I recently spoke at the Sydney CDO Summit, on ‘Change, not Change Management’. That title just about sums up how large organisations should be thinking about digital transformation. It’s not a question of gradual, incremental transition, but rather decisive, concerted and sustained action.
Brownfields are being judged by the standards of the Internet age. Many are failing that test. Brownfields are operating in a Darwinian marketplace: If they don’t change how they do business, competitors will be ready to take their place.
I want to share some of the key lessons of my talk about digitally transforming large organisations: why it’s difficult, and what will — and won’t — work.
The benefits of digital
When organisations use the word ‘digital,’ they’re driving towards a specific idea that’s distinct from ‘online’ or ‘e’.
They’re trying to tell us something about the way they do business:
- They’re offering Internet-quality services. Users with Amazon and Netflix at home won’t put up with needing a solicitor to tell them which button to press on a website when dealing with their mortgage. Too many brownfields think users want to ‘engage’ with them online. Digital organisations realise that no one wants to spend time engaging with their banker or insurer. They just want to get things done.
- They’re operating at Internet speed. They know that six-month release cycles are wasteful, increase risk and make it impossible to take what they have learnt about how customers use their product and apply those lessons quickly. They want to deploy hundreds or thousands of times a day, and respond to what they learn about users quickly, just like Amazon and Google do.
- They know Cloud has changed the economics of IT. Working in IT fifteen years ago, licences and hardware were an extraordinary source of CAPEX. That involved huge up-front risk, which forced organisations to have heavy waterfall-style governance on all tech projects. Cloud has changed the material circumstances of all that. It means organisations can now experiment with MVPs, to quickly learn what users want from their product, and deliver it at Internet speed, before the competition does.
Blockers to transformation – and the iron law of bureaucracy
That benefits of digital are easy to understand, sometimes tricks us into thinking that getting to digital must be easy, too: that it’s just about building a few MVPs, or bringing in some UX people, or making websites a bit cleaner.
The bureaucracies inside large organisations haven’t survived by welcoming change. In fact, an unspoken goal is always maintaining themselves, as-is, with their hierarchies, procedures and cultures intact. They rely on an artefact we discovered when fixing government services in the UK, ‘The Triangle of Despair’: the three big barriers to brownfield transformation that reinforce resistance to change.
1. Inappropriate procurement
The first element is inappropriate procurement policy and practices. Too many brownfields expect a vendor to take care of everything for them in a large tender. They put up with lengthy negotiating processes, and limit their options to what a small group of companies is willing to deliver. That fragments the user journey between the different vendors which set up each segment.
Services aren’t just split between an organisation’s internal siloes, but also between the multiple vendors tasked with creating an element of the service. Worse, it forces organisations to fix a project’s scope at exactly the time when it has the least information: when it initially puts forward a tender offer, before building a prototype and seeing what users want from a product. That makes it difficult to change course without dealing with giant fees for changes in scope.
That’s why brownfields need to reform the way they do procurement. In DTO, that meant building the Digital Marketplace, which let government agencies negotiate a contract in days, not months, and broadened the choice of suppliers for digital projects. In the UK it meant building the G-Cloud.
2. Excessive governance
The second element is the governance brownfields use for IT projects. Laying on heavy, waterfall-style governance on projects for which it’s not suited slows down delivery, and spreads out responsibility across endless boards and subcommittees.
That means no one is directly held to account for the success or failure of the project. It leads to a focus on box-checking and compliance, not control, as a large number of people with a limited understanding of a project are asked to assess it.
There’s no way a board can understand the status and intent of a project from a hundred-page risk document sent immediately before a governance meeting – but that’s exactly what many brownfields expect of them.
In Agile methodology, a small team bears full responsibility for delivery and the SRO can easily browse the backlog at will or attend delivery standups.
Even worse, some brownfields try to use Agile and Lean methodologies inside their delivery teams, but overlay them with project management toolsets suited for other kinds of problems. That means there’s governance at cross-purposes, because Agile and Lean have their own in-built governance.
The result is a lack of clarity about which process is really assuring a team’s success – or worse, making it impossible for a team to properly apply the method that’s right for the problem.
Governance reform means fully adopting Agile and Lean on novel, outward-facing problems. The benefits are obvious. When I worked in the UK, I watched long-serving senior civil servants who were devotees of PRINCE2 experiment with Agile; it gave them so much more visibility and control that some of them now make YouTube videos evangelising it.
That doesn’t mean Agile should be adopted everywhere. One size does not fit all problems.
For instance, Agile:
- is not the way to build a nuclear power plant, or a product processing system in a bank, or a common digital platform.
- is a risk mitigation methodology for dealing with novel, unknown problems. Applying it to situations where you need absolute stability or where you already clearly know the specification is wasteful and discredits the digital team trying to transform the business.
- is the best methodology for new and novel problems, and especially for end-user facing software, where we frequently don’t know what the best solution will be until we actually test it with users.
3. Broken IT
The final element is broken IT. That means ancient systems, using outdated proprietary technologies running a back-office with far too much manual processing.
Broken IT can also be IT that’s engulfed by layers of contracts and systems integrators. When I was at Ministry of Justice, we delivered a service that made it easier for people to book visits with prisons quickly and simply online. To do that, we had to access a prison database that was wrapped in three or four different layers of contracts. If you need to negotiate with vendors to get access to your own resources, then your IT management is broken.
If brownfields want great digital services, they can’t ignore their own broken IT. Instead, as I’ve said before, they should be giving a single Chief Digital and Information Officer overall responsibility for both outward-facing services, and internal infrastructure.
This will enable a common team, operating off of a single strategy, to make the changes to the back-office required to deliver great services to users.
From the perspective of productivity, too many brownfields have employees who use state-of-the art iPhones at home for recreation, while waiting half an hour for their Windows PC to boot up at work.
A deskilled workforce
Finally, each of those three elements feeds off a deskilled workforce. Low internal digital skills at all levels encourage a company to rely on vendors and apply heavy governance to manage risk because they stop trusting their own capability to deliver, while ignoring the problems of their broken IT. Brownfields should be upskilling their workforce so that they have the talent to build great digital services – themselves.
Panaceas — and why they won’t work
Brownfields can’t make the switch to digital if they don’t address these problems. You hear a lot about the magical one-step event that’ll lead an organisation to transform itself. These range from:
- Innovation studios that give a very small number of employees a great time learning about Agile, before sending them back into a change-resistant business with excessive governance and broken IT. This is a recipe for frustrated, demoralised employees who’ll take what they’ve learnt and head elsewhere.
- Hackathons and other kabuki that generate some good ideas, but which are quickly forgotten when inappropriate procurement and deskilling makes them impossible to deliver.
- Lipstick solutions that clean up a digital offering with a neat website, and leave a highly manual back-office and ancient legacy systems in place.
- Incrementalism, or “evolution not revolution” that aims to gradually transition the organisation to digital through “change management” – which misunderstands that the business fundamentals need to be changed, so that the digital team is not sucked right back into the bureaucracy around them.
- Rip ‘n replace strategies that aim for total system change, but take on huge risk and almost always fall short. The only place I can think of where anything like this this ever has worked is Estonia, which aggressively moved towards e-government in the early 2000s. They did it under an absolute burning platform: the perceived threat of Russian aggression. Most of us don’t deal with that, but we will have our own challenges.
- Two-speed, or bi-modal IT that segregates the back office from the front office. A model that forgets that services rely on an ensemble of people, processes, systems, policies, operations and facilities, across both the front- and back-offices. Bi-modal IT creates a fast lane for digital teams working on front-office lipstick, while leaving the back-office untouched or vice versa. This is the opposite of transformation. It is more like conservation. Worse, it creates a cultural divide between the new digital team and a ‘legacy’ team without a real future in the organisation, giving rise to unnecessary political problems and internal conflicts.
- Digital cultism – and ‘Agile Everywhere!’ that suggests all problems should be solved in the same Agile way. Not all problems are experimental ones which are right for Agile. Pretending otherwise just creates a divide between digital teams and the rest of the organisation.
Doing it right
No panacea can make transformation pain-free. But it is possible for organisations to transform themselves. Doing so takes three steps.
1. Transformation requires a clear vision of digital
Organisations need to think about the full journey users take when they access their services. That means:
- understanding what your user is actually trying to get done and making it as simple as possible to do it
- providing a consistent user journey across service entry points
- automating and optimising back-office processes that support them so the transaction can be processed without manual intervention – or “straight through”
- keeping users updated in real time on their status
It also means providing scalable, disposable digital platforms for commodity services that support your product or business, like notification and payment services, from third parties, to make it easier to quickly scale up and replace them with better offerings as they appear.
Ownership and implementation of this vision needs to come from inside, because you are now becoming a software company and you need to insource capability and take responsibility.
Uber, Google, Netflix don’t outsource their product development to potential competitors.
2. Transformation requires radical internal change
Transformation also means undoing the hold that the Triangle of Despair has over your organisation. This means:
- building a procurement process that operates in days, not months
- adopting Agile and Lean methodologies when dealing with novel, user-centred problems
- putting in appropriate governance, suited to the problem at hand and the economics of cloud, not one-size-fits all governance from the age of ITIL and huge CAPEX
- knowing when to build, when to buy, and when to use open source.
- radically upskilling your organisation with the digital and commercial talent to make those decisions and to compete in the global market opened up by the Internet.
- abolishing the divide between digital and IT that stops us from designing services taking into account the whole user journey.
A service isn’t just the user interface; it relies on the people, processes and systems that back it up. Organisations need a single organisation with consolidated oversight for delivering high-quality digital services, rather than splitting up responsibility between a CDO looking after the front-end and a CTO or CIO with responsibility for the back-office.
Abolishing the digital/IT divide is also useful from a human resources and product lifecycle perspective. It means the CDIO can match people’s aptitudes to the right methodologies for the right problems.
I like to think about this in terms of the three categories of digital talent Simon Wardley talks about:
- Pioneers who love uncertainty and exploring new things can be kept away from platforms, but used for coming up with great user-facing services
- Settlers who take MVPs developed by the pioneers, develop them and turn them into mature products
- Town Planners who take advantage of economies of scale to make services into efficient and highly commoditised platforms, at the end of a product’s lifecycle
That creates a conveyor belt to move products from being unknown and experimental, through to being industrial strength, and then to becoming platforms. This process relies on the whole organisation operating at digital speed.
By contrast, a bi-modal IT system which only upgrades some parts of an organisation’s digital capability leaves new products with no way of being industrialised without another massive reorganisation effort, sitting alongside expensive legacy IT in a fossilised state.
3. Transformation requires political will
Finally, transformation requires political will. Transforming by consensus, or by an occasional Hackathon, can’t work. It means looking to make fundamental changes to the structure and processes of your organisation. If you don’t have the political will to make those changes, you’re wasting your resources – and everyone’s time – talking about it.
Digital success requires undoing a Triangle of Despair that creates organisational paralysis and that is the result of years of learnt helplessness. It’s not enough to use a panacea of a few innovation studios or neat websites. Instead, brownfields need to have a clear vision of of how they will compete and thrive in the Internet economy.
Digital success is underpinned by the culture, skills and and the political will to implement it.
This article was first published on LinkedIn.
The full transcript of Paul Shetler’s address can be found on his website.