Monday, 16 November 2015

On communication and fragmentation

The Rosetta Stone, British Museum by Gary Denham

I've dealt with work flows that result in communications for many years and one trend that I've noticed over the past 5 or so years is that unlike previous communications channels, which might have remain relatively small and fragmented before they become widespread, there are now various avenues for communication and people will be comfortable using more than one; even from the same device. For example, I have friends that will use email, Facebook messaging, WhatsApp, Snapchat and SMS (or iMessage) between the same set of people depending on both the context and content of the messages.

Some background

I was recently at a conference where Patrice About from AirFrance made the point about various customer service aspects for airlines now don't have a single way of achieving that service and passengers expect to use the most convenient method available to them, for example for check-in they want to be able to choose between online, at a counter, self-service kiosk or using an app.

This blog post has been in my drafts since the very early days of writing it, it has been interesting reading the new articles that I collected showing how the future might develop. For example although QR codes and NFC have been contenders to provide localised services, neither have really achieved the potential that either has worldwide - although Apple including NFC in its latest products may change that.

So what choices does a service provider have? And which should they choose? Here is a short round up of the main communications channels...


Email has become ubiquitous in modern life as it is quick, asynchronous, "free" in a per message sense. However, for bulk or any kind of commercial send there is a need to consider deliverability - from infrastructure to authentication and reputation. The first quote that sums up email is
The second, which I can't find a reference for, is something along the lines of 
Email is a to do list created by people who don't know your priories or value your time
This really resonated with one of my colleagues, perhaps because it is so easy to send emails it has created a love hate relationship along those lines for me as well, especially since I find typing emails on my phone so fiddly!


Voice messages can be synchronous, which is good in gaining immediate feedback, but bad in that it is a slow send channel to use with the amount of time to send tied to that to receive it. The main value I can see is that this can be very good way of extending the call centre and dealing with the easy cases running an automated script. Difficult cases can seamlessly be transferred into priority call centre queues to be dealt with by a human,


Compared to email SMS never really fully flourished, with smartphones and apps curtailing any serious use of MMS by businesses, although shortly after I started writing this post in 2012 some still thought MMS was the future. In selecting SMS gateway providers there are thousands of choices, but few provide a consistency of service that you can receive with email providers. Some of the issues that you have to deal with when sending text messages include ported numbers, various number formats in collected data, different message formats and character sets available, even differing handset capabilities ... although I'm not sure if any of those features like Nokia lock screen animations even apply anymore!


Interesting recent development, requires smart phone, tied to device. The way that Apple and Google have evolved the frameworks to make other devices such as smart watches hook into general notification services means this has become less of talking point in the last year or so. I must admit I thought this would have had a bigger impact than it has when I started this blog. It has various negative points like being tied to a device or app and requiring network/battery power that email might not have, but does have massive bonus in immediacy that even mailbox alerts don't always provide - I've missed a number of emails recently where the push notifications didn't come through!

Social media

Something I hear in advice to use social media is that "you need to be where the conversation is" but, in my opinion, this only tells half the story therefore it's only half a piece of advice. You need to factor in why the conversation is where it is and what it is about. Buying a service by one channel but then being communicated to exclusively by another channel can be a disconnect, for example WeChat has a bit of an ecosystem to it that if you were to make an airline booking there but got your communications via email it might seem disjointed.

For social media I would argue that you do need a much better idea of the demographics of your users and their needs than you might get away with using email as an interface or communications tool. There are several different models such as public broadcast, public conversation, private notification or private conversation? Then you need to think for your geography and demographic which network? Facebook and Twitter are both popular globally but regional sites such as Renren in China, VKontakte (Russia’s Facebook equivalent) or Hyves in the Netherlands might need to be considered. In fact behind the great firewall of China a local platform like WeChat might be the only viable and operationally effective choice.

KLM have made social media expertise their calling card, although I would imagine a lot of debate goes on internally in treading the line between not over investing in marketing gimmicks - the social seat maps looked like a fine line here - and being careful not to miss the next big thing or not capturing a younger market. To me they embody the choice of channels approach, it remains to be seen how many of these initiatives will last the course or they will become more targeted to the niches they serve.


Finally, this tweet says what I feel about fax ;-) ...

In summary, my advice for choosing the right communication medium is quite simple - know your audience and know what's appropriate in your context but never forget your users' preferences. Also don't forget that expectations change and you need to live up to them for good customer experience...

 Further reading

Friday, 30 October 2015

FORUM: SITA's Europe Aviation ICT Forum 2015

Leaving Athens
Following on from my last trip to the SITA Summit in 2014  this year's SITA Forum in Athens had similar themes, but with a more airport focused audience. I think my biggest take away is that self-service is important; not just as a method in cutting costs but also in improving the passenger experience (#PaxEx) via two main factors - control and information.


To pick out the two salient illustrations of this, for control I'll turn to Patrice About from AirFrance. He made the point that consumer bag tracking devices now available for about $50, and these aren't under airline control or put in baggage with their knowledge. So the industry faces a choice - either find a way to work with passengers bring their own device or bury head in sand and have pax that are better informed than airline agents when a bag has been put on the wrong aircraft. This trend is happening now, the good news for airlines is that bag tag devices are GPS/sim based and most bag losses happen in a hub's bag room ... Where cellular networks and GPS don't work ;) (note that it is in the hubs that bags are lost, not smaller outstations. So hub bag rooms need the bag tracking)

Later in the Q&A session Patrice also made the point that there no such thing as one option anymore, need to offer multiple channels for achieving solutions (e.g. home printing might not have colour needed for green strip). so there may still be a place in the future for home printing, kiosks and human agents. My personal spin is that I wonder how long people will have printers at home? Will the trend of tablets replacing printers mean we have less of them in future? I know that I haven't had a printer at home for a while and prefer Passbook boarding passes.


For information there are two main issues, giving the airlines, airports and government agencies the information they require - and minimising the stress of doing so on the passenger - and giving the passenger information during their journey.

Chris Annetts from Heathrow airport summed up the passenger requirements best with this slide

there was lots of discussion about how to deliver this and who is best placed to do so. There seemed to be a broad consensus that different stakeholders needed to collaborate more. For example some airport apps are great but people tend to use airline apps or if they do use the airport mobile
apps then the experience drops off as soon as they leave where the mobile apps covers.

There was a lot of talk around the "Passenger journey is an emotional one". Overall from a SITA survey they found an average 80% happiness across the journey. Making the booking, just before boarding and on plane are the happiest times. Queueing in security is unhappiest time (down to 64% positive) with the baggage claim step of journey also poor. 

People are moving beyond smartphones ..
Again there seemed to be a broad consensus among the speakers that there is a need to integrate multiple touch points, like kiosk check-in or agent check-in and BagDrop. The question was posed "Can collect biometric data (photo taken at check-in) then use to allow access to restricted zone using facial recognition?" - this simplifies a lot of the current process with security and boarding passes - and again at gate. 

There were some good points made by both Krum Garkov, when talking about the EU safe borders programme, and Todd Frew, when sharing his experience in the Australian Immigration department, that advanced passenger information schemes help push the border to the point of departure; which in turn allows more time for government agencies to process the people requiring entry. Todd also shared that since introducing the electronic visas in Australia the number of fines to airlines for transporting undocumented or "undesirable" passengers that were then refused entry had fallen dramatically, so a win-win for the government and the airline industry there.

What next?

I'll leave you with this final thought - the current process of getting travellers through to their destination is a lot of separate steps, using data with no real connection to previous step .... Could we collect this in one go? (so physical security checks, indentity checks, validity to travel etc.) Could our biometrics be our boarding pass? This could be win-win, less overhead on airports, fewer queues and a seamless journey means fewer anxious passengers.

I think that we can, a definte trend I've noticed in my work at 15below in the past few years is that good customer experience and operational efficiency can often go hand-in-hand.

Further reading

Sunday, 25 October 2015

On post-technical and beyond

Photo by Matt Cornock (flickr)
I have been messing around with computers and first started writing simple code about 30 years ago. Since then I have gained a degree that contained a large "technical" component analysing and solving problems with code. 

My professional code has been experienced by hundreds of millions of people from organising school transport for special needs children, overhauling part of the British rail infrastructure, supporting the programme management of a massive bank refurbishment in the UK to more recently supplying travellers around the word with tickets, information during disruption and the chance to feedback to their travel company. I have also written a content management system in one language that I didn't use professionally and some small open source utilities in another. 

I don't list out these achievements to convince you of my ability as a programmer, far from it, at my best I was merely average. My code was never truly idiomatic in whichever language I was writing and I'm sure my style was idiosyncratic. I have felt a bit of a fraud for a number of years being described as "technical" and this year heard a term I can identify with ... "post-technical". I can't remember where I first heard this term but it may have been from someone related to ThoughtWorks, where it is apparently a bit of a disparaging term.

The reason I like the term is that to me it signals "Yes, I used to code. I can speak your language (to a point). I can understand the trade offs that you are making (to a point). No, you won't have to explain that cutting testing is a bad idea or that unit tests will save time in the long run." 

Although it's not just a label of a state of being. For me it also describes a journey. When I first stopped writing code every day the rest of the team were still using the same tool chain and languages, so it was easy to get involved in pointless arguments essentially about personal preference. Next as the languages and tools changed, this moved to more abstract approaches but probably still framed in too a personal view of how I would have done it.

Given that I haven't written code for pleasure in about a decade, and professionally for around three to five years, I was surprised this month to have my interest piqued by Heroku and playing with Clojure, R or even Prolog! Perhaps I have just got happier with my expectations of the output of my efforts, maybe there is just some idea at the back my head that needs expressing via code. 

Perhaps I should step away from the code editor, embrace being non-technical and do something that I'll be more useful at instead! With the stories and tweets about how everyone should (be able to) code it is easy to be guilty about not coding. So I say if you are also going post-technical rejoice in transfering your skills in being able to analyse a problem or how to work with technology, structure data or optimise processes. 

I would love to hear from any fellow travellers from professional developers to the non-technical world ...

Tuesday, 15 September 2015

On AI and the future

The Future Soon by K Rupp on flickr
Last Friday was the last dConstruct (at least in it's current form). For the past ten years this has been an interesting design conference held in Brighton looking at the culture/technology intersection.

As a Artificial Intelligence graduate the subject matter was right up my street and changed the way I look at things just a little. While at University I had a certain nostalgia for the golden age of AI in the late 50s to early 70s, to have been around while Winograd, Weizenbaum, Schank, Newell and Simon were writing the papers that founded approaches and I was studying all those years later.

What changed after dConstruct was that I now almost wanted to have been born later, with advances like 3D printing, robot kits that retail for around $1,600, and 10 companies with self-driving permits in California it feels like an exciting time to be studying, creating and entering the job market. 

Looking at my Twitter timeline, this doesn't seem to be the general consensus in the press though. With fairly sensationalist stories like How to keep software from stealing your job and Will a robot take your job? not uncommon. I do think the rise of automation is a challenge, one that feels like it could be similar to that faced by the hand loom weavers or labourers when the steam powered drill was introduced. However, I also feel that this challenge will be to find ways of working where the new technology helps you or provides entirely new work and even industries.

That's the optimistic picture of the future I chose to believe in. I can't predict which "creative" enterprises will be safe, since even recipe creation can be a done by computers now! Although there is something utterly believable in the scenario described in Hello and goodbye in portuguese, in this short story a world is revealed in which we slowly work out how to replace jobs with automation and it gives us a glimpse into how humans might react.

This change won't be easy and we have to start thinking about designing that future and working towards it now. Some writers like Andrew Mcafee can't wait for a future when there is no work required by humans and think we are close to this world, although I feel that this chart tweeted describes what work will look like in my lifetime

A couple of areas creeping this way already are better analysis of data in real time and disruption of current technologies in travel to improve how we get around. Going back to the driverless car example, I think that an incremental approach like BMW's will be how we get there. The car they are working on deals with the standized and routine, while the human will for now deal with the more creative.

I don't feel that the world is ready for the big bang completely automated cars yet, as the saying goes "artificial intelligence is no match for natural stupidity", for starters the legislative process and insurance market will probably combine to restrict usage in the short term. That means that we probably have time to reflect and think ... what kind of world would I like to live in? How can greater automated make the world a better place?

Further Reading

Tuesday, 18 August 2015

On storytelling and stories

Stories & Storytelling_1956 by Sterling College (flickr)
In a lot of software projects the tools - such as prioritisation, backlog grooming, MVP - and artefacts that they produce - such as job stories, roadmaps, personas - are useful but ultimately I believe that they are about allowing a community of people to tell a story. 

When "going up" to people at budget sign off level they are generally short of time, so the stories need to be the point and tightly foccused on giving a sense of what has happened and what will happened, including topics such as:
  • What are you doing? 
  • Which business goal does this achieve? 
  • What is the value? (e.g. either cost saving/revenue protection or increase in revenue)
In my experience when going into detail level with implementers they like to hear the answers to these questions: 

  • Why is it needed? 
  • How does this fit into bigger picture? and
  • What will we be doing next?

I have tried experimenting with storytelling in roadmapping before. This seemed like a neat way of expressing at each milestone the context of what could be developed and how the different stakeholders/cast members behaviour had changed ... and indeed how the project team were involved. That project is no longer running and I am using a more traditional roadmap at the moment, but  I may revisit this idea in the future, as I think the format is a natural and relatable way of communicating a vision.

One thing I did notice was a difficulty in creating those stories and which facts to include? What should be left out to prevent a 100 page prose version of a progress spreadsheet? More recently I have found the following that goes some way to describe this problem:
"The facts are really not at all like fish on the fishmonger's slab. They are like fish swimming about in a vast and sometimes inaccessible ocean" What is History? by E.H. Carr

I love that quote on historians and history, and it sums up an important part of storytelling for me. What's the Sea Bass and what are the sprats? that's to say what out of all the facts available are the interesting ones? 

To help improve my skills in presenting these facts and being more "persuasive" I recently went on Storytelling workshop lead by Tobias Mayer. To use E.H. Carr's metaphor this explored a great framework and process to find the fish and catch them. The other, more unexpected, lesson was around "persuasion", it turned out that the meaning of this word I really wanted was "authenticity" and "connection" - in both the scenarios I described with budget holders or implementers these are key facets in communicating ideas according to your audience.

Looking on Twitter these two quotes show what could happen when you make an authentic connection ...

This takes what I have seen about stories being good communication tools to another level, can stories drive behaviour and then build culture long after they are told?

Further reading

Monday, 20 July 2015

On tools and technique

Photo of Bellagio by me
Bit of a parable about chasing silver bullets ...

I have been a keen photographer since mid-2008 when I got my first SLR. Since then I have taken tens of thousands of photos, practising my technique and getting used to the tools available - for example natural light, flashes or particular lenses.

Most of these photos never see the light of day and get written off as I don't like them for some reason or other, but I can't bring myself to delete them and every now and again go back and look to see what I can salvage. Each time I then fall into the trap of Gear Acquisition Syndrome!

I have been doing that recently and looking at what themes I can find in my photos and how they can be edited to fit in a series - at the same time as pricing up the latest Fuji X-T10. This time there are a couple of photos that have not only gone from the "nah" pile, but are now being shared on the Internet with other people!

So, what has changed? ... The only difference has been how I look at the photo and what I see. In the past seven years I have lusted over gear and editing tools and although I do now use Aperture over DxO - as its more suitable for my work - and I have a much better lens, ultimately I had the gear in my hand to produce a photo that I could be proud of. The effort that has paid off with improvement has been getting better at manipulating photos. I doubt the photo that illustrates this article would have been much better if I the different camera I lusted over or the more expensive lens attached that I thought I needed.

What has this got to do with software products or software development? ... well, it struck me that it was a similar situation that I have encountered in my professional life. When things aren't going "right" you might have the tools to hand to fix the problem but you might be making the wrong choices in where to invest time and energy to get the best out of existing tools.

Wednesday, 3 June 2015

On done and successful

"Success never slee... by icon" by MsSaraKelly (flickr)
Do we declare success too soon? That's a question I've heard posed by Gojko Adzic in a couple of talks plus a workshop that he ran with Dave Evans and it has stuck with me ever since. 

To paraphrase most projects or initiatives conflate "done" - for example as in "dev done" - with "success", when really it should be on delivery of value in a business outcome. So it maybe that we declare success in producing a new running shoe, when really it should be when the athlete who buys it runs a personal best. Or maybe creating a Video ad that is actually seen.

What success looks like and when we reach it matters to me as I want to keep focus; I also feel that the goal of success, and what it really means, is a key navigation aid in maintaining that focus. One product that makes me think about this is Google Glass and how the delivery of the product wasn't the final hurdle, there are plenty of stories about how it was a failure, creepy and generally suffered an image problem - the kit wasn't too bad (even if the screen a tad too small to be useful) but something about the vision in what success meant and the place Glass could take in the world just didn't work.

While Google pulled the plug the technology behind the tiny sensors that enable "disappearables" make products like Glass and other wearables look like a steam punk vision of the future. So solving yesterday's problems looks like a pitfall, and there are plenty of products that did this - for example remember mini discs? who would have thought broadband and a nifty encryption algorithm would change the market so much? It's a similar story for APS film, poorer quality but more convenient than 35mm film, trouble is that the gap with digital cameras didn't last long and they are a lot more convenient!

I'm no Steve Jobs but here are what I think are three things to help define what success looks like in your context 
  • Know your market - know if you're catering to a niche or a mass market, then resource and price appropriately (did Google do this with Glass?)
  • Keep an ear out for related technologies - but don't be suckered by shiny toys (although there is money to be made from those who are!)
  • Be in a position to take small bets - even if it means you may cannibalise a small portion of your product or you could create a whole new market

It's also worth considering this very wise advice ...
Whatever you do, don't congratulate yourself too much
Or berate yourself either
Your choices are half chance, so are everybody else's

Baz Luhrmann - Everybody's Free (to Wear Sunscreen)

Further reading

Friday, 15 May 2015

On things changing and staying the same

Mainframe Computer by Dave Winer
I recently spent some time with my maternal grandparents. My grandfather often likes to tell us stories about his past career and this visit was no different.

Previously I have heard about his work in London as an accountant using earlier computer systems, such as those pictured to the left. I don't remember any stand out stories or great surprises from this era, other than how similar it was to my own experience working on a Y2K project more recently.

However, this chapter was more recent and dealing with his time in the small Sussex town they lived in while I was growing up.

I listened to him explain how he had first started to sell personal computers with a business partner and then later to bundle with Sage accounting packages as a reseller- this was when PCs were very much a business tool and before Sage had grown as large as they are now. 

Afterwards it struck me how the two different markets are so different now, large enterprise systems are by and large still massive projects "requiring" special data center setups with expensive consultants. On the other hand if you are buying systems for small to medium enterprises and you have a range of options, from cheap off the shelf computers to turnkey solutions over the internet. I cannot imagine someone selling bespoke systems for general sale. With app stores, open source packaging and delivery over the internet it's not as needed to  rely on a third party to market and sell your software - although arguably that's what app stores do.

I wonder what the next 30 years holds? Will "Enterprise software" still exist? Or will it be like whatever personal computing ends up being in 2045? I am pretty certain - and I'm sure this doesn't make a brilliant fortune teller - that personal computing will be radically different. 

Vaguely related reading

Wednesday, 6 May 2015

on automated and manual experiences

As my running season finishes again, I've been looking at buying the photos of my personal highlights. :ast year I wrote about my experiences in on buying behaviours and usability and this year I have spotted other annoyances and lessons to learn.

Untagged photo search options

This year I started the Brighton Marathon 10km with two other people, so after looking at my photos I went and had a look at theirs. When I did this I noticed that there were some of my that weren't tagged with my bib number.

No problem! I thought, there is a handy search option here. So I entered the both the colour and type of my top and shorts and hit go ... and none of the results that came back contained me (or runners matching the description I had given!).

Luckily I have worked in IT, creating various document and workflow systems, for a number of years; so my eyes spotted a reference number on each indivdual photo - in the example to the right BTK00124.

So my next step was to send off an email via their "Contact us" page containing the references to the missing photos and waited patiently. The next morning I had a very friendly email explaining that I hadn't been tagged as I wasn't "fully shown". The prompt response and my desired outcome meant that my customer experience wasn't impacted too badly.

You don't have to create every feature with code, e.g. tagging any photo with your bib number, but you do need to make this deliberate and easy for people to do. A "tag me in this photo" button would be one improvment, a more general "contact us about this photo" is another that covers more use cases.

Given I like to tie my experiences back to my day job. What can this tell us about travel and passenger communications? For me, it's that automated is both user friendly and useful but sometimes you need to help the users if they need to contact you by alternative methods. So if notifying a passenger of disruption let them know of how they can proceed to change plans using your automated system, or give them the reference numbers and contact details to do so with a human. Making it easy keeps the cutomer experience flowing, while hunting around for who to call etc. can give an impression of poor customer service.

Should we design better products for older people?

This week I've been having a bit of think about products for an older audience, prompted by this tweet by Tom Peters ("The red bull...