Reliable brand tracking data is hard to come by. How do you prevent survey fraud? Reach the right audience? Ensure consistency over time? These aren’t just challenges—they’re the biggest obstacles standing between brands and trustworthy insights.

At Latana, Steph Clapham, our Director of Research, is redefining how brand tracking is done. By refining sampling methods and eliminating unreliable data, her team is pushing the industry forward—making brand insights more stable, precise, and truly representative.

Now, Steph is taking on the most pressing questions we get about improving brand tracking. Explore the full list below or submit your question to news@latana.com—we’ll make sure to answer it.

Brand tracking's most pressing questions

Q. What fundamental research challenges exist in traditional brand tracking?

A. The brand research space has been facing growing levels of fraud and data quality concerns, which have been exponentially heightened since the advancement of AI technology. Industry estimates suggest that up to 40% of completed survey responses could be coming from bots or fraudulent human responses. These concerns can undermine the reliability of the brand data gathered and can be incredibly frustrating. Most traditional approaches to sampling rely on recruiting respondents via incentivised panels, which can be a useful way to engage and motivate people to share their opinions and answer longer surveys. However, offering a reward is what attracts these bots or fraudsters into the surveys, and results in poor answer response and a compromised end dataset. The other issue with panel-based sampling is that panels are finite - they can only ever represent a portion of the population and often only represent a specific subset of those willing to sign up to these surveys. This means that sample sizes may be too small to properly represent an audience, especially when segmenting within the wider audience, and some markets are just unreachable to begin with. 

‍

Q. What limitations do brands face in setting up their trackers using traditional approaches?

A. Most brand tracking solutions will come with a certain degree of inflexibility - and this often stems from the survey design itself. To ensure the survey doesn’t become unreasonably long for even rewarded respondents to complete, brand questions, like awareness, will be asked in a list format - in which respondents are asked “Which of the following brands have you heard of?” and a long list of brand names is presented to them. This approach comes with some vital biasing implications - the types of brands that are listed can influence the likelihood of each being selected, and the number of brands in a list can also have an important impact on the likelihood of each brand being selected. Whenever a brand list is changed in the tracker - for example because a new brand is added or removed - the data for the other brands changes, thus causing a break in the data.

The other major limitation we often see from speaking to global brands is in setting up multi-market or highly complex trackers. Given that panels have limited reach, especially in smaller markets or for niche audiences, there is a likelihood that a single sample provider or sampling methodology won’t be able to provide sufficient access to respondents. If differing methodologies are required to collect the samples, for instance, online surveys in one approach and CATI or face-to-face interviews in another, there may be inconsistencies in quality or in comparability which means a global overview or market analysis is challenging.

‍

‍

Q. What does Latana do differently to address these challenges?

A. At Latana we wanted to find an alternative solution to panel-based research, to reach the 99% of consumers who are not signed up for them. We developed a survey sampling approach that uses interactive ad formats that allow people to answer survey questions directly within digital ads. Leveraging this ad network means we have access to anyone using smartphone apps worldwide, gathering hundreds of thousands of responses every day. This not only gives us vast global reach to gather much higher sample sizes than the industry standard, but it also provides us with a truly diverse audience to address common challenges such as demographic skews and access to difficult audiences. This approach is entirely opt-in and we do not provide an incentive to our respondents, eliminating any potential risk of fraud. Voluntary response means we need to be focused on engagement and have completely re-designed the survey experience, taking inspiration from the feedback mechanisms in e-commerce, ride-hailing and food delivery platforms.

‍

Q. What are the key statistical challenges in translating survey responses into accurate, scalable brand insights?

A. The finite pools of people that can be accessed via panel-based sampling approaches pose challenges to trackers that rely on large responses of new opinions with each wave of data collection. This means that sample sizes are often too small to provide statistically reliable results - this is even more heightened in smaller markets with lower populations. This also means that there will inevitably be repeat respondents in the sample over time, which can introduce a bias into trend data.

Panels will also often overrepresent certain demographic groups while underrepresenting others. This means that maintaining stable sample compositions can be challenging and can lead to larger margins of error, seemingly random fluctuations in the data, and poor reliability of insights.

‍Q. How does Latana ensure survey responses are accurate and reliable?

A. The use of ad networks to collect our samples means we can achieve remarkable market coverage, providing a scalable and inclusive solution for capturing consumer insights across demographics and geographies. We’ve developed a technology that enables us to automatically set quotas on individual answer options to ensure that we always have sufficient responses for every relevant brand metric and audience characteristic in the survey. This ensures we keep margins of error below 2% on all data points. 

The vast sampling framework we access ensures we never have repeat responses within a tracker, to ensure trend data is truly accurate. 

After we’ve collected our sample, we then employ a Bayesian statistical model called Multilevel Regression with Post-stratification (MRP) to estimate subgroup relationships and adjust them to match population demographics for better representativeness, to calculate the proportion of each possible audience combination within the overall population.

Finally, we use the estimates for each of these thousands of cells to weigh the results of the KPI estimates of subsequent waves accordingly. As the dataset grows over time, these estimates become more and more precise, and each subsequent wave is weighted according to those distributions.

‍

Q. How does Latana offer an alternative solution to allow for more flexibility?

A. We realise that trackers require flexibility - it’s impossible to avoid changes to competitor sets - and so we’re asking every question in a “siloed” format, presenting only one brand, image statement or other item at a time to respondents. This means we can completely avoid the problem of breaking trend data when a brand needs to be added or removed from a tracker. This means we can ensure no disruption to trend data for the brands that remain in the tracker, as well as no disruption to funnel metrics on those brands. We’re also able to include a large number of brands in our surveys, each with clear visuals, without the difficulty of presenting cumbersome lists to respondents that might flaw the quality of response we get from them.

We also overcome the challenges that setting up multi-market trackers can pose through the use of ad-based sampling. Ad-based sampling provides full control over the entire data generation process, from the exposure of the survey to the respondents to the data analysis. This means we never have to employ sample sub-contractors and can meet all sampling requirements in a unified and consistent methodology, for full comparability across the market and over time.

Submit your question to news@latana.com—we’ll make sure to answer it.

‍