Back on 26 April I was asked to present "5 Key Challenges for the ISV CTO and How to Beat Them!" at a Ciklum seminar for ISVs
that intended to deliver a hype-free conversation among CTOs, Chief Technical Architects and other key executives grappling with the journey to the cloud. My slides for the session (see below) are already on Slideshare
, but they are mostly visual, so I decided to do this comprehensive (that means long right? - Ed) blog write up following the slide sequence as a companion piece. I was in good company, because the other speakers were Jimmy Gasteen of Precursive
, Liam Hogan of OpenText
and Melissa di Donato of Salesforce.com
. My pitch was intended to do three things:
- Give my perspective on the current state of the Cloud landscape
- Offer my 5 key challenges for the ISV CTO in moving to the Cloud
- Leave the audience with some practical ideas to take action straight away
The current IT landscape is pretty cloudy. IT providers are branding whatever product they have that happens to run in a datacentre somewhe as "Cloud" irrespective of whether it's conventional IT but hosted, a (hosted or shared) managed service, or an actual cloud computing implementation of some kind. It feels a bit like the Wild West out there. I'm a great proponent of Geoffrey Moore's books Crossing the Chasm
and Inside the Tornado. If you look at his life cycle graph, Cloud is definitely over the chasm, in to the bowling alley and heading for, or arguably well in to, the mainstream. However, the average buyer trying to find out about It is met with too many mixed messages, unclear definitions and not enough clarity on the real business benefits - just type Cloud Computing in to Google and see what you get! Another concern I have is that a buyer heading to any of the general Cloud trade shows to investigate and learn about the topic might be left with the impression that Cloud Computing is all about Infrastructure - they're generally wall to wall with booths selling hosting, virtualisation and compute capacity in various Cloud forms. Where are the business apps? Well, if you look at Forrester's projections
out to 2015 published last year, their picture tells the story. Steady growth in Infrastructure as a Service, Platform as a Service, and they separated out Business Process as a Service, but the biggest share of the Cloud market is Software as a Service growing from $25.5 billion in 2011 to $159.3 billion in 2020. I found a US oriented slide from Bessemer Analysts
that paints the picture in product logos. Within the SaaS space there are a host of suppliers (pun intended) from large, but mostly small which is why you don't see them taking big expensive stands at generic Cloud Computing shows. They have a more direct route to their market niche through the web, social media and word of mouth.
Have you read Douglas Adams great book Dirk Gently's Holistic Detective Agency?
like Dirk I believ in the "fundamental interconnectedness of all things."
. Moving on to a different fantasy story, in terms of the usual evolution of IT architectures, we're not in Kansas anymore. I've lived through several major disruptions In my 30 plus years in the industry. From mainframes to minicomputers, to the birth of the IBM PC. Networked PCs and open systems (and by that I really mean UNIX) evolving in to client/server and then the first wave of the Internet and web 1.0. This time "now" is special because we have three significant disruptions happening at the same time. The shift to Cloud is happening at the same time as the rise in Mobile, is happening at the same time as the move to Social. Each one of these alone is as big a challenge to conventional IT as those previous shifts to a new computing paradigm. We've never had 3 such shifts happening simultaneously before, and it presents a huge opportunity (and threat!). Getting back to that Wizard of Oz metaphor, the tornado really has taken us to a different World. You only have to look around at people in the street, the classroom, on the way tothe office - all glued to the web and each virtual other through hand held technology. It's happening in Nairobi, Beijing and Lagos just as much as New York, Berlin and London. People are connected through their smart phones and their iPads. Music, TV, movies, news, publishing have all been disrupted, but actually every industry is being affected by the new, flat world of instant communication and the long tail. We're following the Yellow Brick Road of the digital information economy.
With Cloud we can also see two tracks of development happening. One is retrofitting existing IT solutions in to the Cloud. That's fine, but it's still the same old software, just as a service. You get some of the benefits and you at least make a start. The other approach is to chose solutions that brace the Cloud fully. Pure-play Public Cloud solutions that help you change your business model, make new connections with customers and partners, and allow you or the provider to use the underlying data in innovative ways that would never have been possible or practical with on premise solutions. This is where the Cloud can really make a huge difference, add value and create jobs rather than just give incremental cost savings.
So what are the five challenges for a CTO in an ISV thinking about shifting their traditional software products to the Cloud?
Your business model needs to change from end to end.
As a software provider, every aspect of the way you do business needs to change in the shift from selling up front licences to a subscription based model. You'll find your customers in a different way - from classic direct sales to more use of marketing, the web and word of mouth. You'll have to compensate the sales team a different way. You'll service your customers differently - more self help, better documentation, more training online in place of onsite. The projects will work differently - less big bang mega projects, but instead more of them, but smaller - lean trials of a solution with a few users that subsequently get rolled out in stages to the whole customer community, but only after they've been proven to work. Before the customer took the risk with big up front licence fees and hardware costs. Now the provider takes the risk and the emphasis shifts to better customer service and help for the customer to expand their user population. That's why the old guard providers are struggling to make the shift, and why new players like Salesforce, NetSuite, Workday, Huddle, Xero and Twinfield can come in make and inroads. There is no other solution than recognising what needs to be done and changing all of your processes to suit.
Don't make me think!
(Which is the name of the best book on website usability
, authored by Steve Krug way back on 1999.) Your solution needs to have the user experience front and centre. Your users can go and buy a book on Amazon without going on a training course, and they'll expect your software to be just as easy and intuitive. In any case, you can't afford to do the consulting and training you once did with on premise. You need to understand this and design your solution accordingly. Back in 2007 I met Richard Moross of Moo.com
at an OpenCoffee meet up. I asked him what was the most significant investment he made in setting up the company, and he replied "the £10,000 I spent with Flow Interactive
to improve the user experience of the website", (sadly Flow charge a hell of a lot more these days!). It worked, because creating cards with Moo
is a real joy, and you have so much fun you tell your friends. Go and buy Steve's book, put user trials and testing as an integral step at the start of your design process and focus a lot more attention on the actual user experience.
Your development approach needs to change to an agile, lean approach using SCRUM or something similar. Instead of the dinosaur style 18 month new release cycle you need to have smaller, incremental releases, at minimum quarterly but aiming down to weekly. You need a feedback loop and multiple iterations so customer requirements are met regularly. Involve them in the processs and crowd source new product ideas.
Back in the time of client/server I used to work for a company that supported their enterprise accounting software on 26 different platform combinations of operating system, database and hardware platform. It was a nightmare. More time was spent testing technology glitches in an operating environment than on testing the actual, useful functionality we had added to a release. When we moved in to the SaaS world we were in total control of our own environment and testing got SO MUCH easier with only a few different browsers to worry about. We are still in much better shape than that old world, but suddenly everyone is holding the Internet in their hands, and they want access from as many places as possible. Bring Your Own Device is becoming common in many copanies. Smart phones are the norm and their screens are getting larger and more usable. How many iPads will there be in the iroom at your next meeting? We're back to supporting multiple platforms again, but so be it. You have to have mobile on the roadmap and an iPad app on your horizon.
The smart companies are using social networks and social tools to find and engage their customers in direct conversations. The smarter companies are using social tools inside their company to communicate and collaborate and get their work done more effectively. What I used to call web 2.0 for the enterprise and then enterprise 2.0 has now become social business. You need to be encouraging this approach inside your company along wit adding the social dimension to your product itself. Companies like Saleforce with Chatter and acquisition like Solution 6 for social media monitoring are leading the way. Whether you add micro-blogging in to your product yourself, or integrate a third party tool like Tibbr or Yammer to overlay your process flows, it is an essential direction you should be taking.
So those are the 5 challenges, but I'd like to add 2 more thoughts. First, don't develop everything yourself
. In the past a traditional software company developed too many generic tools to handle things like support or eCommerce rather than buying off the shelf. You need to focus on the expertise that makes your company special. You can buy support functionality from Get Satisfaction or Zendesk. You can buy subscription handling from Zuora. There is a host of options you can use to help spend the time of your development team wisely, on the stuff that really matters.
And lastly, don't forget to worry about all aspects of security
, from data protection, to your internal security procedures to backup and business continuity. This will be one of the first and most important questions you get asked by the prospect. Make sure you have a World Class answer
- The Cloud approach means even the smallest of software start ups can piggy back on best in class Cloud platforms and infrastructure.