Research in the time of viruses- Wine Intelligence

Amid the economic and political turbulence caused by coronavirus around the world, we in the research industry should address two very pertinent questions: is it appropriate or valuable to conduct market research at this particular moment? And how reliable will the results be?

The latest edition of Vinitrac® Global, our survey of wine drinkers conducted in over 30 major markets around the world, has just gone into field (written March 17^th). Over the next few days thousands of wine consumers in key markets will answer an online survey about their usage and attitude to wine, to add to a tracking dataset that has been going in most markets for 12 years or more.

The context in which they will be answering their questions is unique in the history of the survey. Never before have our respondents been confined to their homes by government order, as those in France, Italy and Spain currently are, for example, whilst many other countries and states are adding to this list at the time of writing. Even in countries without such restrictions, bars and restaurants are often closed, and opportunities for socialising have been limited by government advice on ‘social distancing’.

For some of the questions in the survey, this context won’t matter as much, such as an individual’s preferences for certain packaging or product ideas, or recall of certain brands, regions and countries of origin. In other questions, we are expecting more of a contextual impact, though it’s not clear in which direction: how much wine you are drinking right now may go down because you are anxious about your health and the economy. Or it may go up under conditions of house seclusion.

Then there are those regular questions we ask about where you are buying and drinking wine, particularly the frequency of consumption in an on-premise location (bar, pub, restaurant) or your propensity to buy wine in travel retail. Here, we fully expect to see a significant decline in usage frequency in most markets, particularly strong in those with restrictions in place. In terms of spend levels, we just don’t know: people could be cautious about their money, or they could be splashing out on the rare occasion they make it to their favourite restaurant simply because such a trip has become so uncommon.

So yes, we may will see differences across a range of measures, but as researchers, our job is to measure and track behaviours as accurately as possible which includes during times of crisis. We did this during the global financial crisis of 2008-10 to enable our clients then, as now, to have an accurate and clear reflection of the reality of the marketplace and our consumers’ sentiment.

We also know that Coronavirus comes on top of a series of other challenging impacts to our category of wine – tariffs, taxes, some economic slowdown, bush fires, competition from other categories and more recently the oil price crisis). The world is changing for the foreseeable future, and we want to be able to share with our clients’ a real time and accurate view of the world and wine’s place in it.

For the benefit of our clients and other businesses puzzling over how best to deploy research efforts in the current environment, we have addressed a few frequently asked questions below. Feel free to reach out to me directly at lulie@wineintelligence.com if you would like any further guidance or views.

1. Can we still rely on the responses from online surveys collected during March 2020 as they are being collected in such unique times?

As suggested above, some of the standard questions we ask around brand and country / region recognition will most likely not be affected. Additionally, specific client questions about brand packaging, brand imagery and perceptions should also be answerable without any contextual bias. However we realise that standard questions around recalled usage frequency in on and off premise will be impacted. To help respondents distinguish between current times and more ‘typical’ times, we have updated our question set to include:

On-premise: We are now asking for frequency of consumption before Coronavirus impacts and also since any impacts
Off-premise: We are now asking for frequency of consumption before Coronavirus impacts and also since any impacts

2. What’s happening with response samples & rates for online surveys?

One of the benefits of conducting all our quantitative research using online interfaces (computers, tablets and smartphones) is that respondents can tell us about their preferences from the comfort of their home as well as elsewhere. We have been in touch with all of our partners in the research fieldwork world, who help us direct respondents to our surveys in Vinitrac®. They are reporting consistent, if not higher response rates from consumers currently – their hypothesis being that more people are home based currently than were previously, and they may have more time on their hands than usual. We use quota sampling (so we use quotas which are a combination of age, gender, state / area of residence, income and socio-economic group), meaning each survey is representative of wine drinkers in that market, so there is no change here.

3. What additional procedures are we putting in place to ensure the quality and validity of online responses to surveys:

As always, when we get responses back from our surveys, we first filter these through our quality control process. This includes a series of simple algorithms that check whether respondents have filled in their responses in a logical and consistent way and have taken a reasonable amount of time to do so. The software identifies ‘speeders’ (those who dashed through the questions too fast), ‘straight-liners’ (those who seemed to answer the same for most if not all things), and ‘random walkers’ (nonsense or illogical responses). Finally, we have some expert human eyes to interrogate outlying data points and eliminate those that seem fishy. All of these will be applied in this wave, and the fieldwork managers will be especially attuned to weeding out invalid responses.

We will than compare the data we collect with the historical dataset for each market. At Wine Intelligence, we have the advantage of holding data that, in some markets, goes back well over 10 years, so we can easily see if there are any spikes in a particular data collection wave and what those are, by looking at all of the key indicators of wine behaviour that we measure. For this March 2020 wave, we will put extra auditing and analysis in place at this stage as well.

4. Is it still ok to run all types and style of projects given the current world environment?

Our view is that the virus crisis is not affecting consumer memory and preferences, so studies like brand health, label tests and imagery statements are still appropriate studies in these circumstances

We may see changes in brand health numbers, but we’ll look at showing rankings as well as ratings if this is the case, so you can still see how brands are performing relative to each other.

However we are conscious that questions regarding intended behaviour / future plans are clearly impacted by the lens of the pandemic, so we are advising clients to avoid these, unless of course they require this understanding at a tracking level to determine what impact the crisis may have on future business. We have also included a series of additional questions about behaviour change as a result of the crisis, which we will feed into our syndicated research and form part of briefings to key clients over the next few weeks.

Author: Lulie Halstead