Artificial intelligence may dominate technology conversations, but for many organisations, the biggest obstacle to AI adoption is not the technology itself.
It is integration.
This was the topic discussed at a recent UK NashTech Leaders Lab event attended by 20 senior technology leaders from sectors including higher education, banking and enterprise technology.
The IT leaders, who are all part of the NashTech Leaders Lab community group, spoke about fragmented systems, poor data quality, legacy architecture, organisational culture, partner relationships and the challenge of proving long-term value when it comes to AI.
‘Listening in at the Leaders Lab’ is our latest series designed to give you direct insight into the minds of some highly experienced technology leaders.
Let’s get into the first in the series…
The discussion opened with a fundamental question: why does integration matter so much?
For Mohan Kandola, Vice President, Global IT Strategic Business Partner at Goodyear, the answer is simple.
“When you look at applications, ask whether they can operate in isolation. In reality, they can’t, because business processes depend on systems working together. That’s why integration matters across data and processes. If you want to run a business effectively, you need strategic integration across the board.”
His point resonated across the room. Integration is not simply about connecting technology. It is about enabling business processes, customer journeys and organisational agility. But it can be challenging, and the challenge becomes even greater when organisations attempt to layer AI onto disconnected systems.
Chris Price, Fractional IT Director, agreed, highlighting the growing complexity organisations face.
“Everybody is working to different standards. You can try to put some kind of master data layer over the top, but even then, it remains a major challenge.”
Many leaders acknowledged that while AI is generating excitement, basic integration and data management challenges remain unresolved in many organisations, and it’s highlighting so many weaknesses.
A recurring theme throughout the discussion was that AI is not ‘creating’ integration problems. It is just revealing the problems that were there all along, such as poor process management.
Tony Colson, Director of IT at the British Psychological Society, argued that technology challenges often stem from deeper business process issues.
“Data maturity and process maturity go hand in hand. We often talk about process ownership as if it is a technology issue, when in reality it should sit with the business. For me, there is also a question of skills, not just technical skills, but skills in understanding and managing business processes. That links back to documentation as well. Do the people who own these processes have them properly documented and do they have the capability to define what a good process looks like? That may be one of the missing pieces we still need to explore.”
Without clear ownership, consistent governance and documented processes, AI risks amplifying inefficiencies rather than eliminating them.
Andrew Morrish, Interim Executive at Ratiopraxis and Technolink, echoed this concern, arguing that many executive teams have yet to fully embrace digital thinking.
“In many businesses, it is still not a digital-first conversation at board level. As a result, decisions are often shaped by how things have always been done, rather than how they should be done. That holds back every conversation that follows — around funding, prioritisation and the ability to make strategic changes. Until organisations can have that conversation properly, many will struggle to take the next step. AI will remain a distant pipedream: something confined to proofs of concept or limited use cases that produce outcomes of questionable value.”
As a result, organisations often struggle to make the strategic decisions required to modernise technology estates and prepare for AI adoption, with consequences that seem, frankly, predictable.
Without the funding, the understanding, and the confidence, AI initiatives will remain trapped in pilots and proofs of concept rather than delivering meaningful business outcomes. Therefore, the boardroom ‘sell’ becomes increasingly important.
While business leaders readily understand investments in customer-facing innovation, infrastructure and integration programmes often struggle to attract support because their benefits are realised over time rather than immediately.
Mohan Kandola described the challenge:
"With some solutions, the benefits are not immediate. The value often comes later, once the integration with other systems is in place. That can make it difficult to convince stakeholders because, at first, you are asking them to trust that the benefits will follow. But without that investment, organisations stay stuck with legacy systems, continuing to patch things together rather than addressing the root problem. In many ways, one of the biggest challenges for technology leaders is persuading the business to back the right long-term approach, even when the payoff is not instant."
James Longmore, Director of Digital Services at Aston University, compared the situation to constructing a building.
“We often need to sell the right way of doing things to obtain funding. I sometimes use the construction industry as an illustration: if you are building a new office block, people can see the foundations going in and they understand why they are necessary; they see the concrete being poured. In technology, it is harder to visualise. I have been guilty myself of walking executives around a new data centre to explain what has changed and what has been invested in. At Aston, for example, we spent a large amount of money and time renewing our network because it needed to be done. But when the question comes back — what are we going to get from this? — the honest answer is often: nothing immediately. The value comes later. In three or four years, it gives you the platform to build the things you want to build and the security and longevity that comes from a refresh. But that makes it a difficult conversation, because the response is often: isn’t AI going to do that for us? And the answer is no, not unless we put these foundations in place first. As a technology leader, you can end up taking the easier route and pushing these changes through gradually, almost by stealth, because you know they are necessary but can't persuade others. The bigger question is whether the board truly understands what is required underneath the surface and what the ongoing cost and effort involved in those foundations is”
The conversation highlighted a growing tension for technology leaders: knowing the importance of foundational investments but struggling to communicate their value to boards focused on short-term outcomes.
The question of value surfaced repeatedly throughout the roundtable.
Charlie Houston-Brown, Managing Consultant and Chair of the Digital Forum of the Staffordshire Chambers of Commerce, noted that many organisations are struggling to quantify the return on AI investments.
"You can spend money on AI, use up tokens and incur costs, but still struggle to say exactly what value you have gained."
This challenge becomes even more complex when organisations are investing in architectural improvements, integration platforms and modernisation programmes that create value over many years rather than delivering immediate financial returns.
Tom Lee, Commercial Director at NashTech argued that AI is introducing entirely new commercial uncertainties.
"Businesses may know what they want to achieve, but they do not know what it will cost. Organisations now face a growing mix of software licences, token consumption charges, data access fees and cloud costs. As a result, business cases are becoming harder to construct and harder to defend.”
Peter Truman, Software Development Manager at Weatherbys Private Bank, suggested that stronger governance and board-level involvement may be the answer.
“CFOs can be extremely helpful because they can ask the simplest and most important question: What value are we actually getting for the money we’re spending? Companies can often be paying for an expensive SaaS product without clear ownership, proper review or enough challenge to the suppliers. A finance ‘challenge’ can be a lever to drive engagement in structured service reviews, with relationships becoming more strategic. The same issue now applies to AI. Costs can rise quickly, so organisations need much stronger oversight. In banking, that means knowing where AI is being used, what models are involved, where data is processed, and how those systems are governed, tested and monitored as they change.”
Whether reviewing SaaS platforms or AI services, leaders increasingly need visibility into costs, ownership, performance and outcomes. The group then discussed how important partners were in supporting them with this visibility.
A recent NashTech survey found that only 32% of technology leaders view their third-party software development providers as strategic partners, meaning partners who are aligned with long-term goals and involved in high-value initiatives.
Chris Weston, Senior Technology Consultant at NashTech, referenced this recent NashTech research further to show that organisations increasingly want strategic relationships.
“97% of technology leaders in our survey said they would be willing to pay more to a technology partner who provides greater long-term value. The question is whether that happens in practice."
Sharon Penfold, an interim technology leader, argued that partners should be explicitly encouraged to challenge.
“Partners should be expected to do more than simply say yes and send the invoice. I’ve seen this happen far too often: a business sponsor or product owner asks for extra features, and the supplier or consultancy immediately says yes. Before long, the scope keeps growing, and so does the bill. To address that, I developed a partnership agreement that requires suppliers to challenge requests and be much more transparent about what is needed and why. I would strongly recommend that approach. It should be written into the contract. For me, that is one of the most important choices an organisation can make when deciding what kind of partner it wants to work with.”
Tony Colson reflected that many organisations still have suppliers rather than true partners.
"I would say I have a lot of suppliers, but not many true partners. The challenge is that it is hard to build genuine partnerships when procurement is saying, ‘You are three years into your contract, so what do you want us to do?’ That creates another dynamic for technology leaders to wrestle with. It becomes a question of how we make the case, how we explain the value of partnership, and how we educate our peers across the organisation.”
The discussion highlighted how procurement processes, governance structures and commercial models can unintentionally discourage the behaviours organisations actually need.
Kuldip Sandhu, Managing Director of IQS took the conversation further.
“Contracts alone are not enough. You also need to assess culture. You need to create the right incentives for strategic partners through the contract, with clear outcomes, as well as clear responsibilities, asking yourself, ‘Will this partner go above and beyond to solve problems in the right way, or will they simply do the minimum?’ That challenge becomes even sharper with AI, where the pace of change is extraordinary.”
As AI evolves rapidly, organisations increasingly need partners capable of adapting, challenging assumptions and supporting continuous improvement rather than simply delivering predefined outputs.
Another key theme focused on balancing standardisation and differentiation.
As organisations consolidate and integrate systems and move towards global operating models, they often discover significant inconsistencies in data, processes and ways of working.
Mohan Kandola observed:
"If the core systems are becoming more standardised, your uniqueness has to come from elsewhere."
However, Bill Schindler, a Freelance Consultant, offered a reality check.
“I’ve seen how difficult global standardisation can be in practice. For example, we had 38 different SAP instances around the world, with little or no real integration between them. Over seven years, the business invested around three-quarters of a billion pounds, working with multiple major partners to bring that landscape together. But even then, the outcome was not a single source of truth. It became several versions of the truth, shaped by local requirements such as language, financial regulations and regional complexity. For me, that shows that in large global organisations, a perfect single source of truth may not always be realistic. The more useful question is: what level of consistency and utility do we actually need to run the business effectively?”
As the conversation drew to a close, participants reflected on how familiar many of these challenges felt.
The technologies may have changed, but the underlying questions remain remarkably similar.
Chris Price captured the mood perfectly.
“I remember having very similar discussions back in the late 1980s and 1990s about the classic buy-versus-build debate. At the time, we were promised things would become much easier in the future. But here we are again, still going round the same cycle. The technology changes, the labels change, but the underlying tension remains: do we buy, do we build, and how do we avoid repeating the same problems? In many ways, it feels like a perpetual loop.”
His conclusion drew smiles around the room.
The Leaders Lab revealed a reality that many organisations are now confronting.
AI is forcing organisations to revisit some of the most fundamental questions about technology strategy, operating models and business transformation.
Organisations that have invested in integration, improved data quality, strengthened governance, standardised where appropriate and built genuine partnerships that focus on outcomes rather than transactions are finding themselves in a better position to embrace AI.
Ultimately, to be AI-first, organisations need to become integration-first too.
If these Leaders Lab insights reflect challenges your organisation is facing, speak to a NashTech advisory expert. We can help you assess your integration maturity, identify where disconnected systems are creating risk and shape a practical strategy for building stronger foundations for AI and digital transformation. To request to join the NashTech Leaders Lab community group, contact Natalie Whittlesey - leaderslab@nashtechglobal.com .