Getting Started with Brand Tracking

06.15.2023

When it comes to brand tracking, many organizations are likely to default to metrics like clickthrough rates and impressions to judge success. However, this data only relates to surface-level actions. Although strong performance in these metrics is generally symptomatic of strong brands, it may also result from any number of contributing factors.

Most importantly, surface-level metrics don’t offer insight into how people feel about your brand. In this case, seeing doesn’t equal (brand) believing.

Genuine brand tracking provides far deeper—and more actionable—insights. It evaluates the connections people form with brands, and what those mean to them, and informs decisions that deepen brand-audience relationships. But brand tracking also requires a perspective and process shift many organizations don’t know how to implement. So, where do you start?

The almighty dataset

The process for brand tracking centers around building the right dataset—extensive, bias-free and flexible—through quantitative surveys. That dataset enables brands to derive insight regarding what people feel, along with myriad other psychographics and relationship information. And, crucially, the right dataset allows stakeholders and leadership to derive insights or facilitate more granular segmentations that they may not prioritize at the time of the survey.

Ultimately, you can’t derive insight from information you don’t have. And if you didn’t collect it initially, there’s no means to do so until you conduct another round.

Strategic considerations for brand-tracking datasets—construction and frequency

Brands tracking their customer relationships must emphasize survey and sample construction along with survey frequency as major strategic considerations. They directly affect the gathered dataset’s potential insights, segmentations and for how long it accurately reflects audience sentiment.

Simply put, with the dataset existing as a “moment-in-time” snapshot,” brands need to know how long that photo must last and make its uses as flexible as possible for that period before they take it.

To that end, your brand-tracking dataset requires enough depth, breadth and overall quality to enable your various teams and stakeholders to adapt it to their changing needs until conducting the next round. This may mean oversampling—in total or for specific segments given more weight with respect to other strategic considerations (e.g., new market entry, launches or rebrands targeting new segments).

For example, perhaps an infant clothing company is considering a new campaign that targets expectant moms who’ve already raised at least one child. Before fielding the survey, it’s imperative that:

The sample includes that segment
The survey includes relevant questions to enable segmentation
The platform or analytical method supports distinguishing that segment’s other answers accordingly (while keeping individuals anonymous).

That’s the best method for the brand to ensure it has enough data from a statistically significant group to derive the insights it needs. And this approach must be conducted frequently enough to track performance over time (after establishing baselines) and to refresh the data so that it accurately reflects the landscape into which the campaign is launched.

Providing brands with datasets of that quality necessitates survey context and consistency. Context must inform sample groups, survey questions and intervals so that the dataset and the potential insights it offers remain specific to the brand. And because brand tracking over time relies on historical reference and the changes between each refresh, consistency enables long-term performance tracking and keeps analysis efficient.

Compiling your brand-tracking dataset

Broadly, the process for effective brand tracking follows seven steps.

Step 1: Conduct initial research.

Your initial research heavily informs brand-tracking surveys, helping you determine your strategy for data collection. Building your dataset with the most pertinent information requires understanding:

The right audiences to sample from (e.g., consumers, industry professionals, other companies)
Brand attributes those audiences consider important (e.g., yours, competitors)
The survey questions to include, based on “derived importance” (covered more below)

And with each brand-tracking refresh, this body of research increases its utility, enabling you to hone in on what audience, brand and survey characteristics will provide the most valuable data.

Step 2: Determine goals and your measurement framework.

After incorporating your research, it’s time to determine your goals and how you’ll measure success. For an initial brand-tracking survey, this mostly revolves around the insights you’re seeking to achieve. With following rounds, you’ll also want to begin refining or adding key performance indicators (KPIs)–which initial research should inform—and track their changes over time (e.g., brand awareness, buyer consideration).

Understanding your goals and KPIs is also an exercise in psychographic analysis—as they should also be informed by the attributes and motivations that make target audiences (and brand believers, in particular) most likely to purchase something.

Step 3: Determine the necessary sample size.

The significance of any brand-tracking insight is based on statistical differences found within your dataset. However, statistical significance requires your sample size to be large enough that you can attest that the information demonstrates a pattern rather than chance. A 10% difference among 500 participants carries more significance and insight than a 20% difference among 50 participants.

For many organizations, the minimum preferred sample size will be at least 200–250 survey participants. But this number increases depending on the information you’re looking to gather, how representative the sample is of target audiences and the segmentation flexibility you want to enable. Using a tool like a sample calculator can be useful in understanding how big your overall sample is and for your specific audience and subsets you want to learn about.

To ensure your dataset accommodates every segmentation-based insight you’re looking to gain, consider:

Oversampling—More participants will only reveal stronger statistical significance when it’s uncovered. When you want to dive deeper into comparing attributes for a specific audience, or cut your data into smaller segments, you should consider oversampling a specific audience so you can retain statistical significance and dig deeper.
Typing tools for pre-sampling—Enabling participants to pre-segment themselves along any desired criteria (e.g., brand preference, participant demographics and psychographics) will make working with the dataset easier and will uncover additional insights. This criteria can be based on initial research or business intelligence in order to segment your audience in the initial survey and establish necessary sample minimums for statistical significance.

Step 4: Build surveys based on “derived importance.”

An important rule for brand-tracking surveys is to generally avoid direct questions. They’re more likely to introduce bias by cluing participants into the information you’re seeking to gather, which can affect their responses. Similarly, a participant’s literal survey response to one question may not accurately represent how they feel, what they mean or how they interpreted the question.

Instead, you want to build your surveys based on derived importance. This involves tagging a series of questions and statements (i.e., “Agree or disagree?”) on the survey’s back end with their association to different needs, wants, values and other psychographic characteristics. By building your insights from an “aggregate of an aggregate,” you can more effectively filter out responses that might inadvertently create outliers or negatively affect the data.

Step 5: Evaluate for potential bias or leading questions.

All too often, organizations self-sabotage their dataset by including questions that introduce bias or inadvertently influence answers. For example, consider the difference between the following:

Rate the app’s user experience from 1 to 5, with 1 being “poor” and 5 being “excellent.”
How did you find the app’s intuitive user experience, with 1 being “extremely clunky” and 5 being “extremely intuitive”?

Take the time to comb through each question or statement and ensure you’ve removed all possible bias or other elements that can impact effectiveness (e.g., clarity, leading adjectives, consistent rating scales). Polluting your dataset with these can invalidate all the information and the insights you plan to derive from it.

Step 6: Field the survey.

Finally, it’s time to field the survey—either independently or in conjunction with a panel provider that will help ensure your recruitment meets certain thresholds (e.g., sample size and specifications)—particularly for target audience segments. However, if your brand chooses to field the survey independently, it will need access to a self-service platform like Qualtrics to perform this validation and set the survey’s length.

Regardless, analysis planning is crucial—specifically how the gathered data should be provided in tables (i.e., “data cuts”) and the banners and tabs to use. This can either be provided to a panel provider or completed via self-service in platforms like Qualtrics.

And remember to be considerate of participants’ time and exchange enough value (e.g., monetary compensation or rewards) to incentivize them to complete the fielded survey.

Step 7: Ensure data accessibility for utility and ongoing efforts.

Research and data analysis are worthless if no one uses them.

Because your brand-tracking dataset serves as an ongoing utility for decision-making, you need to ensure it remains easily accessible for all stakeholders, leadership and relevant roles. For example, brand-tracking data can (and should) help inform significant decisions like leadership’s annual decisions regarding strategic initiatives or creative considerations for mid-year campaigns targeting specific audience segments.

Data must be “nimble” and readily available, but that isn’t always the case depending on the platforms and partners brands rely on. Ensuring this accessibility may involve:

Dashboards so people can look at data from different perspectives on demand
Data architecture supporting exports (i.e., using brand-tracking data in other platforms and systems) and imports (i.e., combining data from other sources)
- Seamless data architecture will need to leverage APIs and may require custom-built connectors, but these efforts are essential to accessibility and ensuring derived insights are based on hard data.
Generating easily digested reports for stakeholders and other relevant parties

A centralized platform, system or database is also essential because it enables you to compare the most recent survey against historical data and monitor KPIs over time.

When to conduct brand tracking

As mentioned above, organizations benefit from ongoing brand tracking, with survey intervals strategically evaluated with respect to brand context and available resources. For ongoing monitoring, it’s generally about striking the right balance of insight and cost.

However, organizations expecting significant events on the horizon will also find value in ad hoc brand tracking to “test the waters” or refine their strategy. For example, brand tracking can prove invaluable prior to:

Brand redesigns or refreshes
Initial public offerings (IPOs)
New product or service launches (especially if targeting less familiar audiences)

Deeper insights from brand-tracking datasets

Campaign engagement and similar metrics are beneficial, but they simply don’t tell you enough about the connections audiences form with your brand and how those change over time. That deeper insight requires brand tracking, which—albeit challenging to get off the ground initially—any organization can perform by following the steps above.

Just remember to prioritize collecting the right data set for future segmentation, compare the latest surveys against historical data to monitor performance over time and ensure accessibility to make it easy for stakeholders to leverage the data.