advertising, algorithm, alpha, audience, Ben Poynter, beta, blogger, business model, coffee, colour dna, colourdna, commercial model, community, Community Engine, community graph, Community Management, image upload, innovation, interest graph, love, Manly, manly grill, media, mojito, northern becahes, pinterest, platform, recommendations, relevance, restaurants, social commerce, social media, social network, social objects, social platform, social profile, startup, stickiness, technology, Terry Brissenden, user experience
First thoughts on ColourDNA
I’m not normally the sort of blogger that has to jump on every new thing and try and tell the world about it; If you do happen to read this blog, then you know I don’t really do it to get attention – just to document thoughts around business, media, technology and of late – I’ve spent the last year and a half helping get a social commerce startup called Community Engine off the ground.
The thing that happens when you’re working in a startup though, is you’re always, always, always looking over your shoulder for someone else that is building the same thing you are – the fear that they’re doing it better, or faster, or if they’ve been doing it for longer than you have is always present. Are they talking to the same capital investors as you are? And so on.
Anyway, I noticed a few months ago a crew calling themselves ‘Colour DNA’, who were building a platform to ‘colour your world.’ I can’t quite recall what exactly it was they said, but I could clearly see that they were building some kind of social platform that would build and drive a very strong interest graph. I think the thing that grabbed me was that they were using a very similar concept to my own business which was to build what we call a ‘community graph’.
The community graph concept is an extension of the social graph, in that an examination of the community graph brings you places, people, businesses and groups that might be relevant to you based on what you actually like and do. It’s a bit different to the social graph, which looks at your collective social network and tries to recommend you various things based on what your ‘friends’ are doing. The problem is of course, that for most people, they usually don’t have the exact same tastes and needs as their friends, family, colleagues, acquiantances and so on. In fact, when it comes to identifying businesses, hobbies, groups, activities that I might be interested in, I have very little in common with most of my social graph. We are separated by distance (not helpful for finding local businesses), age, lifestage (most of my friends have families now), and a whole raft of other things. The community graph, therefore, only looks at the things I am interested in – then compares me to other people that are also interested in the same thing – then, it figures out what else I might like as a result of that.
So I finally got my alpha-invite, and have been tinkering around in there for the last couple of days.
When you first sign into ColourDNA, it’s a little hard to figure out what you’re supposed to do and why you would want to do it. It’s a very sticky user experience once you get into the swing of things, but it could do with a few walkthroughs to get started. For example, as you get started and you get prompted to fill in your profile, it prompts you to start selecting things that you love – you’re presented with a categorized list of things that include brands, events, communities, media – it’s quite a comprehensive list. The reason for this is to start building your ‘colourDNA’ – which is to say – so the algorithm can start to figure out what else you might like. Unfortunately, for the untrained user it’s not really clear ‘what’s in it for me’ and so it’s a potential drop-out point for a lot of users.
If you do start to select items though, you’ll see both a positive and a negative affirmation on all social objects throughout the site – you can either ‘love’ something, which then gets added to your list, or you can say ‘it’s not for me’ which presumably removes that item from your interest graph. When you do select something you ‘love’, you then get a little popup telling you that X% of people who loved this item also love item X – which then prompts the user to stick around and to decide if they also ‘love’ the second thing. There’s an example below.
The site is really sticky while you’re building out your interest graph. It’s much much stickier than Pinterest, and refreshingly, they haven’t copied the noticeboard style of design like everyone else has of late (yes, we followed the noticeboard concept at Community Engine too!) .
I’m a bit challenged by what users are meant to do next once they have built out their interest graph though. Once you get tired of ‘loving’ things, there needs to be some kind of prompt, or message, or something that suggest what you do next. After a day or two of using the site, I’m kind of looking for a reason to come back, which isn’t quite clear yet. Are you meant to interact with other users in communities? Are you meant to review local businesses (there isn’t a business focus yet)? To help get through the hype cycle, it’d be good to build in these kinds of prompts and perhaps initiate community management sooner rather than later.
There are communities that any user can create within the site. For example, I went to find my hometown of Canberra, and as no one else had done so yet, I had to create an item to ‘love’ called Canberra (I must be an early adopter?). Then, I also had the choice of creating a community called Canberra, to which various social objects could be attached. For example, I could create social objects for the Canberra Raiders and Act Brumbies (local rugby / rugby league teams), and add them to that community. People can ‘join’ a community, without having to ‘love’ it.
The ability to create a community around any social object is a great way to self-propagate a social network. The advantage of this in comparison to existing social networks is that communities are created around interesting topics. Behind the scenes, there is clearly the opportunity to use these self-created communities to understand the linkages and connections between socialized objects – with a view to being able to recommend specific new things to users. EG: if 60% of people in the Canberra community also like the ACT Brumbies, then the ACT Brumbies would be a highly recommended community to each of those people.
Now, this is where I can see future potential in the model and I hope I’m not giving too much of the game away here.
Once you’ve gotten past the initial excitement of finding every brand, location, topic, sporting team, product, invention that you ‘love’ (This morning I ‘loved’ Mojito, The Hangover Movie and a good cup of coffee – I must have had an interesting night), you start to realise that you want to find things in your loca area. For me, I wanted to find my favourite restaurants in and around Manly. Turns out I had to create them myself, which was really easy to do, so I did. I might just add at this point that the upload image function is fantastic, with both zoom and rotate abilities on any image you upload. Much more seamless and usable than most other sites I’ve used.
A business profile – like any other community profile on the site – can have all of the information about the business, photos, and so on. The more people that ‘love’ a particular restaurant (or any other business), the more this starts to get recommended to relevant people. I’d imagine for example that as more people from Manly beach start to sign up for ColourDNA, they’ll start to see these local businesses recommended to them. In fact, on your profile screen, you even see a prompt asking you to add restaurants in, which is a dead giveaway for where the platform is headed.
At this point, it’s way too early to know exactly how they’ll monetize that, but I can clearly see the opportunity is there to access businesses at all levels that might inspire a passionate following in their customer base. I wouldn’t be surprised to see some sort of transactional model develop here, as advertising would be inappropriate for an output based interest graph as powerful as this.
I’ve rambled on more than enough but if you’re a social media junkie or someone looking for innovative audience-building – ColourDNA seems to be one to watch. I’m sure we’ll all be reading their ‘one year later’ slideshare in a with interest. As with all new social platforms though – how they grow into a cash-positive business model that doesn’t rely heavily on advertising will be the deciding factor in how long they last (or before someone buys them for their data/staff/technology, like hunch.com got bought last year).
From → Uncategorized