I had breakfast a few weeks back with Robert Scoble. Not for the first time, the conversation turned to the subject of context.
It’s an often abused word and a subject close to both our hearts.
In a past life I founded Rummble (created in 2006 to “fix” recommendations for places – we were rather early to market!) and it’s a topic which Robert has written about many times before and is now I understand writing a book about.
I’ve certainly heard the word context more and more the last two or three years. It’s banded around at conferences as if until now no-one has been thinking about it. When announcing Rummble five years ago at Mike Butcher’s first ever London tech event, I was told by Jason Calacanis that the vision was “sh*t and irrelevant”. In one sense he was right. Rummble was far too early to market and ultimately failed as a consumer service.
Five years on, still few if any services are yet truly contextual (as Robert discusses in his blog) but today I’m excited that contextual search and relevance is more relevant, even fashionable. It’s edging its way out of computer science departments up the start-up agenda.
Context is not to be confused with personalisation. Personalisation filters a set of information to prioritise the things most appropriate to you and your taste.
Contextualisation changes information of surfaced to you depending on the circumstances in which an event occurs, it is dependent on the setting, e.g. environmental factors such as location, time, weather, or a human behaviour such as walking, driving, flying, sleeping, eating.
There is plenty of personalisation out there and some of it really quite good. Your TIVO-clone recommends TV shows. Google personalises your search results. Facebook filters your feed of news from your friends. But most incumbent services don’t join the data dots in the way that they should to produce contextual understanding of what you’re actually doing as a human being.
So I agree with Robert, that there is huge opportunity in this space. He describes the problem of there being too much noise. And there is so much noise because services throw data us when we don’t have use for the data. When information isn’t relevant right now, it’s as bad as junk mail in your post box, only continuous.
The next big thing
One of the next David turned Goliath tech start-ups of this decade will certainly be some clever service or app which deals with that “noise” by understanding your own context.
It may be a highly contextual son-of-Siri for your Google Goggles, or may be a wristband a la Nike Fuel which instead of telling me whether I’m exercising enough, understands my entire day: where I should be, what I should be doing, by which mode of transport and in which order.
And there-in lies the problem with applying context and relevance to your service. Doing this sort of clever technology requires far more engineering hours than building an Instagram app, which has a single context of whenever the user takes a photo.
I wonder then what that does to the odds of a start-up in this space succeeding?
The holy grail
It’s already almost impossible to get funding in the European ecosystem for consumer-facing services which don’t have an immediate revenue stream so how will start-ups inventing the next big “context-aware” service survive?
Even in Silicon Valley the appetite for investing in consumer services is on a down curve, because it’s so hard to be heard amongst the thousands of apps and websites being launched. And not just in America but in all of the recently spawned start-up ecosystems around the world. Most of the clusters on this list simply didn’t exist five years ago, not in any significant way anyway.
If future start-ups will need more engineering complexity to have a minimum viable product, what does this mean for boot-strapping start-ups and their potential investors?
Hopefully it means that investors will be forced to be more adventurous in their choice of investments, buying-in more heavily to the big visions of entrepreneurs and understanding that this technology will not only take time to build and perfect but that the market is going to take time to mature.
Well, that’s unlikely to happen.
Currently things are going the other way. Enterprise services are once again becoming more fashionable targets of VC money, and in Europe they never went out of fashion!
Dinosaurs to the rescue
Perhaps it means some of the disadvantage that has befallen the corporates of the world, due to the ease and ultra-low cost of create compelling services on the web, now will be reversed.
Cloud services, third party APIs, better programming languages and universally compatible browsers (well nearly) have all made setting up a starting up a start-up uber easy, even if winning in the marketplace still isn’t.
But at least start-ups now have a chance to compete. I can’t imagine Huddle, who is now a serious challenger to Microsoft Sharepoint, would have survived long had they needed to write desktop software, advertise at exhibitions, send junk mail and advertise in trade journals and then ship CDROM’s out the door.
This evolution has increased innovation and democratised the production of computer software (i.e. apps and websites) leaving big companies struggling to respond quickly enough. Corporate hierarchies with all their controls, checks and balances are not good at lean or gutsy, dynamic risk taking on new products (though that is changing)
But as services become more complex and the software behind them has to become more sophisticated (and contextual certainly requires sophistication) does this mean some of the advantages of the grass roots tech-entrepreneur are being eroded ?
Back to school
Students who have a lower cost of living and can invest two years of their study in creating the very technology they will later commercialise may have an unfair advantage in the world of Web 3.0 (let’s accept for a minute that context is what Web 3.0 will be renowned for). There is good precedence for this; Google (built as Backrub at Stanford) and Siri (built at SRI the non-profit lab) are just two of many.
Even if the window of opportunity for great financial success from simple apps will only be open a short while longer, I am sure there will be other opportunities created by the launch of new killer ‘personalised’ services. The provision of API’s as part of many modern services enable developers to build upon one service to create another and that seems something that is unlikely to go away.
In fact, as the Internet in general continues on a path toward everything being more “open” I wonder whether the days of walled gardens for almost any type of service, are numbered.
The Internet of Things, a very hot sector, is by its very nature the poster-boy for openness and interconnectivity. The data itself is the new gold rush (and not just social data). Over the next five years we’ll be stumbling over more and more of it, as everything from your thermostat to your toaster starts to engage in digital conversation with you and everything else around you.
In summary then, I think the execution of innovation with mobile apps, which is the future of everything, will indeed become harder. It will take more engineering time to get to an MVP.
But these things do go in cycles. Setting up a newspaper 350 years ago required a rudimentary press, a lot of patience and of course some creativity of what you wanted to say. By the mid-1990’s the cost of launching a newspaper was millions of dollars, a large workforce and a lot of sector expertise. Then along came the Internet which made publishing a newspaper as cheap as the cost of a telephone call with a modem and learning a few funny tags called <h1> <bold> and <center>.
Precision manufacturing has for years been the privilege of large companies. Yet 3D printers are set to hit the $500 mark by the end of 2013. That brings complex physical manufacturing back to the home, where it all started before the industrial revolution nearly 300 years ago.
It seems as one thing becomes harder, opportunity elsewhere becomes easier.
Personally, I’m still dreaming of creating a mobile recommendation app which gives an instant intelligent answer to the simple question “Where should I go next?”
To get to a version 1.0, which answers that question exceptionally well (in other words, an MVP that does to existing recommendations services what Google did to Alta-Vista search results) I would guess is probably at least two man-years of development. Understanding context takes time. But somebody somewhere is working on it, along with other fantastic solutions for the barrage of big data we are surrounded by today and will be drowning in tomorrow.
Expect the next Google or Siris to appear very soon, on a mobile device near you.