Perspectives
Intro: Growth in an uncertain time
As business leaders stand on the cusp of a new year, many are bracing for disruptive change and ongoing economic and geopolitical uncertainty. A recent survey by the US-based Conference Board found CEO confidence declined in late 2024, on the back of a generally weaker outlook. There is no shortage of potential causes for concern, with another poll of global CEOs by KPMG identifying supply chain, operational and cybersecurity issues as the top perceived threats to business growth.
Top threats to growth, as seen by CEOs: 2015-2024
Yet, leaders remain, on balance, hopeful – about the economy, and the potential of emerging technologies to drive revenues and efficiency. Most are also sticking to, and investing in growth plans, with an eye to positioning themselves to seize future opportunities.
“There's always going to be some part of the economy or globe seeing demand,” points out Mike Mason, Chief AI Officer, Thoughtworks. “If you don’t grow, you’re shrinking. Stagnation is death, and no business would want to admit to that. So leaders need to be thinking about and acting on growth most of the time.”
“If you don’t grow, you’re shrinking. Stagnation is death, and no business would want to admit to that. So leaders need to be thinking about and acting on growth most of the time.”
Mike Mason
Chief AI Officer, Thoughtworks
What has changed is that growth is no longer necessarily something to be pursued at all costs. “Organizations have always sought to grow and provide shareholder returns,” says Chris Murphy, Chief Client and Revenue Officer, Thoughtworks. “But they are doing so in a much more focused and intentional way now than three or four years ago.”
This trend is changing approaches to investment in technology, which is increasingly viewed as vital to expansion, but is also being assessed more rigorously.
“We're coming off a post-pandemic technology surge,” Murphy notes. “There was a rush to reorganize business towards a more virtual model, which had the effect of bringing forward a lot of technology spend. But with the costs of capital rising, many of the business cases that previously made a lot of sense now don’t. Companies have had to rapidly reorganize around the question: What does return on investment look like in a new world where economic uncertainty is higher and money is suddenly more expensive?”
“People are still cautious with where they're spending their money, and it's hard to get businesses to commit to longer-term projects because of that,” says Andy Nolan, Director of Emerging Technologies, Thoughtworks. “They want to see results early and to prove value quickly.”
In this edition of Perspectives, Thoughtworks experts set out the blueprint for a growth-supportive tech strategy that keeps spending and ambitions in check while delivering value where it counts, enabling the business to strike what Murphy frames as a delicate balance.
“Almost no organizations have the luxury of ignoring cost and complexity,” he says. “But at the same time, not one has the luxury of stepping back, of not focusing on the customer, or of not driving new growth opportunities.”
i. Level-setting: Integrating business and tech strategy
The stakes around technology investment have risen amid an increased cost focus, and with the rise of generative AI (GenAI) inflating budgets and expectations.
“I've been working on machine learning and AI for 15 years, and in the last two years, I've spoken to more boards and C-suite executives than in my entire career,” says Nolan. “The conversation is often a mandate from the board to work out what the organization is doing around AI, and how it’s being adopted to achieve growth or cost optimization outcomes. It’s exciting, but also a little bit scary, because there's a disconnect between reality and where the hype is.”
One study by Google found almost three-quarters of enterprises are realizing returns on GenAI investment. Thoughtworks experts have also helped organizations extract significant efficiency and productivity gains from GenAI projects, typically by applying the technology to augment roles rather than replace people altogether.
“Growth and efficiency driven by AI is on everybody’s agenda - not necessarily cutting staff, but reducing cycle times and getting better outputs,” says Mason. “I was recently talking to the CIO of a food and beverage company using AI for marketing. The last thing he was going to do was lay off anybody from his marketing team. But he's got a list of 500 wines that he wants to market, and the team can only do the top 100; if they’ve got AI support, they can get to 200 or 300.”
Despite such examples, a recent survey by Gartner showed proving the value from AI projects remains the primary barrier to greater adoption.
“Some of the numbers that we're seeing thrown around give businesses a false sense of AI’s potential benefits,” Nolan says. “We're in a window of opportunity now where organizations need to quickly demonstrate tangible value. In the next six to 12 months, we’ll probably start seeing the tide turn a bit on AI adoption. Funding may dry up for some of these initiatives, or attention will pivot to other areas of the organization.”
Extracting value from AI depends above all on ensuring AI and tech investments are linked directly to business goals. “Technology strategy needs to be at the heart of your business strategy, and they need to be co-created,” notes Mason. “Every organization is trying to bring those things together to get better alignment, and also better insight from technology as to how it can be applied to their business.”
In the most effective organizations, this alignment is reflected in leadership. “Technology essentially requires two functions,” Murphy explains. “The first is strategy - linking the technology landscape with the business landscape, so that you're building the right digital products and assets to achieve your organizational goals. The second is execution; that is, having the capability and the knowledge to deliver those high-quality digital products and assets to market quickly. In businesses that achieve both, you often see technology leadership on the board, reflecting the importance of understanding the changing technology landscape.”
When clients ask for help applying AI to their business, Thoughtworks “tries to understand the business strategy first and foremost, and based on that, identify opportunities to leverage AI to accelerate that strategy,” Nolan explains. ”Looking at the organization with an AI lens can be helpful, but it also can be dangerous, because you might chase after things that are distractions. If you can tie your AI project back to strategy, it will get the visibility, the support from leadership and the funding that you need to make it happen.”
AI initiatives tend to fare poorly when the integration with business strategy is absent, and “everyone’s running off and doing their own thing,” as Nolan puts it. Such projects may not only fail to support growth; they can actually cause more friction.
“AI on its own is not necessarily going to solve an organization’s problems – but it will push on the sore spots,” says Mason. “If you don't have scalable platforms or clean data that's able to move between systems, AI is unlikely to fix that for you. You can get teams across a company all spinning up small AI experiments, pulling in different directions, not sharing learnings, maybe even spending a lot of money through duplicated efforts. That’s a technology governance problem, made worse by the excitement around AI.”
ii. Cultivating data as an asset, and modernizing legacy infrastructure
Leveraging AI for growth opportunities may highlight weaknesses as often as it enables transformation. And a major weak point for many enterprises is data.
“It’s no secret that most organizations have faced challenges in capturing, organizing and governing data,” says Murphy. “But all of a sudden, that challenge has been thrust into the limelight as a foundational issue that needs to be solved quickly to take advantage of rapidly progressing opportunities in the AI space.”
Data has become even more critical with the rise of AI because it serves as a differentiator, Nolan points out.
“There are plenty of off-the-shelf tools that organizations can simply pay to start leveraging AI and get some type of productivity boost,” he explains. “But it's not really a competitive advantage if everyone else has access to the same solutions. Organizations that have already invested in their tech modernization, and their data infrastructure, are in a better position to adopt AI than those that haven't. The reality is any organization that wants to build a defensible moat around itself needs to leverage its own data.”
"The reality is any organization that wants to build a defensible moat around itself needs to leverage its own data.”
Andy Nolan
Director of Emerging Technologies, Thoughtworks
This is feeding a new focus on data strategy and governance – or “understanding how you can ensure the integrity and the origins of data, right through to modern mindsets around treating data as a first-class product that can be enabled as an accelerator for a company's progress,” as Murphy puts it. “All that supports the rise of data platforms as a foundational organizational asset.”
“There's a renewed question about who should be in charge of data, whether it’s a chief data or chief AI officer,” Mason notes. “It’s definitely not one size fits all; the important thing is to have clarity around who is going to be driving data initiatives.”
Nolan has witnessed a welcome rise in businesses setting up data and AI governance groups to ensure new tools are adopted “in a sensible way, so that they don't have issues with ethics and bias, or front page newspaper articles written about them.”
To enhance both access and transparency, Thoughtworks advocates a decentralized approach to data “that maps to the reality of data within a business, which is that it is owned by different parts of the organization, with different kinds of expertise,” says Mason. “The democratization of data access, thinking about data as a product with a set of consumers, is increasingly a critical way to succeed.”
However, achieving this requires a similarly integrated and decentralized approach to infrastructure, which for businesses running on aging legacy systems is easier said than done.
“Technology is a key enabler for most organizations in growing and achieving their business objectives,” says Murphy. “But it can also be a key bottleneck. Decisions over decades have added cumulative accidental complexity. Adding new functionality and bringing new products to market that are critical to achieving growth can be a time-consuming process because you’ve got to integrate with and remain reliant upon this miasma of legacy applications.”
While turning technology into an accelerator isn’t an overnight process, Murphy notes Thoughtworks has seen many clients achieve success through a ‘thin slice’ approach. This entails “building out slices of new functionality that realize value for the organization, that show return on investment while also modernizing key components of the legacy estate,” he explains. “That's particularly relevant given the normalization of the cost of capital.”
Emerging tools like Thoughtworks’ CodeConcise allow organizations to harness AI to assist with a modernization process that has become both urgent and continuous.
“We’re working with a number of customers who have large legacy code bases, combining GenAI with more traditional analysis techniques in order to build knowledge graphs that represent and enhance understanding of the legacy system,” Murphy explains.
iii: Growth pains: Facing talent limitations
Beyond infrastructure, preparing teams for a technology-driven growth run can be another challenge. Many companies are still contending with tech skills shortages, sometimes with serious consequences for growth. Research from IDC has found nearly two-thirds of North American IT leaders feel a lack of skills has resulted in their organizations missing revenue growth objectives.
At the same time, taking advantage of new technology doesn’t always entail employing an army of specialists.
”The barrier to entry is much lower than it once was,” says Nolan. “A lot of tools have matured to the point where a normal full-stack engineer can develop a GenAI solution without really having deep AI expertise. A slight stretch of existing roles to add AI elements is therefore a cost effective way of starting to leverage the technology.”
This means that in hiring, the focus should be on talent and aptitude to learn rather than specific skills, notes Murphy.
“We’re seeing a lot of organizations chasing unicorns, competing for a very small and very expensive group of people who have core skills in data, infrastructure or architecture, but forgetting that technology is always in a state of flux, and that soon those skill sets will be dated like any other,” he says. “You really need to be building your technology organization so it’s filled with people who are curious and have a growth mindset, to be able to adapt and respond as the landscape changes.”
“You really need to be building your technology organization so it’s filled with people who are curious and have a growth mindset, to be able to adapt and respond as the landscape changes.”
Chris Murphy
Chief Client and Revenue Officer, Thoughtworks
One of the most important, yet rarest, skills Thoughtworks experts see is the ability to position and explain technology in a way that makes clear to the business how it can support growth strategies.
“There's a big distance between seeing an ad for a solution and actually getting it to work within your organization, so that actual business people who understand your business process get value out of the tool,” says Mason.
Having technologists who can serve as a bridge with the business is especially important in what Nolan cites as the toughest aspect of talent development – uplifting skills at the C-suite and board levels.
“Not having someone in the room that really gets the technology and how it works can be dangerous, because people buy into the hype and problems start to bubble up,” he explains. “It doesn’t have to be a technologist, but you need someone with an intuitive understanding to help set up a sustainable AI strategy, explain the importance of data governance and ethics and bias, and get things on the right track.”
Setting the tone from the top can encourage the rest of the organization to learn about and experiment with AI in a responsible and safe way.
“That means actually providing AI tools, not walling yourself off from ChatGPT and insisting no one’s going to use it,” Mason points out. “It’s better to provide approved tooling where you have an idea of what’s happening with data, and there are guarantees that if someone uploads a company PowerPoint, it's not going to be used to train a future version of an LLM.”
iv. Controlling costs and tech debt
In addition to rethinking approaches to talent, companies are subjecting technology spending to more scrutiny.
As Murphy points out, many firms have been disappointed over the past decade when cloud migrations that were intended to lower costs failed to achieve that result. This is contributing to caution towards investments in AI – caution that Mason believes is, to an extent, warranted.
“$30 a month for an AI tool might not sound like much, but when you're at 10,000 employees, you can do the math,” he says. “Depending on someone’s role, it can be difficult to quantify productivity gains. At times you’re relying on people telling you how they see AI tools saving them time and helping them create better outputs.”
Much like cloud, the costs of AI investments can quickly outweigh any efficiency benefits if implementations are not carefully planned or target the wrong things.
“There's only a certain amount of process optimization you can do with AI before you hit a floor and it’ll just cost you more money to have a human to review the output,” says Nolan. “We're nowhere near replacing humans with AI systems at the moment, and I would actively advise businesses not to do that.”
“We're nowhere near replacing humans with AI systems at the moment, and I would actively advise businesses not to do that.”
Andy Nolan
Director of Emerging Technologies, Thoughtworks
Seeing AI as a full-time equivalent (FTE) or cost savings opportunity is “looking at it through the wrong lens,” Nolan adds. “The creativity brought with new advances in AI to achieve revenue growth by developing new products and new business models is much more exciting.”
One example is a Thoughtworks client that commissioned the building of an internal chatbot to help steer users through company documentation. “That sparked a sense of innovation that everyone could participate in,” says Nolan. “Employees could develop their own prompts to optimize their jobs and improve things they cared about. That sort of uplift might not be a business benefit, but it will help with retention and the employee value proposition.”
The risks of investing insufficiently in technology also can’t be ignored. When technology infrastructure fails to keep pace with business growth, “you end up creating tech debt that you have to pay back later,” says Nolan. And at some point, tech debt will accumulate to the point that growth slows or halts.
Telltale signs of trouble include tech investments producing diminishing or no visible returns; the pace of improvements or new product launches slowing; and products that are launched becoming increasingly unstable.
Avoiding this requires “freeing up dollars from keeping the lights on and putting those dollars to use creating better products and services for customers,” says Mason. “If you're drowning in tech debt, it becomes impossible to do that.”
Murphy points out that traditional and digitally native businesses alike inevitably confront the same difficult question when the technology function reaches a certain scale: “How do you enable your technology teams to operate effectively and efficiently, to embrace the latest opportunities and the advantages that you have with your legacy data and technical estate, rather than be constrained by it?”
“Businesses that get it wrong spend all their time dealing with legacy issues and don't bring new products to market,” he says. “On the other hand, enterprises that only seek to bring new products to market and ignore technical debt start becoming very inefficient.”
Tech debt is another challenge that can be resolved through a thin-slice strategy that, rather than attempting to fix everything at once, concentrates technology initiatives and investment around specific functions cutting across all products and aspects of the business, “addressing legacy while also addressing the future,” Murphy notes.
Iterative value
vi. Partnerships and creating virtuous change cycles
In trying to ensure technology supports the enterprise’s growth trajectory, enterprises should also bear in mind they don’t have to go it alone. In fact, attempting that can hold the company back.
“The technology landscape is now so broad that it's a very big ask for an organization to have capability across every aspect of it,” Mason explains. “A strategy of minimizing third-party dependencies can cause organizations to suffer, because they're forced to put people in roles where they don't have the required expertise.”
“The technology landscape is now so broad that it's a very big ask for an organization to have capability across every aspect of it.”
Mike Mason
Chief AI Officer, Thoughtworks
“Especially with technologies like AI, it’s quite easy to do the wrong things, so securing that expertise through a partner is really important,” he adds. “The other thing that partners bring and that enterprises can benefit from is cross-industry perspective.”
“The ecosystem is exploding at the moment, and it's hard to distinguish useful tools that could have an impact from those that are more hype,” agrees Nolan. “Having trusted partners that can help you navigate that fast-moving space is key. There's nothing better than learning what others in the market are doing in a transparent way.”
Even so, “business strategy is now so integrally related to technology strategy that every business needs to have a strong, robust and leading internal technology organization,” Murphy says. “Organizations that are successful have a core function that understands the business and existing tech deeply, and is critically involved in setting strategy – but also one or more best of breed partners who are able to act as inputs into that strategy, as well as bring the external thinking that that organization needs to thrive.”
“It's really important to be specific about which capabilities are genuinely important to have in house, and which you’ll work with a partner to evolve over time,” Mason notes. “You also need to know enough about a particular technology to be able to manage a partner who's supplying expertise in it - otherwise you’re completely beholden to them.”
The best technology partners don’t just act as advisors; they work directly with clients to identify and realize opportunities for additional value.
“Running apps, for example, has often been this process where you take a shiny new piece of software that you've spent years building, and then try to find the cheapest possible set of people to hand it over to for support,” says Mason. “That's where applications go to die. You've got all this value locked up, and the software can’t continue to evolve. Thoughtworks’ approach is to revitalize running systems in production to unlock value and add new features. Organizations need to make modernization continuous – while optimizing outcomes over time.”
“If organizations are able to get the balance of in-house and external, of tech and strategy, right, they get a self-reinforcing loop where new products are successfully launched using the latest capabilities and create higher revenue growth and better margins, enabling greater reinvestment,” says Murphy.
Along with solid partnerships, developing an in-house ‘early-warning system’ by paying close attention to new technology trends, and building the enterprise’s capacity to act on them, can help drive future growth. Thoughtworks experts encourage clients to engage with future-focused industry research like the company’s closely followed Technology Radar and Looking Glass reports; to network with technologists and peers at industry events; and to seize on opportunities for education and skills development.
“External sensing is becoming critical - understanding not only what's going on in your industry, but what's going on in the technological space, to be able to see around corners and to predict what the next trends might be, and start placing bets now,” says Murphy. “A lot of those bets will be wrong, but one or two might be right, and they’ll give you that competitive advantage.”
"When developing systems, it's okay to make a decision based on the information you have today, but make sure that the architecture you're developing is evolvable, so that you can switch pieces out as this landscape continues to change.”
Andy Nolan
Director of Emerging Technologies, Thoughtworks
“You can make a bet today that might not be the best bet in 12 months’ time,” says Nolan. “So when developing systems, it's okay to make a decision based on the information you have today, but make sure that the architecture that you're developing is evolvable, so that you can switch pieces out as this landscape continues to change.”
“The key organizational capability that’s going to distinguish 21st century winners from losers is the ability to embrace change,” says Mason. “This year AI is a big thing. But in a few years’ time, we might see quantum come into the fore, or breakthroughs in extended reality. If you get an insight into where things are going, you’ll have less and less time to capitalize on it. You need strategy, technology and organizational structure lined up so when that insight comes, you’re ready to maneuver.”
Technology blueprint for growth
Perspectives delivered to your inbox
Timely business and industry insights for digital leaders.
The Perspectives subscription brings you our experts’ best podcasts, articles, videos and events to expand upon our popular Perspectives publication. (* Required fields) (* Required fields)