The Future of Fraggl


You'll know by now that I'm working with the smart folk at AdaptiveLab on our Twitter curation app Fraggl. The core concept and methodology behind Fraggl - curating the best content shared on Twitter through a combination of human, social and algorithmic curation and delivering it straight to your inbox every morning - has not changed since its inception (James calls it an 'awesome catcher' which I really like).

But we have iterated a lot over the past year that the beta has been live. It's been fascinating how the algorithm has had to change in response to changing use of Twitter which has impacted the balance of social signals that we use as a key part of the curation process. And we've also made improvements to design and content presentation in that time.

So, continuing in that vein we've made some more ongoing improvements including ensuring that every link is presented with a snippet of text that tells you more about what you'll find, changes to deal with representation of content behind paywalls, and foreign language content (sorry non-English speakers but concentrating on English language content first). 

We also have a shiny new website to improve on-boarding and better explain what we do. But the big news is that we have now created a brand new version of Fraggl specifically for the health sector that is designed to surface forward-facing, interesting content and trends around the intersection of digital and health. The source list for this content has been curated by Pritpal S Tamber, a well-known specialist in the future of the health sector. We've been trialling the new product for a few weeks now and it seems to be surfacing some fascinating content so we'll be working to improve that over time. In the meantime, you can sign up to Fraggl health here.

The health product hints at a broader future direction that we want to explore with Fraggl. We already have products that cover advertising and health. This joins the bespoke version we created for Google. And there is lots of potential to develop this concept to cover a broader spectrum of verticals, enabling people to get insightful, tailored, relevant content delivered to their inbox every morning. We're really excited about where this might go. Our subscriber numbers are growing at a faster rate now than ever before, and we'll continue improving the product as we go, launching new versions when we can. We'd like to thank everyone who has been a part of the process so far, and given us such valuable feedback along the way. If you're interested in developing your own vertical, or a version of Fraggl for your company, get in touch. And don't forget to subscribe.

Organisational Structures from Scratch


Ashley Freidlein, CEO and Co-Founder of Econsultancy wrote a recent post on the Econsultancy blog as a first take answer to the question: 'With a blank sheet, what organisational structure would you choose for marketing and digital?'. It's an interesting question, and the answer he came up with incorporated a structure that brought improved customer focus to the organisation, and integrated marketing and digital to a far greater extent with other other customer facing and support functions and roles.

I think that there are some big questions around how we design our organisations in response to the ever changing environment in which we find ourselves, and so I built on Ashley's post with a response of my own which you can read here. As you'll hopefully see, asking ourselves what it would look like if we started with a blank sheet leads to some interesting possibilities, but also the question of whether it isn't time to have a more challenging debate about how we structure organisations (particularly functions like marketing, customer experience and digital) for a digitally-empowered world. I'm interested to know what you think.

Distributed and Destination Thinking (Redux)


It's been almost three years since I wrote about the importance of combining distributed and destination thinking into strategies, and quite a bit longer since I first started talking about it. But the concept has become ever more relevant, and I was reminded of it again last week in reading about Facebook's plans for Messenger. It's not hard to see where a strategy that focuses on people being able to send money to each other (and surely soon to companies too), integrating with third party apps, and enabling Messenger to be integrated with online retailers and other businesses as a direct channel for customer service, receipt and shipping notification, is headed. Like many aspects of Facebook itself, Messenger is becoming a service layer. As well as incorporating interaction via a destination site or app, they are (in the words of The Guardian) weaving themselves 'into the fabric of the internet'.

People may think that the term 'platform' is hype but to me, this is what it really means. The same is happening with Facebook ads capability, as their ad platform Atlas increasingly enables you to amplify the use of Facebook's huge data repository beyond Facebook itself to find and target users across the internet. A capability that, as Simon points out, potentially goes a long way toward solving critical issues of targeting, attribution, measuring reach and optimising frequency, even across devices (and in doing so highlighting the weakness of cookie based systems).

Google takes a similar distributed approach of-course, with it's Ad Sense network, embedded search in third party sites, development of Doubleclick, cross-device analytics, joining up of offline and online. And the similarities don't end there. At F8 Facebook also unveiled an embedded video player, bringing it more into direct competition with YouTube, and meaning that Facebook video (and no doubt the advertising that accompanies it) will now not only sit on the site itself, but be woven 'into the fabric of the internet' via countless other sites.

All of which puts the discussion around Facebook's rumoured plans to directly host publisher content on the site into an interesting light. Media owner strategies have traditionally been focused far more on destination models. On bringing as much traffic back to owned media assets likes websites and apps and keeping it there as long as possible to derive value. Even distribution of content onto third party platforms is typically focused on bringing traffic back to destination sites which might be monetised through strategies that have typically focused on maximising ad impressions, often through maximising page impressions (leading to ridiculous tactics like splicing long articles into multiple page views, or creating image galleries that can generate larger numbers of impressions, or putting banners on timed rotation). Such tactics may provide a short-term minor boost to revenues but sacrificing user experience to revenue is a slippery slope (and surely a lesson we learned fifteen years ago with the explosion in commercial pop-ups).

There is significant potential downside of-course from hosting publisher content on Facebook (captured in these questions from John Battelle), not least potential loss of control, loss of revenue, loss of customer data. But as Battelle also says, there is reason enough to test and learn (whilst walking into it with your eyes open) and it could be that in some situations it could well make sense

Whilst at SXSW I heard Jonah Peretti talking about Buzzfeed's counter-intuitive 'network integration' strategy which values impressions and views that occur in the stream without necessarily an expectation that traffic will brought back to the site. Whilst they might get over 420m referrals in a month back to the site, their content attracts over 18 Billion impressions across networks like Facebook, Pinterest and Twitter.


Peretti described the value he sees in going where the eyeballs are. For most publishers the big question is of-course about monetisation. But my impression was that Buzzfeed believe strongly in the value of the impact and reach that those 18Billion impressions bring. So much so that this distributed approach is a key part of their over-arching strategy:


This requires producing content for a single story in multiple formats for wide distribution, and the payback is the data and learnings they get which feeds back into the content generation work. Buzzfeed have of-course concentrated on native advertising rather than banner ads, which makes sense in the context of this network integration strategy. Whilst banners are destination focused, and arguably failing on mobile, native is highly portable and works in the stream. Which makes me wonder if, in the context of this Facebook strategy, there isn't greater opportunity for publishers to think more laterally about how distributed content might work for them.

But on a broader point, whilst combining both destination with distributed content might make for a sound content strategy, it is often the latter that is harder for people (used to destination thinking) to get their head around. But whether it is people talking about what we do, highly portable content like mobile cards, embeddable content, or the notification layer in mobile, or integrating with other services, there are surely many ways in which we can develop greater opportunity from distributed thinking.

Image courtesy

Why Small Teams Work


(N.B. This post is part of an occasional series I'm doing, drawing from some of the thinking that's going into the book I'm writing - any feedback is appreciated)

'From the Founding Fathers in politics to the Royal Society in science to Fairchild Semiconductor’s “traitorous eight” in business, small groups of people bound together by a sense of mission have changed the world for the better.' Peter Thiel

There is a story (seemingly from a former executive) that whilst at an offsite retreat where Amazon’s senior staff had gathered, some of those staff suggested that employees of the company needed to start communicating more with each other. Jeff Bezos apparently stood up and declared to all in the room: "No, communication is terrible!". Bezos was referring to the potential for over-burdensome communication to slow everything down. Yet many managers at many large organisations still loudly advocate the need for more communication. It’s the kind of rallying cry that very few others will disagree with. The kind that feels like it is just what is needed to solve a broad range of internal issues that need attention. 

I've written before about the power of small teams, and am fascinated by their potential for bringing greater agility and speed of delivery to organisations, and also for generating significant change. But this is not just conjecture. There is plenty of research into how small teams can do this, some of which is summarised by Janet Choi in this excellent blog post on the subject. Janet draws on the work of J. Richard Hackman who was Professor of Social and Organizational Psychology at Harvard University to make the point that the issue with larger teams isn’t necessarily the size of the team itself, but the number of links between people.

As group size increases, the number of unique links between people also increases, but exponentially. So whilst a small team of 6 creates 15 links between everyone, a larger team of 12 will generate 66 links, and a team of 50 has no less than 1225 links to manage. This exponential increase means that coordination and communication costs are soon growing at the expense of productivity. Hackman, writing in The Psychology of Leadership, explains that:

“The larger a group, the more process problems members encounter in carrying out their collective work …. Worse, the vulnerability of a group to such difficulties increases sharply as size increases.”

Leaders, says Hackman, may often create oversized teams in the faulty assumption that ‘more is better’ for team effectiveness, or due to emotional considerations such as sharing responsibility and spreading accountability across larger numbers of people, or for political reasons such as ensuring that all relevant stakeholders are represented:

“For these reasons, individuals from various constituencies may be appointed to a team one by one, or even two by two, creating a large politically correct team - but a team that can find itself incapable of generating an outcome that meets even minimum standards of acceptability, let alone one that shows signs of originality."

Janet also mentions research conducted by Bradley Staats, Katherine Milkman, and Craig Fox (The Team Scaling Fallacy: Underestimating The Declining Efficiency of Larger Teams) showing that larger team sizes can lead to overconfidence and an under-estimation of time needed to complete tasks. One experiment conducted by the researchers set different groups the task of building the same Lego figure. In spite of the fact that the larger teams were almost twice as optimistic about how long they’d take to complete the task, four person teams took 52 minutes whilst two-person teams took only 36 minutes.

It’s tempting in digital transformation to think that since the outcome is so important and speed (in delivery of transformation or digital development) is often such a factor, more people will lead to a greater chance of success. But we underestimate the increasing burden of communication at our own cost. Small, nimble teams can achieve amazing things. So rather than throw numbers at a problem, ask yourself this - what's the smallest number of people you can put together to achieve a result? It's likely to be less than you think.

Image courtesy

Google Firestarters Austin - Engineering Strategy - The Event


Saturday saw our first ever Firestarters event in Austin, held in the Google Fiber space during SXSWi. We'd themed the event around the intersection of user experience design and strategy which, judging from the audience feedback and debate on the night is rich territory for discussion right now. Perhaps unsurprising given the growth in importance of UX within both clients and agencies, and more generally how tech-savvy design resourcing, expertise and practice is in ever increasing demand (as encapsulated nicely by John Maeda's Design in Tech Report that was launched at SXSW).

Our first speaker was the brilliant Oonie Chase, Director of Experience at Wieden & Kennedy, who talked about how UX was influencing briefs in interesting and sometimes uncomfortable ways at W & K, but also how the two disciplines will likely start start from a different place - planning from the lens of the brand, UX from the customer perspective, and what they can teach each other. She used a quote by Dave Terry at W & K ("It's entirely wrong, but it's golden in its wrongness.") about UX:


User Experience practice challenges the idea that you need scale immediately - you might start small to acquire learnings, whilst advertising has a ‘need for glory’. UX is something that is co-created over time so not necessarily perfect. It doesn't worship creative and 'doneness' in the same way as planners do. But whilst UX can be overly focused on getting from A to B, planning can teach UX about not losing sight of the ‘soul’ and emotion of what you do. Rather than aiming for a Minimum Viable Product, perhaps it's more about achieving a 'Minimum Lovable Product'.

Chloe Gottlieb, SVP Exec Creative Director, R/GA, complimented that nicely by focusing on where planning and UX (or rather Experience Design, which is the term they prefer at R/GA) overlap. Great planners, said Chloe, are inherently creative and great Experience Designers are inherently strategic, but it's how they work together (as opposed to working in parallel) inside the agency that is becoming increasingly important. Both planners and experience designers are pattern seekers, consumed by consumers, and obsessed with culture and behaviour, but:

"While UX folk might veer more toward architecture, engineering and design - closely observing customer needs and how to add value to them over time - planners are more like poets, anthropologists, psychicians – finding tensions in culture and bringing them to light."

UX insights might lead to products, services and platforms (systems), planning insights lead to brand stories, content, comms. R/GA's own progression has amplified the overlap, growing into making more branded interactions and systems for campaigns, becoming more strategic (e.g. thinking about functionally integrated services that tied products and services together), just as planning comes from brand and storytelling towards having to create strategies that could lead to products, services or communications, and is evolving to find ways to stay involved as work is made, tested, evolved over time. 


So it is far from binary, there is plenty of overlap, and planning and UX need each other more than ever. The intersection is where the magic happens, particularly now that most brands are primarily experienced through interfaces (in fact 'Brand is Interface'), product and message are integrated more than ever, and experiences are not separated into silos. Strategy becomes 'truth over time', requiring more system thinking, experimentation, tweaking, QA than ever before. At R/GA, getting to insights involves multidisciplinary teams working together on the problem, looking for patterns and collisions, and briefs are the crystallisation after the ideas are baked. The way they know they have a great insight is when it enables creatives to take giant leaps.

Ian Spalter, UX Lead at YouTube, added a third unique perspective. Moving to a tech company from an agency background, said Ian, meant coming to a place where big ideas are not as important as big releases, where impact is more important than inspiration, and from an environment run on creative disciplines, to one dominated by engineering. His tool set and collaborators have also changed (he counts algorithms amongst his new set of collaborators). Planners dig for insights, and uncover or manufacture a truth, and creatives make fictions (through stories and designs). So the flow of the process is that from an insight you get a big idea, and from that big idea you tell or make a story, and in software especially you create an experience for the customer. But in software the distance between creating a story and creating an experience can be a long and winding road. But increasingly both marketing and product now focus on the essentials of the experience they are trying to create at the end. This may be in a story like form, but stories illustrate a promise, and all promises are lies until we keep them. So 'planners tell awesome lies', lies we can believe in.


And whilst software companies understand their customer’s behaviour, they rarely understand their customer.

Russell Davies (Creative Director, UK Government Digital Service) built on his Firestarters UK talk by focusing in on how usability increasingly trumps persuasion. In a pleasingly controversial talk, he talked about how the product is the service is the marketing, and why this meant that experience design was the future. A brilliant product will always be better than a parity product with marketing. It was once hard to produce brilliant products, but as everything becomes increasingly digitised it's easier than ever. So companies that are still set up around persuasion need to design around the needs of the user, and be set up to deliver the best experience possible. This means no new ideas until everything works ('fix the basics'). And that user experience is killing marketing since if the product is good enough there is no need to over sell it. And about how everyone should be concerned about making user experience better.


The questions and debate afterwards picked up on just how topical a subject this is right now for strategists and agencies, and we had no lack of interesting (and sometimes controversial) opinion on the night. You can see a Storify of some of the feedback and conversation here

My thanks as always to Ben Malbon and Google for hosting, to our amazing speakers who made our first Google Firestarters in Austin such a success.

Thanks to DDB Worldwide and ImageThink for images used in this post.