If there were an award for the most common term at a conference anno 2023, it would undoubtedly go to 'ChatGPT'. Artificial intelligence and its associated AI tools continue to explode. While there is still an "it won't be that big of a deal" attitude, the impact will be of unprecedented magnitude. Are researchers - who by nature have a more critical and analytical view of the world - underestimating this disruptive turn of events?

In no time, AI and ChatGPT will impact the way research is conducted. They will facilitate the researcher's life, speed up (research) processes and increase agility. For now, few are using it in their daily jobs. But beware, even artificial chat robots are not always right if you choose to use the tool anyway! Ask ChatGPT literally the same question ten times. Probably a different answer will appear ten times. Moreover, the medium will never add that the answer is 100% correct, or even 80% (or less) reliable.

It is evident that the blend of new technologies, AI, surveys, ... is getting bigger and bigger. Yet humans as researchers remain - always - necessary. The tools complement existing research and will not simply replace it. See it mainly as a 'spark of creativity' in the innovation tunnel. For example, AI can contribute to data analysis, but the need for human inspiration and creativity to build hypotheses and find stories in data is inevitable. Human interaction with customers remains indispensable. The future role in which we generate impact for companies and brands by 'humanising the data' can only be achieved through storytelling and through creating visual impact.

How do companies choose an agency? And above all, how will they continue to approach research companies? It is and will always be a people business, there is no doubt about that.

First of all, word-of-mouth advertising is the best sales technique. Apart from good performance on client projects, one does not have to do much sales wise to achieve this. In addition, customers readily follow researchers when they change research firms, because they believe in the researcher's demonstrated approach.

But customers especially value agencies that don't just do what they ask and dare to challenge them. Find out the deeper need behind the client's question. Dare to be sharp and think along with the business as a whole. Taking up the challenge on your own initiative is highly appreciated and even creates valued continuity between clients and research agencies, resulting in long-term partnerships. Finally and obviously: be unique!

Data quality can be approached from two angles. On the one hand, we need to guard the quality of research panels. If even possible: improve them. Now, research should be protected from clickfarms and other IT bigwigs who go straight for the incentive through data intelligence. For instance, they engage in questionnaire completion only to receive rewards. Bots and speeders also distort data. The latter can be solved by adding one or more control questions during the survey.

On the other hand, biases arise in research due to questioning, social desirability, over-rationalisation in questionnaire design and drawing wrong conclusions from datasets. In other words: garbage in, garbage out. Especially at the beginning of the research, think more thoroughly about which questions to ask and the structure of the questionnaire.

There are still too direct and incorrect surveys, and that can have major consequences. For instance, the war Russia started in Ukraine was partly based on a poll showing that trust in the Ukrainian government was never that low and the people would never fight for the country if there was a war. This was questioned too directly, without any context, leading to dire consequences.