Brand health measurement is vital, yet so many businesses avoid it or resort to using basic KPIs. They end up with findings that are not agile or actionable, and sometimes these results are so general that they could be misinterpreted.
In this guide, we’ll talk you through how to develop an approach to brand health that can instigate meaningful change within your organization. We’ll also provide detailed how-to guides on:
- Incorporating competitor intelligence
- Measuring brand health across different audiences
- Blending digital datasets to generate better insights
- Setting up appropriate alert systems
Throughout, you’ll hear from Brandwatch Senior Solution Strategist Marsha Robie. She works with some of our largest clients, setting them up to work faster and smarter with the help of digital consumer intelligence. Having worked on brand health projects with businesses across a wide range of sectors (including travel, pharmaceuticals, financial services, retail, and more), she’s got a wealth of experience and expertise in best practices.
“Assessing brand health can be tricky. It takes some finesse to customize for every client. But if you do it right, it can become critically important within the organization,” she says.
But this trickiness means businesses can procrastinate in setting up helpful brand health measurement systems.
“Because a good brand health measurement system will include a number of brand attributes and could require some fine tuning, developing one can seem daunting. Many will say ‘I’ll get to it, next quarter or next year.’ The can keeps getting kicked down the road.”
Let’s explore how Marsha helps our clients to tackle brand health, leverage the insights, and stop kicking the metaphorical can down the road.
How to approach brand health
Starting with objectives
It might seem obvious, but starting with high-level business objectives is the best place to begin. From there, strategies can be built out that ladder up to company goals and generate actionable insights that make a difference.
“When deciding how to measure brand health, you start by considering your company’s and your brand’s overall objectives – all the way up to the company’s mission statement,” says Marsha. “We want to ensure we always, always have an eye on those goals because, when approaching something like brand health, it is really easy to get caught up in discussion of KPIs and metrics and forget to consider what you are trying to achieve at a broader level.”
Marsha will start by asking clients how they’re currently measuring brand health. Often the answer will relate to simple metrics like net sentiment or share of voice, which are both red flags.
“I’ll say, ‘Let’s look at your objectives. Let’s look at your mission statement. Let’s talk about your industry. Who are the key audiences you’re trying to reach? What are your overall goals? What outcomes are you trying to achieve?’ Then we can work our way down to a level of granularity that helps them convey actionable brand health information to leadership.”
Marsha has been surprised by the number of teams she’s met at large enterprises who get bogged down in the metrics before thinking about the wider implications of their work. Often, leadership teams will ignore what social or marketing teams are telling them because the metrics (like simple sentiment analysis or social share of voice) just don’t include the nuance that’s needed for them to take action.
Sentiment or share of voice analysis might not sound like a bad place to start when thinking about brand health, but it’s important to note the potential drawbacks.
- There’s no context in share of voice numbers. If Business A has 30% more online mentions than Business B, what does that actually mean? What’s driving those mentions? More isn’t always better.
- There’s no nuance in sentiment. “90% of conversation around our brand is positive!” That might sound great, but what are the mentions about? Sentiment analysis is all about measuring the tone of a piece of text. The positive driver of those mentions might have nothing to do with your brand, sector, or industry.
Both of these metrics are great ways of discovering topics, issues, and opportunities that can be further drilled into, but neither should be seen as a golden metric, especially in their out-of-the-box form.
Go deeper
Going beyond these basic metrics can make measuring brand health intimidating – it’s perhaps more comfortable to focus on rudimentary analysis than to start with the larger problems.
Marsha is skeptical of out-of-the-box brand health measurement tools for a few reasons. Not only do they miss out on focusing on what’s really important to the business, they can also skip over important differences that might come with a particular sector or category.
“If anyone thinks there’s a one-size-fits-all approach to brand health measurement, they’re kidding themselves. There is some fundamental consistency in how we approach brand health measurement programs, but we can be very customized and build out measurement according to what‘s important to a client’s industry and their specific customers’ needs.”
“I think clients who are looking at that more granular level to measure brand health are on the right path,” says Marsha.
In the ‘how to’ section of this guide, we’ll take you through an example of how this can work in relation to retail brands using machine learning to monitor shopping experiences.
The importance of a yardstick
Brand health measurement is impossible without something to check performance against.
There are three key ways to make this comparison:
- Measuring against your own previous performance
- Measuring against your competitors
- Measuring against ‘best in breed’ brands (those you might not compete against but admire)
“In any brand health measurement, I would always include some kind of competitive yardstick. It could be direct competitors or I think it can also be very valuable to compare against best-in-class brands. Regardless, you have to have a yardstick to help you define what good looks like and what bad looks like.”
Whatever your approach, keeping your methodology consistent over time is key.
Action is the priority
Like with any research, the most important thing is what’s done with the findings – how do they effect change?
“The ability to take action is key,” says Marsha. And with brand health measurement, that can mean a lot of different things, from uncovering white space opportunities that weren’t previously known about or vulnerabilities where course correction is needed.
For many people conducting brand health measurement, there will be various key internal stakeholder groups that can take the insights and act upon them.
While there is a lot of nuance to brand health measurement, socializing insights needs to be a simple process, especially for time-poor leaders.
“It’s critical to set up a brand health measurement program that allows you to take a quick pulse check of your brand’s vital signs at a frequent cadence, and then allow for a more comprehensive, deep dive examination of brand health at longer intervals.”
This simplicity at the point of delivery is something that Marsha helps clients work on from the start. By making sure everything ladders back to key business objectives, and by creating a measurement set up that allows for quick analysis at different cadences, it’s all about helping the client work faster and smarter.
In practice: How-to guides for common use cases
Now you know how to start thinking about brand health measurement, we can get into the ways you can do your best work with digital consumer intelligence solutions.
How to customize brand health measurement
As Marsha mentioned above, there is no one-size-fits-all approach to brand health measurement. And, since all measurement activity should ladder up to key organizational objectives, the way you define brand health will be different depending on your brand, category, or sector (and beyond).
Marsha gives the example of a retailer: “One of your functional objectives might be to make the shopping experience easier for your customers. So something you might want to measure is how are people talking about their in-store experiences, segmenting that discussion by the most important customer experience categories. How about your online commerce site? How are they talking about about ease or difficulty of experience locating items using the app? How seamless or difficult is the checkout process? Are your customers complaining or complimenting your payment options? This type of granularity is going to lead to actionable insights; standard SOV will not.”
Finding out what people are saying about shopping experiences is easy with the right tool. For example, using Brandwatch Consumer Research we collected online mentions of The Fresh Market (a North Carolina-based grocer with stores in 22 states).
By searching within these mentions, we can easily find out how people are talking about their experiences shopping with the brand. But we want to break that down further to see how they’re talking about shopping experiences both online and in-store.
To do this, we use Custom Classifiers. These can be trained to recognize that a mention is about a particular topic. In this example, we’re training Custom Classifiers to recognize when someone is talking about going into a store versus talking about ordering online, by dragging and dropping relevant mentions into the different category buckets. Classifiers can be trained with as few as 10 mentions per category, but we’d recommend training with at least 25 posts.
Then conversations about in-store and online experiences can be analyzed separately, revealing potential opportunities or vulnerabilities in each setting that the company can address.
One of the coolest things about Custom Classifiers is that they can be applied across various queries. This we can train them to recognize the difference between in-store and online conversation about our own brand and also about our competitors, all at once.
From here, we can analyze our own performance over time as well as how it stacks up against competitor performance. For example, we could see how we compare to competitors when it comes to angry conversation about in-store experiences.
How to measure brand health across different audiences
“Something we can do so much more quickly with our customers now is slice and dice by audiences,” says Marsha, referring to our newly launched Social Panels feature.
This allows brands to search for particular groups of people based on criteria like gender, location, interest, profession, or bio keywords to analyze their conversations.
Here’s a simple example:
We searched for people who are interested in movies – for example, those with ‘movie buff’ or ‘film lover’ in their bio. By applying this panel to queries around different movie theaters, we can start to build an understanding of which theaters self-confessed movie lovers enjoy visiting most, and what attributes are most important to them – is it snacks? Price? Seat comfort?
Below, we split out the conversation from movie lovers around two competing movie theater brands.
An analysis of key attributes shows that movie lovers prefer Theater Chain 2. If it upgrades it’s seating, Chain 2 could widen its competitive advantage among movie fans.
Theater Chain 1
Theater Chain 2
“The Social Panels feature is just a complete game changer,” says Marsha. “It allows our clients to take the pulse of their different consumer groups in near real time. And I think that’s particularly important right now.”
How to blend digital datasets for better brand health insights
“I think, as with all things, brand health measurement is best when it includes many different digital data streams,” says Marsha. “I would highly recommend our clients consider integrating owned data, data from primary research, search data, and other publicly available data streams.”
With Data Uploads you can analyze any text-based data you own directly in Brandwatch Consumer Research such as online or offline surveys, chat and call logs, reviews or community forums. This means that multiple sources of brand health data can be analyzed alongside each other in the platform.
So much data is siloed across an organization – by connecting with different departments and bringing datasets together, analysts can get a 360 view of brand health.
Here’s an example of how it works with some publicly available data from San Francisco International Airport’s Screening Checkpoint Satisfaction surveys, released this year.
Once uploaded to the platform, the data can be analyzed in many different ways.
For example, comparing the topic components for low (1,2,3) and high (4,5) rated responses gives a quick picture of why people had a good or bad experience by pulling out key topics in their text responses. Those with bad experiences generally referenced long waiting times.
The data can also be compared to social media conversation coming from the airport (using a query that’s geofenced) or reviews that come in later once people have departed. In some cases, there might be a difference between the two datasets, which can be a good thing (this means that you’d be missing out on important feedback without that data source). If the themes are similar, the two datasets can help validate the need to prioritize fixes for particular issues.
“I really encourage our clients to think broader, think bigger,” says Marsha on how clients can combine the data at their fingertips to get to deeper insights.
How to set up appropriate brand health alert systems
While the methodology must be kept consistent, brand health measurement shouldn’t be a ‘set it and forget it’ activity. Not only should it be checked at appropriate cadences, but changes in the data should be monitored continually.
Marsha has a great example of a client who initially had a very narrow view of brand health.
How might that work in practice? It’s pretty easy with Brandwatch Consumer Research’s Alerts and Signals.
Alerts can be set up so that particular incidents online will immediately spark notifications to key stakeholders. This could be a brand mention from a journalist, or a post containing a particular keyword or phrase. Alerts are a very flexible and highly customizable tool.
Signals are much like Alerts, but will monitor online conversations for things that a brand might not even be aware of. For example, it will notify stakeholders when there is a change in the conversation like a spike in negative sentiment or around a particular topic.
Both Alerts and Signals help with continuous brand health monitoring, enabling brands to react quickly to opportunities and crises without anyone needing to be logged into the platform at the time.