What Do Clients Think and Do About Quality

Quality has become a regular subject of industry discussion, often framed as a technical concern for sample providers and research agencies. The client view is more direct. Poor-quality data threatens the credibility of the insights function and the business decisions it supports. Clients expect suppliers to manage quality, but they also recognise their own role in setting realistic requirements, funding the necessary controls and designing research that respondents can complete properly.

This was the central theme of a Global Data Quality (GDQ) session at Esomar Congress in Valencia on Thursday 3 September. Anita Watkins, CEO of the Insights Association, and Ray Poynter presented the first GDQ study of brand and client-side researchers. Vinay Ahuja of Procter & Gamble and Jennifer Huber of Campari, the drinks company, then brought the results to life from the buyer perspective.

A useful first view of client practice

The study surveyed 173 people in brand and client organisations across 29 countries, recruited through GDQ association partners and industry networks. It was a convenience sample, so it should not be treated as a representative census of all research buyers. It is nevertheless a valuable early view of an active group: 80% commission or run research at least monthly, 84% field at least some surveys in-house, and 77% work directly with agencies.

The findings do not support the idea that clients simply hand quality over to suppliers. On average, respondents used 4.9 quality checks. The most common were reviewing poor or nonsensical open-ended answers (71%), comparing results with expected or known findings (67%), and looking for contradictory answers (61%). Only nine of the 173 respondents said they relied on the provider alone.

But involvement does not always translate into confidence. Only 33% felt confident that poor-quality respondents are removed before data reaches them. That leaves a sizeable assurance gap, especially when research results support high-value commercial choices.

The commercial consequence

When asked to choose the main business risk from poor-quality data, 43% selected loss of trust in research and insights. A further 39% selected making the wrong business decision. Wasted budget and rework mattered far less. This makes sense: a weak study is inconvenient, but a confident recommendation built on a false signal can damage a team’s standing and alter a major decision.

Quality concerns also affect purchasing. Forty-five per cent had removed a supplier from future consideration, while 39% had paid more for stronger sample or controls. Yet 18% had taken none of the listed commercial actions. The challenge therefore has two sides: some buyers respond firmly to quality evidence, while others remain concerned but do not change the way they buy.

A quarter of the open comments spontaneously raised AI, bots, synthetic responses and related risks, despite no prompted question on this subject. This makes clear provenance increasingly important: buyers need to understand where respondents came from, how they were recruited, what controls were applied and what the provider removed.

What buyers want from providers

The requested assurances are practical rather than abstract. Sixty-two per cent wanted disclosure of sample source and respondent recruitment, and the same proportion wanted cleaning rules and the number removed. Sixty per cent wanted information about fraud and bot detection. By contrast, only 15% selected independent certification or audit among their top three assurances. Certification provides useful discipline, but clients also want evidence that relates to the specific project in front of them.

The briefing and contracting stage offers a clear opportunity. Fifty-seven per cent always or often specify minimum quality requirements, but only 39% have a right to review or reject poor-quality data. Fewer than one quarter specify quality metrics or benchmarks, or minimum bot and fraud detection standards. One in four have none of the listed contractual provisions. Clients and their agencies can make quality more visible by agreeing the requirements before fieldwork begins, rather than debating them after a problem emerges.

Vinay Ahuja made the buyer challenge explicit during the panel: “we get the data quality we are willing to pay for”. His point went beyond technical cleaning. The aim is to hear authentic voices from real people. Great research requires solid provenance and research designs that allow real participants to tell their real stories. Supplier qualification and ISO 20252 certification give Procter & Gamble an independently audited foundation. Research design, execution and processing then determine the quality of each project. A genuine participant will still give weak answers if a client pushes an unreasonably long or poorly designed questionnaire.

Jennifer Huber described the same responsibility from the perspective of a smaller buyer. Trust in agency partners matters, but clients should welcome challenge when internal stakeholders want to add more questions or compress the timetable. Transparent conversations about risk are much more useful than discovering issues late in the project.

Making quality visible

The GDQ study points to a straightforward agenda. Buyers should specify the quality information they need and use it in supplier selection. Providers should make provenance, cleaning rules, removals and fraud controls visible as standard, rather than burying them in technical appendices. Agencies should help clients protect respondent engagement through sensible questionnaire design. Professional bodies should turn their guidance into material that procurement and insights teams can use quickly.

GDQ can help close this assurance gap. Forty-five per cent of respondents wanted clear industry standards or benchmarks, 39% wanted standardised quality reporting and 35% wanted a practical checklist or toolkit. The study also found higher confidence among the small group who used the GDQ Pledge, although the base was too small for firm conclusions. The direction is encouraging: common language, visible reporting and practical guidance can give buyers a stronger basis for asking the right questions.

The message from Valencia was clear. The industry already has tools, standards and practical guidance that can address many quality risks. Quality improves when buyers, agencies and suppliers choose to use them throughout the research process. That choice carries a cost and protects the trust on which research teams depend.

Acknowledgements and recording

This study and session were made possible by GDQ and its association partners: MRS, the Association for Qualitative Research, the Canadian Research Insights Council, EPHMRA, Esomar, the Insights Association, the International Association for Measurement and Evaluation of Communication, Intellus Worldwide, QRCA, SampleCon, The Research Society, VMÖ, and the Global Research Business Network. Thank you also to Esomar for making time for this important session during Congress in Valencia.

The session recording is available through the GDQ webinar resources page.