your recommendations.

Tip number two: This one’s very important. When recommending an experience – be it a film, a restaurant, a book, a play or a hooker* – make sure you wait until the end of the experience before you submit your recommendation. There is nothing more embarrassing than recommending something halfway through, just so you can be the first among your friends to do so, only to find that it takes a turn for the crap the moment after you click send. By then it’s too late – your friends will be flocking to your recommendation in their droves and you’ll look like you don’t know what you’re talking about.

If you’re in any doubt, simply rent the 2001 Steven Spielberg film ‘AI’ and watch until about the half-way point. Not awful is it? I mean it’s not good by any stretch, but it’s not awful, and if a couple of your friends were Brian Aldiss fans, or – for some unfathomable reason – Jude Law fans, you might recommend it. But wait… keep watching… Wait until the first false-curtain ending. Oof. And then the second, and the third. See how it gets incrementally worse with each ill-conceived apparent finale – they’re under water! his parents are back! some fucking aliens have arrived! – if you’d recommended that to your friends, you’d have looked like an idiot. Even the Jude Law fans would hate you. And those people don’t even hate Jude Law.

Tip number three: Express gratitude, when it’s due. There’s a feature right next to each recommendation that allows you to thank your friends for their suggestions. If someone suggests something you like, click the link. It will make them feel fuzzy and warm and will ensure you’re first in line for their next brilliant recommendation. Of course, if they recommend ‘AI’, unfriend them IMMEDIATELY.

Tip number four: Recommend outside the Box. Not literally of course; Rob and Scott would hate me for suggesting that. No, tip number four is not to restrict yourself to only recommending stuff like books and films and bars. Go wild – recommend something a bit quirky that you’ve fallen in love with. If I could remember the first person to recommend Mini Babybel cheese to me, I would track them down and marry them on the spot. Be that person, but to someone else. And probably not for Mini Babybel cheese – that’s probably just me and Sarah Silverman

Tip number five: Yeah, ok, this one’s contractual, but it also happens to be true: send in your valuable feedback. There’s a tab on the left hand side of every page of the site. Click it.

Recommend Box is still very much in beta mode, which for the benefit of non-web people can be thought of as ‘we’ve got the basic functionality done and there’s an invite a friend link, that’ll do for now, let’s launch and see what happens’ -mode. If, in the next couple of months, you send a constructive suggestion for a feature you’d like to see on the site, there’s a very, very good chance it will be incorporated in future versions and you will be thanked profusely. If you wait until after the first two months, there’s a risk that Recommend Box will have already gone super-viral and Rob and Scott will simply be too busy laughing at their own success and brilliance to give a damn that you think the ‘Store’ sub-category should actually be called ‘Shop’. Actually, they probably already don’t care about that – but there’s pretty much everything else to play for. So get to it. Click that tab. Make history!

That’s it. Use those tips; don’t use those tips. But, whatever you do, trust me on the ‘Box.

(* Recommend Box – I imagine – in no way condones using the site to recommend a prostitute. But if you do, I’d suggest popping her in the ‘people > social’ category.)

So we are back from the valley, feeling inspired and driven to get some stuff done. First change (that anyone might notice) is that I added the UserVoice feedback system created by Richard White. This allows people to leave ideas or vote on existing ideas to improve the service. So, if you have an opinion on something please let fly using either the feedback tab on the main site or by visiting our UserVoice page.

We're in San Francisco until the 30th. We're getting to meet a load of people. Yesterday we went to a great barbecue. Tonight we are going to the Moo.com party at 5pm, then to a Glass House event with Michael Birch of Bebo speaking and then onto a Web Expo warm up party. I think that the best thing about being here is that it is really inspiring.

I'm pleased to announce that we have secured angel funding from Brendan McLoughlin the founder and CEO of Fubra. The funding will be used to develop the product.

With any internet startup the question of when to launch is a difficult one. Should you hold back until you are reasonably happy with the service or throw it out there as soon as it has some base functionality?

The disadvantages of throwing something out there early are:

- The press / bloggers etc. may just say 'So what?' because the service isn't yet significant.
- Users may join and then say 'So what?' and not invite their friends to join.
- Potential investors may say 'So what?' and judge your service (and the success of it) in it's launch form.

These are all genuine dangers which can be damaging to a company and be demoralising.

However, there are also significant advantages to going live fast:

- You can view how people use the service and what difficulties they have.
- You will, whether you ask for it or not, receive a lot of feedback.
- It is easier to prioritise development.
- It is easier to see where the challenges are.
- You aren't guessing anymore.

In most industries you need to get it very right for launch. For instance, you can't release a movie and then tweak it continuous. The most you can do is some pre-launch screen testing and then re-edit based on feedback. On the internet you can launch and then continuously iterate. One of the most extreme example of this approach is Bebo, which was launched after a few weeks of development as an online self updating address book, and finished as a social networking platform. You can view a webcapture of their launch site here. I know that Mick favoured the throw it out there approach, after his experience with launching BabySittingCircle which he and others spent 10 months developing as a 'perfect' service, only to ditch it a few months later because they couldn't get anyone to use it.

Another example is Moo.com which started life as a social networking site, where users could give out cards to people. I remember when I first met Richard arguing with him that it wouldn't work. I was a 'So whater' and I was right, but that wasn't the point. Richard had spotted an opportunity early (social networking -meets- cards / printing), and just needed to find the best way to address it. He did and Moo.com was born.

I think that we are very fortunate to be in an industry where we can / have to experiment. At RecommendBox.com we believe that the flow of recommendations to people is VERY inefficient - that is our core belief - so we are off on a mission to change that. Along the way we hope to pick up users, fans, employees, partners, suppliers and investors. We aren't anywhere close to our destination yet and we don't expect to reach it easily. But we know the start point and know the finish point, and are very sure that if we get to the finish point, we will find something very very valuable.

I had the pleasure of entertaining Sarah on Saturday night. She is over from San Francisco and I was asked to round up a few entrepreneurs from the London startup scene to meet her. The group included the founders of Moshi Monsters, Moo, Kindo, Viapost, WAYN, Buildabrand, Trusted Places, Firebox and Fubra.

It was a fun night and we talked a lot about the difference between the startup scene here and in SF.

Myself and Scott are heading over to SF in on the 16th April to support the Web Mission.

It has been interesting reading the reviews of RecommendBox on various sites.

The most pleasing part for us is that all of them acknowledge that what we are doing is different. They get that we are creating an environment for friends to share recommendations with each other and that this is different from the array of open review / recommendation sites that currently exist.

However, it is clear that we still need to work hard on explaining why we think that a closed system is so much more powerful that an open system. So here an attempt:

You are surrounded by a huge amount of knowledge. Say you have 50 friends, each of them may have 20 experiences in their mind which they would recommend to you. Therefore there are 1,000 things that you would benefit from hearing about.

Now the problem is that at present the flow of these recommendations to you is limited in many ways:

1. People tend to get most of their recommendations from the people closest to them / the people they see most often. RecommendBox is designed to help people to broaden the circle that they receive recommendations from.

2. People aren't that good at making recommendations. We tend to make recommendations in a sporadic manner. RecommendBox hopes to encourage people to make more recommendations, to more people, by providing an environment that reminds people to share recommendations and also provides permission to do so.

3. People trust recommendations from their friends. Open systems can be manipulated and spammed.

4. The human brain, filled with knowledge of a specific person, is very good at judging what someone will like.

5. Open systems require substantial data before they become useful.

So let's now consider a couple of points made in the Mashable Article on RecommendBox.

'I emphasize the fact that these items are shared with existing friends on RecommendationBox because the only people you’re able to network with are those you’ve invited or that have invited you. There’s no larger network to which you can offer your recommendations.'

Opening up to a larger network is exactly what we don't want to do. If I don't know someone then how can I make them a really good recommendation. If I am recommending to a crowd, all I am actually saying is that I like something. But that crowd doesn't know me. This type of recommendation is very weak in comparison to a recommendation made by a friend to a friend.

'As much as people would like to get as personalized as possible, by looking to their trusted network, asking for such a huge time commitment from users as well as limiting recommendations to pre-existing friends is more detrimental than anything, making RecommendationBox little more than an extension of an email blast feature.'

Firstly RecommendBox doesn't ask for a huge time commitment. You can join, connect to a few friends, and then sit back and wait for recommendations - so receiving recommendation is a low time commitment, with high reward. Then when it comes to making recommendation, again this doesn't take much time. You create a recommendation (1-3 minutes) and then you choose who to send it to (less than 1 minute). So for instance I recommended a book to 20 people yesterday and it took me 2 minutes in total. Phrased differently it took me 2 minutes to give something of value to 20 people who I like. That is a great effort to value ratio.

Secondly, yes you can do an email blast, but we don't because we don't have explicit permission to do so. A huge number of people have my email address and I can't remember the last time I received a recommendation by email, without me prompting it. If I prompt it, yes I do get responses, but I don't feel comfortable blasting out an email asking for recommendation, and therefore only do so when I am truly desperate. RecommendBox provides the permission to make and request recommendations.

All of this said, our service will in the future provide recommendations from outside your personal network. For instance we will use collaborative filters to provide recommendations. But we require data to do this, so the first stage is to grow.

Further coverage:

- Techcrunch

- Killer Startups

It is with great excitement that today we announce the launch of RecommendBox!

The idea is simple - everyone is surrounded by friends with knowledge of amazing things BUT access to this knowledge is sporadic, inefficient and limited. Our goal is to change this and to become THE PLACE where people go for recommendations.

There are lots of sites out there that provide reviews, but we think that reviews and recommendations are very different things. When you write a review YOU give YOUR view on something. When you make a recommendation YOU are suggesting to someone that THEY will like something. This distinction is very important to us and forms the basis of our concept.

The version of the site that we have launched today focuses on delivering our core functionality, that allows people to recommend things to friends, and request recommendations from friends. However this version simply represents the start of the journey. Over the coming months and years we will be building on this core, with the goal of creating a powerful service that delivers people the best recommendations for them.

Please join us on this journey...