“Please expect a delay in delivery. I spent the last half hour at the bomb shelter. I will complete the job and deliver ASAP.”

Emails like this are not unusual in the inbox of Ruth Partington, founder and owner of Empower Translate. It would be hard to quibble over missed deadlines, given the circumstances. But conducting research in zones of conflict demands flexibility. These are turbulent times, and every geopolitical tremor sends a fresh shock wave over those on the ground.

The Global Peace Index 2024, published by the Institute for Economics and Peace, reported that there were 56 active conflicts – the highest number since the second world war – and 92 countries involved in conflicts beyond their borders. 2024 was, the report said, ‘the fifth consecutive year that global peacefulness has deteriorated.’

It’s not just conflict that is having a dramatic impact on the business of research. With political, social and economic conditions intractably unstable, organisations around the world are facing an unpredictable and high-risk trading environment.

Commissioning and conducting research is a complex undertaking involving decision makers and stakeholders from functions across the business. In addition, a single piece of strategic research can require an intricate network of specialist agencies. Each stage in the research process, each link in the supply chain, will be impacted by shifting political alliances, regional conflicts, and overnight switches in public opinion.

In such a precarious climate, Clifford Young, president of Ipsos Public Affairs in the US, is finding a growing demand from clients for faster turnaround of projects, especially tech-enabled forecasting and foresight. 

Not only is political risk baked into a business’ core operations, but the specifics of individual research projects require a high level of scrutiny when dealing with divisive themes or controversial clients, as seen with Stagwell’s recent work for the Israeli government.

The trickle-down influence from global issues to local impact is especially evident in the research work that Verian executes for governments and public sector institutions around the world.

“Public sector clients increasingly need research that connects international developments with domestic outcomes,” says Craig Watkins, chief executive at Verian UK. “There’s a shift from purely UK-focused studies to research that tracks how global instability affects migration, energy, trade and service demand at home.”

Fieldwork in Burkina Faso © ORB

Global issues, local needs

The type of research localisation work undertaken by Empower requires cultural sensitivity and hyper-localised language skills; this is to ensure that how market research is conducted and communicated reflects the culture of a very specific region. Think in terms of dialect rather than language. There are approximately 7,000 living languages in the world, but the number of dialects – local variations – is far greater.

Because the ebb and flow of language work reflects the vagaries and shifting alliances in world affairs, the pipeline for Empower’s work is difficult to anticipate, whether it supports insight designated for commercial or humanitarian outcomes.

Whereas 10 years ago, the industry might have more readily predicted an increase in requirements for research into certain markets, this has now become more difficult – and therefore more difficult to resource in advance.

According to an update issued by the International Monetary Fund (IMF) in July 2025, the current trade environment remains precarious, with ongoing trade uncertainty. In the summer of 2025, trade rifts saw some US brands boycotted and a decline in tourism into the US from Canada.

“We’re seeing unprecedented unpredictability in the markets where global brands want – and are able – to operate,” says Partington. “Especially for researchers, having access to on-the-ground, local resource is more important than ever.” 

The rise in conflict

One trend that is certain for Empower is an increase in work commissioned in response to zones of conflict and war. Partington says: “We have definitely seen an increase in requirements for both conflict-driven and conflict-impacted research. What we're being invited to do now is work with researchers on more humanitarian and philanthropic studies.”

The UNHCR reported in 2024 that 123.2 million people worldwide were forcibly displaced as a result of conflict.

The term ‘conflict’ needs to be defined. Partington offers a two-sided definition of conflict: “In research terms, when we talk about ‘conflict-driven’, we are talking about going into a war zone and talking to people, for example for humanitarian purposes, or to people displaced, for example, refugees in the UK.”

By contrast, ‘conflict-impacted’ is where research is impacted ‘indirectly’ by conflict.

Translating English language surveys into a local language or dialect, or performing live interpretation during interviews and home visits, is incredibly nuanced – even more so in environments where misunderstandings can have real-world implications.

“The way in which you word questions and the way in which they're translated is absolutely critical,” says Johnny Heald, UK chief executive, ORB International. He uses the word ‘jihad’ to illustrate how words can have multiple, very different meanings. “Many people think it only relates only to suicide bombers. What you and I understand to be the meaning of the word might be very different. Your ‘jihad’ might be you go and clean the mosque on Friday because that’s your job.” In fact, ‘jihad’ is a common word used to describe any number of charitable efforts a Muslim may undertake to support their community and pursue a virtuous life.

Life on the ground

Running an effective research localisation business like Empower requires a careful balance between achieving something that is culturally sensitive and empathetic (especially considering the traumatic circumstances in which people may find themselves) and yet highly practical: how can Empower protect and pay their moderators where there are no banks, no internet and no infrastructure?

Within these constricted circumstances, the ability to put people on the ground in any part of the world and at short notice is a key differentiator for agencies. North Korea aside, there’s rarely a community that cannot be reached, says Heald, whether it’s running focus groups in Gaza or talking to defected Isis fighters. Where access on the ground really is impossible, talking to displaced people can be used as a proxy.

Running the risks

Safeguarding is a priority for employers. Tragically, the impact on employees can extend well beyond post-traumatic stress disorder (PTSD). The fate of local workers in Afghanistan, including interpreters and fixers for the US military, who were left behind at the mercy of the Taliban, raised ethical questions around the responsibility of employers (in that instance, foreign governments).

Since 2009, an estimated 300 interpreters working with the US military in Afghanistan have died while seeking a US visa, and the UK’s recent data breach implicated not just employees but their extended families.

It is customary to run detailed risk assessments for any country or environment before work starts. As well as physical risk, ORB will run cognitive testing of all research materials to understand how respondents will interpret questions.

“Interviewers will say ‘I cannot ask that question’,” says Heald. “So, for example, I was in an Islamic republic in West Africa and the client wanted to ask a question about LGBTQI+ issues and everyone in the training session said there was no way you could ask that question. 95% of what we do is face-to-face and random probability sampling, and we won't send people into a potentially harmful situation. So, we'll just take that question out.”

However, in a war zone, no amount of due diligence is completely fool proof, and when three interviewers working for ORB were kidnapped and murdered, Heald couldn’t be sure of the motive: ”We didn't know whether or not they were being targeted for what they did, or whether it was random. We stopped working for three months. It was the interviewers in the end, who kept coming back saying: ‘Please, please, we need to start again, we need to put food on the table.’ So, we started again.”

As we have seen, the languages in demand at any one time reflect the geopolitics of war and regional instability, and the subsequent displacement of people.

For Empower, more of today’s work comes via government agencies, NGOs, and philanthropic or humanitarian organisations. What Partington describes as ‘rare language’ work is on the increase as more people from every corner of the globe are being displaced and the need for ‘niche’ language services becomes more acute.

Another international agency, Ronin, has worked on a resettlement program over the last nine years, interviewing refugees in the US –  a typical example of what Partington earlier defines as ‘conflict-impacted’ research. For Ronin’s project, Armenian features on their list of language requirements in 2023, but it’s missing in 2025, and instead there’s a new requirement for Kibembe – a Bantu language from the Democratic Republic of the Congo.

Simon Glanville, managing director at Ronin, says it is non-negotiable that all telephone interviews with refugees in the US are conducted in the participants’ native language.

The impact of being heard, listened to, cannot be underestimated, especially for people whose existence is under threat.

“Response rates are actually surprisingly high because people are almost delighted to be able to express an opinion,” says Heald. “I've been in focus groups where people start crying. It’s like a therapy session.”

Winning the confidence of participants is never more important than in environments where people are likely to be mistrustful of strangers asking questions.

The changing battlefront

As well as these continually evolving language requirements, at ORB, Heald has seen a fundamental change in the topics being explored in research. Once again, this is reflective of the real-world changes in conflict and warfare, and in particular, the new technological battlefront; that of disinformation.  

“There is absolutely a war going on in the information environment right now,” says Heald. “Russia is very adept at communications and disseminating, whether it’s through RT very openly, or through bot farms."

Identifying how people access information and what they are absorbing is critical to understanding the experiences of those impacted and driven by conflict. Information distributed online can permeate even the most remote environments and its influence can be immediate and visceral.

“We need to understand that anything that is connected to the internet should be considered a potential weapon of war,” says Partington. “If you're connected, you can be hacked.”

With war increasingly taking place in the tech space as much as the physical one, protecting data is paramount.

Language data inequality

The advent of artificial intelligence (AI) has accelerated the development of technical warfare. But aside from weaponry, there are unintended consequences of embedding AI into our everyday systems which exacerbate the marginalisation of communities, especially those impacted by conflict.

Generative AI also poses a double risk in that it is biased both towards data emanating from WEIRD societies (Western, Educated, Industrialized, Rich, and Democratic) and with the dominant written language being English (British and American).

The large language models (LLMs) that drive generative AI solutions, for example ChatGPT, have been programmed with a WEIRD psychology, which is not representative of the global population.

This was demonstrated by research from Harvard University which showed the more distant a country is culturally from the US, the lower the correspondence between AI responses and human responses. This is especially pertinent given the enthusiastic adoption of synthetic data by the research sector.

Partington underlines the importance of ‘language data equality’ – without which many global communities will not have a voice. 

Representation and bias are well documented weaknesses of any AI. But are there positive aspects to AI that can help clients achieve their objectives?

Glanville says yes, with limitations. AI references historic data that is unlikely to reflect the unique and shifting experiences of a diverse spread of under-represented people and communities.

In reference to Ronin’s project interviewing refugees in the US, Glanville says: “Ultimately, the kind of the interviews we’re doing, particularly on a job like this, are really what the project is all about. You can't really automate that if it feels out of step with somebody being comfortable in an interview situation.

“I've no doubt that over the years things may change a little bit, but at the moment, we know that a properly administered questionnaire will often yield things that no one would really expect someone would say. And that gets to the heart of what market research really is.”

Regardless of whether a research supplier is working in a zone of conflict or not, quality is now the watchword for the entire sector. Data is under vigorous scrutiny from clients who are much savvier than they may have been in the past when ‘big’ was enough, with concern around the authenticity of respondents and widespread survey fraud.

“As an industry, we need to be collecting really good quality data that people feel the value of having to pay that extra money for,” says Glanville, who also sits on the Global Data Quality steering group.

When providing localisation and language services in such a sensitive and high stakes environment – an area of conflict – there is little if any margin for error. Missteps can have fatal consequences; or at the least offend or confuse, and invalidate research. Many agencies at the forefront of localisation are conducting work that above all else supports life-changing humanitarian outcomes for millions of people around the world.

But the bigger picture is how the issues discussed in this report reflect the wider challenges to the global research sector. War and conflict, global politics and public sentiment have a direct and often immediate impact on the daily work of all researchers, regardless of speciality or brand. The tectonic plates underpinning world politics lurch in a new direction, a tsunami of opinion is unleashed, and whole markets are disrupted, all in the time it takes to share a single social media post.

Watkins sums up the position not just for Verian’s public sector clients, but the future of all insight work: “As geopolitics moves faster, our insight systems need to move faster too…The future of [public sector] research will be defined by agility, global awareness, and the ability to connect macro risk with local response.”