Managing AI: How organisations can adapt traditional business management tools to succeed

Taking advantage of the opportunities that artificial intelligence brings isn’t just about installing new technology. It requires organisations to prioritise strong governance, invest in their people and implement robust innovation frameworks. These pillars improve the likelihood of AI adoption enhancing an organisation’s operations and service offerings, while effectively managing the associated challenges.

Traditional business management models such as “People, Process, IT” and “SWOT analysis” have limitations in the current environment of rapid change and innovation. An evolution in these models is necessary to ensure that agencies like ours, and other organisations, can adapt whilst using tools that are familiar, well-understood and make it easier to thrive.

SWOT analysis for agencies in the age of AI

Like it or loath it, SWOT analysis is widely used to guide businesses through change. But, in immature markets, strengths and weaknesses are hard to define. When technology adoption is unclear, it’s better to focus on opportunities and threats rather than current advantages and shortcomings.

Key threats to the current agency model and services

Using our own sector as a guide, a lot has been written about the potential threats that AI brings to our industry including:

Relevancy: With ongoing advancements in AI, agencies might have to revise their services, since some traditional options could lose relevance for clients.

Fee pressures: The efficiency gains enabled by AI might lead clients to expect lower fees, potentially undervaluing the enhanced productivity and higher quality agencies can deliver

In-housing: Clients may see AI as an opportunity to bring more capabilities in-house, reducing their reliance on agency partners for certain tasks or services.

Dogmatism: The rapid pace of technological change presents a significant challenge to agencies dependent on rigid processes.

It is important to remember that these concerns are familiar territory and predate the existence of AI by some time. For agencies that take initiative, the chances to benefit far outnumber the risks.

Agencies have always been required to navigate evolving market conditions and changing client needs. True expertise and adaptability are still at the core of their success.

Agencies are uniquely positioned to draw upon the collective experience gained across a diverse set of clients and sectors. They are well positioned to quickly develop an informed understanding of what strategies and solutions are effective in the marketplace today. This insight enables them to anticipate shifts, stay ahead of trends and lead clients in the adoption of new approaches – ensuring they continue to deliver meaningful outcomes in an environment shaped by rapid technological advancement.

Opportunities presented by new technology

As mentioned, I think AI’s benefits can easily outweigh threats if businesses address challenges and adapt effectively. Emerging technologies ultimately create new possibilities, enabling agencies to innovate at pace.

Brands are expected to increase their marketing technology budgets, creating opportunities for agencies to bring these new innovations from concept to reality much quicker.

New technology capabilities are also driving channel shifts in end-users. Brands need to respond to this and will turn to agencies to help them be present in new channels whilst restructuring existing ones.

Finally, these new technologies allow individuals in agencies to develop deep and broad capabilities as well as having deep and broad sector experience. Combining these factors will allow agencies who invest in their people to respond faster and more flexibly than ever before to client needs.

The big opportunity: A shift from output to outcome

Perhaps the biggest opportunity for agencies is in producing higher order work. The mark of a great agency is its capability to direct a process that leads to the right output for a desired outcome, rather than simply focusing on the act of production itself. This approach ensures a clear connection between what is created and the broader result the client seeks to achieve.

The true worth of an agency is not in the mere creation of a high-quality documents, visuals, code or adverts. Rather, it is defined by the strategic decision regarding which output to pursue and the process that leads to that decision. Whether creating the output itself shifts from primarily human-driven to largely AI-driven, the intrinsic value provided by the agency remains in its ability to identify and justify the chosen output.

Leveraging technology to enhance value

Historically, agencies have consistently adopted the latest technologies to expedite task delivery – be it advanced printers, cutting-edge Apple Macintosh computers, or specialist software such as Photoshop or Visual Studio. These technologies were often costly or complex, necessitating specialist knowledge and expertise. Today, however, recent advancements have led to tools that are both affordable and accessible, thanks to natural language interfaces that promise ease of use even for non-experts.

Despite these advances, the role of the expert is more critical than ever. As AI technologies elevate the baseline of what is considered average, the capacity to distinguish oneself and add genuine value increasingly depends on the expertise required to extract that final, elusive 1–5% improvement that makes an ever increasing difference.

The importance of AI skills and expert deployment

The effective use of AI tools has become a skill in its own right. These tools have evolved beyond simple chat interfaces into complex suites comprising assistants, agents, and workflow automation. Determining which tool or combination of tools to deploy demands careful consideration and genuine expertise. It often requires the implementation of bespoke configurations tailored to specific needs.

Agencies can deploy AI tools at scale and deliver specialised outputs. Clients may have the same tools but rarely the expertise required to match the quality or outcomes agencies achieve.

Emerging products

Inevitably new markets and new channels driven by technology require new products. This too presents an opportunity for agencies prepared to invest in R&D and product development. IDHL Labs is part of our response to this opportunity and already seen AI specific capabilities launched within our creative and performance marketing product sets.

Empowering people, evolving processes, leveraging technology, driving innovation

I started this article by saying that, traditionally, building a business in established markets has relied on the well-known formula of investing in “People, Process, and IT”. However, the game has changed. In this rapidly evolving and less understood environment, businesses and agencies need to do more. The current environment demands ongoing investment not just in people, process, and technology, but also in real innovation to ensure organisations remain at the forefront of their sector, delivering meaningful services to clients.

People: Navigating change and empowering adoption

People are often the main barrier to adopting new technologies. Change is challenging, frequently met with resistance due to ingrained habits and organisational culture, making it hard for even willing individuals to adapt.

To address these challenges, it is essential to invest in comprehensive training and growth enablement for your teams. Training should not be limited to the technical aspects of using new tools either. It must enable effectively working with AI. Skills such as prompt engineering should already be a fundamental requirement for anyone entering the workplace.

Empowerment must be enabled to make informed decisions about when to use AI and when it may be more appropriate to rely on human judgement. The relationship between AI and human input is best understood as a spectrum rather than a binary choice. It is not simply a matter of whether something is human-created or AI-generated, but rather the degree of human and AI involvement at each stage of the process.

Our colleagues know not to rely on AI-generated outcomes. Instead, we advocate scenarios where AI may produce an output that is then human reviewed, in conjunction with efforts where both human and AI collaborate to produce outputs. It is equally valid for a human to create something in its entirety or with an AI reviewer involved.

It is vital for colleagues to be trained when to recognise that additional scrutiny is required and when to double-check AI outputs. They need to be aware of the potential biases and hallucinations that are commonplace when working with LLMs – taking steps to avoid perpetuating these counterproductive or harmful outputs.

Furthermore, teams should be conscious of how these tools operate, particularly regarding data privacy. They need to be aware of whether the data they input into AI tools could be used for further training or might risk exposure to the public domain. Handling sensitive information, therefore, demands careful curation and a thorough understanding of data governance protocols.

Ultimately, effective and ongoing training of personnel is fundamental to ensure that teams are well-prepared to work with AI technologies in a secure and responsible manner.

Process: Automation, reliability, and human oversight

Automation is often the go-to consideration when discussing adoption of AI into the workplace. Automation of manual, repeatable processes has huge potential for many organisations and one of the prevailing ambitions is that AI will finally fulfil this potential.

The current landscape of AI tools is also empowering colleagues to create automated processes that previously needed the expertise of a developer or automation specialist. For example, when working with a new client, IDHL historically had a manual checklist shared across multiple teams to ensure smooth onboarding of the client into delivery. Today AI agents are being deployed to do the manual checks automatically and flag any anomalies with relevant team members in real-time.

Consequently, organisations will see more automated workflows across business units. Managing and updating these systems will become a challenge, especially as oversight and human involvement must be carefully considered. With autonomous agents now able to make decisions, it is increasingly important to clarify where human input is needed within these automated processes.

Expecting AI to seamlessly automate processes is not without its own challenges. LLM’s are nondeterministic, meaning they do not always produce identical outcomes from the same inputs. For processes which demand reliability and consistency this can be a deal breaker. A greater opportunity may exist where human intervention in a process is currently required. Making a statistically informed and safe judgement call might be where the apparent unpredictability in LLM outputs can be an advantage.

Quality control is essential. As with manufacturing, AI-driven automation needs to follow strict standards. Organisations must maintain high levels of quality as automation expands, with responsibility for overseeing automated outcomes likely shifting to senior staff instead of less specialised employees.

Information technology: Navigating rapid change

Currently, IT isn’t the barrier; technology is advancing faster than businesses and individuals can adapt. The main challenge is the urge to chase every new product, increasing business risk through frequent pivots and changes.

Compounding this situation is the market dynamic, where major software vendors are aggressively competing for market share, often driving down costs in a bid to secure dominance. The outcome of this intense competition is crucial, as the provider that succeeds in this current 'land grab' over the coming 6 to 18 months sets themselves up as the dominant player for the next 5 to 10 years. Established giants such as Google and Microsoft appear well-positioned to prevail, given their capacity to fund innovation and ensure their products reach key users on a global scale.

For organisations, it is therefore prudent to limit the number of platforms and providers used and deploy their solutions consistently throughout the business. Having many point solutions risks needing to integrate these further down the line. Creating the right balance between deploying general purpose AI capability and specialised AI solutions (such as software development and image generation) is something each organisation should consider.

Over time, new features in the technology will plateau, meaning choice of provider is less critical than many may think at present. First mover advantage has been a key consideration in the past, but in the current environment, its impact is fleeting, often lasting only days or weeks.

Innovation

While the elements of people, processes, and IT provide a strong framework for integrating established technologies and achieving steady, incremental growth, the current pace of innovation demands a more direct and proactive approach. To remain competitive and responsive to market changes, agencies and, in fact, almost all organisations, must dedicate more resource to technology-driven innovation.

Enhancing innovation does not necessarily mean hiring dedicated R&D personnel. AI-driven productivity improvements create capacity within existing teams; organisations should consider utilising this excess capacity in whole or in part to driving technology and AI driven innovation activity.

At IDHL, our approach has been the creation of IDHL Labs - a collaborative initiative bringing together passionate individuals from across our organisation. The purpose of IDHL Labs is to accelerate the adoption of new technologies and to foster a culture of innovation throughout all our teams. We recognise that, as the business environment experiences unprecedented disruption due to advancements in AI and related technologies, innovation will be a key differentiator in the next phase of organisational development.

In uncertain times, success demands experimentation and testing of new ideas. While many attempts may fail, those that succeed will drive the business forward and create key advantages. These successes will empower our people, enabling them to achieve the marginal but vital differentiations that are increasingly difficult to secure in a crowded marketplace. Equally every failure is still a success in our eyes, helping us to divert resource away from unsolvable challenges and evolve hypotheses towards genuinely valuable outcomes.

Moreover, we see value in progressing swiftly through this period of upheaval towards a more stable future, where a return to “people, processes, and IT” is once again the norm. Getting to that point depends on sustained investment in innovation.

Governance in the ‘Age of Innovation’

Governance is a critical consideration given the uncertainties (perceived or real) surrounding AI. The importance of the “human in the loop” has already been highlighted, this represents just one aspect of a broader governance framework.

A comprehensive approach to governance must include clear guidance on tool and provider selection – ensuring safe and appropriate use within the organisation whilst not stifling the ability to experiment.

Managing technological risk and promoting responsible AI usage

In a landscape, where new products launch daily, the risk of selecting a provider that may quickly disappear from the market is greater than ever. The evaluation framework for approving new tools should take this into consideration, working to the modern best practice of composable architectures will help to reduce this risk.

It is equally important for colleagues and teams to understand the implications of unprofessional or careless use of AI tools. Presenting unvetted outputs especially those containing inaccuracies or so-called “hallucinations” - to customers or the wider market can be highly damaging to the organisation’s reputation, just ask Deloitte. When working with AI generated outputs colleagues need to understand that when they accept the output, it becomes their output and they assume full responsibility for it.

Strategic data governance for sustainable AI success

Governance goes beyond tool selection and user behaviour to include ongoing management of organisational data. AI initiatives need high-quality contextual data, making investment in robust data storage and classification essential for long-term success.

Given the pace of change in the AI marketplace, some organisations may benefit from adopting a “wait and see” approach to direct AI technology adoption. A short-term focus on strengthening document governance and data cleanup may pay dividends setting these organisations for greater success over the next two to five years, as new technologies will have greater impact when implemented across a well-organised data and document ecosystem.

Ultimately, the way in which data within the organisation is organised, categorised, and managed will have a profound impact on the success of future AI ventures. Effective governance, therefore, is not just about compliance or risk avoidance - it is a strategic enabler of innovation and long-term competitive advantage.

Where this leaves us

The simple truth is this: AI will force organisations to move faster than traditional management models were designed for. But the answer isn’t to abandon those models, it’s to evolve them. The organisations that will thrive are the ones that continue to invest in their people while modernising systems and processes, strengthening governance and creating the space for real innovation.

The future belongs to those that treat AI not as a shortcut, but as a catalyst for better thinking, better systems and better outcomes. Good luck.