There are many pieces in the puzzle when it comes to designing good market research projects but arguably none is more important than selecting who will be part of that research.

In our experience, this is part of the process that clients can sometimes struggle with. While a research project may be commissioned by the market research team, the outputs will often need to serve multiple functions – such as marketing, sales, PR and research – as well as regional teams.

Attempting to support these different objectives without overshooting your research budget may require some trade-offs. Getting those decisions right is crucial for maximising the value of your research project. With that in mind, we’ve outlined three key tips for optimising survey sample selection during the research design stage.

1). Engage all relevant stakeholders from the outset

When kicking off a research project, the flagship content piece or whitepaper often receives the bulk of clients’ attention, while the main internal sponsor of the research may have the biggest say in deciding sample specifications.

This is entirely understandable but can sometimes lead to wider objectives of the research campaign being overlooked. For example, sponsors based in the UK or US may not be fully aware of how client-facing teams in other countries and regions want to utilise the data in their markets.

Will the research findings be presented purely on a regional basis or are there key individual markets where the findings will need to be segmented for local press and client engagement purposes? Would a different audience mix – such as a different profile of asset owner – make more sense in certain countries, given the structure of local markets? Coordinating with regional and/or local teams at the outset can iron out these issues before it’s too late.

Another important consideration is which organisation and investor types to include in the survey – whether institutional investors, gatekeepers, end investors or a combination. For an asset owner survey, for example, you may want to ensure the sample is a reasonable representation of the wider market to make it press-worthy. But there might also be important nuances between say insurers and pension funds on a given topic that demand a more granular approach.

Again, engaging different stakeholders such as PR teams and channel distribution heads early in the process can help ensure these factors are baked into sample design. It’s also important to remember that if the research audience is niche, for example chief investment officers within the institutional space, a minimum of 30 respondents can be sufficient to enable statistical analysis between specific respondent subsets. Allowing for a more granular analysis does not necessarily mean increasing your fieldwork budget.

2). Adopt an outward mindset

Another potential pitfall in sample selection is focusing too much on your firm’s internal approaches to segmenting audience types. These approaches might be based on client segmentations or the regional structure of your business, but they may not necessarily represent the wider industry.

There can also be a temptation to only include certain audience segments that your organisation already sells to without considering wider groups. An example would be producing a piece of research that seeks to be representative of institutional investors but excludes groups like insurers or single family offices because they are not a business priority. While it may still technically be an institutional piece, it will only voice a section of the market.

Similarly, within global studies, it can sometimes feel natural for firms to adopt regional groupings such as EMEA or the Americas that mirror how their own organisation structures its business units by geography. Depending on the research topic, and the extent to which regional analysis is going to feature, this may not always be an optimal approach. This is because market dynamics between, say, Europe and the Middle East or North America and Latin America can differ substantially.

3). Avoid positivity bias

When developing your research theme, it’s good practice to canvass the views of subject matter experts from across the firm as they can provide innovative ideas and market insights. However, it’s important to remember that while you may have selected a topic – such as ESG or private markets – which mirrors your firm’s strengths, your organisation’s perspective may differ from the market at large. Where differences do exist, such as lesser-known investor challenges or muted progress on a certain trend, uncovering these variances can often generate the most valuable insights from a study.

This is why it is crucial that the survey avoids positivity bias as much as possible. Positivity bias can arise both from improper sampling methodologies and the framing of questions in the survey itself. From a sampling perspective, being too prescriptive with screening of respondents is one issue to be aware of. For instance, requiring asset owners to allocate to private markets to participate in a survey on the topic might seem sensible. But in reality, doing so would narrow your lens on market behaviour and lose potentially valuable insights into the barriers facing non-adopters.

At CoreData, continuous engagement with our proprietary database of industry professionals – which covers hundreds of thousands of investors globally – provides us with a strong vantage point to understand the full breadth of industry sentiment across core research themes.

When it comes to survey design, there are multiple approaches to help mitigate positivity bias. One example is ensuring respondents are presented with contrasting questions and perhaps different routes through the survey to cater for those with opposing views or at different stages of adoption for a given trend.

Questions dealing with the barriers, challenges or difficulties respondents might encounter not only allow for a greater range of sentiment but also ensure research outputs can address both sides of the issue. Our recent asset owner research highlights how important this is to an institutional audience.

The wording and format of questions play a role too. For example, a question such as “Do you believe sustainable investing is essential for long-term returns?” could be re-written as “How do you assess the role of sustainable investing in contributing to long-term returns?” This can also be extended to map out a range of potential scenarios drawn from both positive and critical angles.

Being mindful of these issues in sampling and survey design will lead to better data – and ultimately better insights.

Daniel Zorilla Fernandez is a senior research consultant at CoreData Group, a global specialist financial services research and strategy consultancy. To find out more about our industry insights and research programmes, you can reach him at [email protected]