The false promise of technology

Part 7 of a series on toxic technology.

For all the wide-eyed futurology surrounding the potential of digital technology, it commonly fails to meet expectations. Start-ups fail considerably more than they succeed. Digital ‘transformation’ initiatives commonly overrun and under deliver. Many products and features fail the ultimate test – users just don’t find them useful. This isn’t surprising – building with digital technology is experimentation, trying something new. 

In some respects, the optimism of digital is a virtue. It creates the conditions for the boldest reforms, and the biggest ideas. But where this optimism becomes a problem is when it comes with a disregard of pragmatic pessimism. Optimism is toxic when it leads to avoiding questions of feasibility, viability and the conditions for success.

Digital technology is a particularly dangerous place for this blinkered optimism. It’s an unpredictable medium, in an industry bursting with hype. If you’re willing to believe it, the worlds’ hardest problems can be solved with AI, or blockchains, or internets-of-things. If you do believe it, you’re almost certainly going to be disappointed.

In markets

Companies can tap into broader market optimism for the potential of hyped technologies. This is typically by expressing an intent to use it, rather than showing proven success. In November 2019 there were a flurry of articles reporting that HSBC would use blockchain for a ‘custody platform’ by March 2020. 

“HSBC aims to shift $20 billion worth of assets to a new blockchain-based custody platform by March”

November 2019 Reuters news article

There is no corresponding press coverage in March 2020 to indicate this really happened. But the announcement had already served its purpose to generate positive market sentiment for HSBC. It’s unlikely blockchain was a necessary technology to solve the problem, but its use turns a bland corporate IT initiative into a story about supposed innovation.

In governments

Governments frequently make assertions of future delivery. This helps cement strategy across huge bureaucracies, and encourages public engagement. It is also low risk to reputation because governments are pretty effective at managing the message around its technology delivery failure.

A current example of a government promise on technology comes from the UK government’s National Data Strategy:

“We will drive data discoverability across government through developing an Integrated Data Platform for government, which will be a safe, secure and trusted infrastructure for government’s own data.”

UK government National Data Strategy, September 2020

This announcement sets false expectations:

  • That much of the UK governments’ data can be put onto a singular ‘platform’.
  • That a named thing needs to exist called the “Integrated Data Platform”.
  • That such a thing will come into existence reasonably soon (perhaps within the next spending cycle in the UK government).
  • And, at some point, funding will be allocated to make this all happen.

The language may appear subtle, but published strategies start to shape the funding pots, and the £multi-million programmes that emerge. The complexity needed to solve data in government is immense, but a ‘single platform’ is appealingly simple to fund and easy to announce. The wider strategy does explore the complexity of data in government, but these more concrete, announce-able things will have a more enduring impact.

In politics

As technology plays a greater role in society, we’re likely to see ever more technology-based promises from parties seeking votes. This from the 2019 UK Conservative Manifesto is an example:

“We will use new air traffic control technology to cut the time aircraft spend waiting to land”

2019 Conservative Manifesto

This promise of a real-world outcome, achieved by introducing new technology, is hugely appealing to many voters. But how this will be done, and the uncertainty involved, is not forthcoming. Which technology would be used, and why? When would aircraft waiting times reduce? And to what cost to taxpayers? Instead of a strategy, we’re asked to vote for a promise that cannot be made, and the belief that a government will be competent enough to deliver on it. Whilst this pattern of trust in political promises isn’t new, the unpredictability of digital technology will increasingly expose the gap between intent and delivery.

Expectation vs reality

It’s hard to trust assertions about when and how digital technology can deliver outcomes. The internal complexity and interdependencies of digital technology mean even the most experienced professionals rarely get their estimates right. Most software developers can tell you the time they spent days solving a problem they thought would take a few minutes.

Technology initiatives don’t overrun or under-deliver – instead, people have irrational expectations of certainty. CEOs, Ministers and managers crave this certainty to build a reputation as a person-of-their-word. Customers, workforces and electorates crave this certainty because it reduces anxiety around something they need, or want.

This mismatch of expectation and reality isn’t just a disappointment or a financial write-off. Tactics and fire-fighting must make up for failed plans – reducing the quality of digital technology. These are the ideal conditions for toxic technology to emerge – poorly designed, insecure and unstable technology which is harmful to the organisation and its users. Sometimes organisations work this way for years, reinforcing the culture of false promises. Inevitably it’ll backfire spectacularly such as the glitchy dystopia of Cyberpunk 2077 or the IT meltdown at TSB. If organisations don’t learn from consistent failure to meet expectations, then they normalise the accumulation of toxic technology. 

The accumulation of broken promises

The tech industry narrative of success is dominated by survivor bias of the few commercial giants – each an example of when bold predictions came to pass (at least in their origin stories). Yet the majority of organisations are filled with toxic legacy and broken promises. The historic strategies of these organisations did not anticipate today’s glacial pace of change, system workarounds and disruptive incidents.

Inside established sectors, like government and traditional banking, these false promises repeat cycle after cycle. Somehow, the next Big Technology Transformation will solve all the problems, despite being funded, governed and promised in the same way as previous attempts.

The powerful – CEOs, government ministers, or those with the ability to gain media or community attention – have a unique role in how promises are made. There is pressure on them to make promises – relieving anxiety that something needed will be delivered, or creating a buzz around a future product. At worst, this is a tool of command and control management – creating unrealistic expectations in public to motivate teams in private. But even the most throw-away announcements can be exploited by command and control cultures. Leaders must take care with announcing new investments, plans or strategies – they communicate intent, but they also set expectations far beyond the leader’s control – and these expectations are not always reasonable. 

Making better promises

We still need to make promises, and to do so with optimism. It’s essential to positive, forward-looking organisational cultures. Optimism is what encourages teams to take risks and try something new. Promises are how trust is built. 

But we all need to make better promises when it comes to digital technology.

Leaders should balance their optimism with pragmatic pessimism. They should ensure this is woven into the culture, practice and process they encourage. They must make sure they don’t consistently over-promise and under deliver, even when the culture might reward it. They must wary of where the burden of expectation will fall, as often it will be with others.

Funders and those that govern should recognise that upfront large-scale approval for funding and headcount encourages false promises. They should work to reform accounting and governance based on greater trust within organisations, making it more incremental and iterative. They should listen to digital leaders and delivery teams as peers, putting aside their role in ‘holding to account’ to collaborate on strategy and reform.

Technologists should use their experience to counter naive optimism. They must remember what caught them by surprise – when the simple became complicated, and the complicated became simple. They must be aware of their own bias in setting expectations of delivery, particularly. when estimates depend upon their expertise. Technologists must be skeptical of hype, even when technology is exciting and new.

Designers should value and understand their medium of delivery as this is where viability and feasibility becomes clearer. Like an architect appreciates wood, steel and construction techniques, a designer of digital technology must appreciate software, data and the craft of technologists. This can be done best in a multidisciplinary team – finding the right experts for the right design challenges.

Product managers should embrace risk as a peer to value as a core part of their discipline. This helps ground bold ideas in achievable reality – it ensures product managers have a more complete set of information with which to make decisions, and make promises.

Delivery professionals should embrace uncertainty. They shouldn’t accept the false certainty of specific and singular estimates (e.g. “complete by July next year”) – ensuring they consider a range of potential outcomes from the best to the worst cases. They should be skeptical of estimates and conscious of bias. Delivery professionals should take a lead in exposing and challenging unreasonable expectations from leaders. Delivery professionals should hold to account those who consistently set poor expectations or provide inaccurate estimates.

Sign up for the next instalment

The next part of this series will be published soon. Sign up for email updates below.

Success! You're on the list.

Culture eats technology for breakfast

Part 5 of a series on toxic technology.

Culture in organisations is what people value, and how they behave. Some organisational cultures are undesirable, and will cause and embed problems until the culture changes. Some of these problems impact technology – culture can reduce its quality, and lead to it becoming toxic. Poor cultures can encourage neglect, poor decision making, and prevent people from caring about what is important.

These cultures are unfortunately very common, and even become codified in policy, process and practice. The emergence of undesirable cultures is perhaps inevitable – people’s individual motivations, preferences and biases lead towards them. But they can be changed once observed. Through leadership and the momentum of shared purpose, people can change their behaviour, and start to value what matters.

Command and control

A command and control culture means teams aren’t trusted. It means the more senior someone is, the more they presume to know what’s best for the organisation. Teams are commanded to deliver outputs rather than outcomes.

Toxic technology will emerge naturally in command-and-control cultures, because feedback loops don’t work. When new information is discovered, and it conflicts with the highest-paid-person’s opinion, then it’s disregarded. Often, the most senior people never even see new information – throughout a command and control culture, people are incentivised to follow the plan, not do what’s best for the organisation. This prevents new information flowing up the hierarchy, where it can influence strategy.

The problem is that command and control is pervasive. Long-term detailed business plans, target operating models and annual budgets all default to command and control – and they exist almost everywhere. Managers must preserve their reputations by keeping to promises and commitments. If growing technology risk is exposed as new, and contradictory information, it has to contend with these forces.

Command and control practitioners might point to under-performing or self-preserving teams to legitimise their methods. But if teams have psychological safely, they will be more open about their shortcomings in capability or experience – they will ask for the help they need. Command and control prevents this honest dialogue between managers and the managed.

Sometimes organisations must make major command and control decisions – to become competitive or to secure survival. Even in these circumstances, trusting in people and teams is a better foundation for the hardest change – such as redundancy. Reverting to pure command and control in these times could poison the culture – meaning that what survives is doomed to fail.

There are also circumstances where rigid standards-based controls can be used to exert influence at scale. This blunt instrument can be useful in making a big cultural shift, but its effectiveness is time-limited. If compliance rises high, then governance simply becomes friction. If compliance remains low, it could be unachievable – leading to the standard being subverted or ignored. Consistent failure to meet standards is often a signal of teams burdened with technical debt. Being punitive to the non-compliance motivates teams to hide, and therefore continue to accumulate, toxic technology. Crucially however, even if a standard is temporarily rigidly enforced, it should be done with trust-by-default – assuming everyone is doing their best to achieve it, but they may need help.

Everything is awesome

In some cultures it can seem that everything is awesome (when it is not). When communication is dominated by an organisations’ success stories, failure becomes a dirty secret. Bad news is discouraged or even suppressed. These cultures can be very fun for the majority, but they are usually time-limited. Once the disparity between communicated success and observed failure becomes stark, more people will become cynical and disengaged. A counter-narrative will emerge amongst the disaffected.

Most organisations should tilt the narrative towards the positive – it helps make for a happy, optimistic working environment. This is particularly important for external communication – it helps with hiring, and pride for the current workforce. But, when everything is supposedly awesome, it can be challenging to raise awareness of risk. Toxic technology can be allowed to grow because talking about it doesn’t fit the organisation’s positive narrative. If organisations experience the consequences of toxic technology – such as a cyber attack or system outage – these cultures don’t handle crises well. They are likely to be under-prepared, and shocked at how this could happen.

An extreme version of this culture exists where routine failure is celebrated as success. Large-scale outsourced public sector IT since the 1990s has largely been a series of calamitous failures. Yet, many embedded in the industry celebrate the same activity as unbridled success. The means the business case structures, delivery methodologies, organisation structures and funding models have remained largely unchanged despite routine, spectacular failure. This pattern has been perhaps one of the largest sustained contributions to the global mountain of toxic technology.

Shiny things

A powerful distraction to tackling legacy is human nature — we are attracted to the new. The technology industry is rife with the problem of under-valuing the old, and over-valuing the new.

Organisations focus on delivering new technology, often to the detriment of improving existing technology. We build new products. We digitise the analogue. But organisations perhaps don’t realise that software ages like fish? – if we look away for too long, it begins to rot. Legacy accumulates quickly by following the fashions — Javascript frameworks are a good example of this effect, where popularity rises for passing moments in the long history of web technology.

Not invented here syndrome”, is the tendency for technologists to craft a solution to a problem that has already been solved. This can be caused by the challenges of discovering existing solutions to problems, and how much to trust them. But more significantly, creators of technology have the desire to create. As a former software engineer, I know that creating something completely new is emotionally rewarding. Particularly when compared to configuration and composition of existing technology.

Technology that has been around for decades, such as SQL databases, web frameworks, and web servers can be harnesses to solve a vast range of problems. So many problems can be solved by “putting strings into databases (and taking them back out again)”. Technologists must be very user-focused and goal-focused to choose the boring technology that works best. Unfortunately, early career progression as a technologist can depend upon exposure to a variety of technologies – incentivising trying out new things where possible.

The pursuit of shiny things affects entire organisations when they prioritise reputation over sustainability. To some degree this is legitimate – reputation drives sales, inflates the share price and gets investment. An inflated reputation also helps gain momentum when transforming organisations from the inside. But reputation-over-sustainability doesn’t last – at some point there are consequences for neglect.

The love of shiny things runs deep. Design and systems thinking connote a sense of freedom often only possible by starting anew. Organisational, political, and professional leaders are valued and rewarded, for their output – measured through the creation of announceable things – not the outcomes they influence through teamwork. It takes a lot to resist these forces, and spend some time polishing up what used to be shiny.

Too many heroes

Organisations love a hero. They celebrate the achievements of individuals or teams who achieve impressive feats, against the odds. They might deliver something new in a heroically a short timescale. They might avert catastrophe by working all hours to resolve a problem. A culture of everyday heroics has innocuous beginnings, with people receiving praise for going beyond the usual expectations of their role. But when entrenched, it makes the organisation fail to see systemic failure, or appropriately recognise systemic success. Rapid delivery might be the result of years of bold investment in platforms. Heroic ‘all-nighters’ by the operations teams are often the result of systemic neglect of technology.  

Organisations shouldn’t stop celebrating these ‘heroes’. But they must find balance. They should strive to celebrate systemic success or failure. This is hard, because systemic change often doesn’t have an event, or moment, with which to associate the fanfare. They should celebrate the teams, past and present, who made the success, or averted failure possible – not just the individuals who made the most visible impact. Celebration creates incentives, and an imbalance towards the short-term, highly visible success stories. Incentives are needed to encourage longer term impact, where individual credit is less likely.

Improve culture by observing it first

Identifying and cataloguing cultures is not scientific – it’s just too hard to characterise and categorise collective values and behaviours . But culture still has a powerful effect. What’s important is to find ways to spot undesirable cultures, talk about them, and improve them – before they do too much damage to the organisation, and its technology.

Continue reading part 6.

Sign up for future instalments

Sign up for email updates below.

Success! You're on the list.