Turn Global Data into Insights in Your Command Center
Now that we’re in the fourth quarter, many communicators are evaluating their 2017 campaigns and making strategic plans for next year. Media and social analytics have a unique role to play in this 2018 planning process, and the establishment of a centralized, global analytics strategy for a dedicated command center is well worth your consideration.
The path to global measurement is similar to a U.S. program, but there are some major differences that require attention. Here is a five-step guide to help turn data into global insights while collaborating in your command center – relevant and actionable insights to shape and drive future strategy.
1. Choose Markets, Languages and Tools
First, review your markets and rank their order of importance based on your business goals. This focuses your analysis and insights on your priority regions.
Pin down the languages that are important in each of these priority markets. You might already have this information available, based on the translations used in different countries for your corporate website. If not, conduct a language analysis study. I did that for a client a couple of years ago and discovered that six languages were used in conversations about consumer electronics in India, but only one language was used in that country to discuss the technology industry. (As an additional benefit, this study resulted in client savings on website translation costs the following year.)
Find out which media channels are used most frequently in each priority market. If you’re not sure, conduct a channel analysis so you won’t waste time and effort on media that are not being consumed by your target audience. Analyzing your presence in the right media channels helps you understand which campaigns work, where they work best, why, and what to do next.
Be sure that your media and social analytics tool has the functionality to pull data in those markets, languages and channels. This is crucial because all tools do not provide all media sources in all countries and languages.
2. Select Metrics
Metrics should always connect to your business and communications goals. You might want to analyze brand reputation, customer loyalty, purchase interest, executive visibility, thought leadership or engagement. Each of these requires different metrics, which are the mainstay of a successful cross-functional command center. For more details, see my earlier post on metrics that matter.
3. Create Search Strings and Validate Data
Data is inherently messy, particularly when you are analyzing a large volume of mentions in different languages and countries. This leaves you open to error, so your collection process needs to be meticulous.
Carefully constructed search strings form the foundation for all analytics; they help eliminate noise, so your results represent the best information to mine for trends and insights.
The amount of search strings you build depends on your business goals, markets, languages and number of competitors. I have crafted ten search strings for a small company and dozens for a large corporation. Your strings should contain keywords for company names, messages, products, issues, industry trends, and topics that are most important to your organization. A best practice is to have an overarching search string, plus individual strings for specific topics.
Once you have an overarching string, build mirror ones for competitors in each country. These results help you to identify competitive tactics, messaging and positioning, which is good fodder for insights and recommendations for your future strategy.
After all your search strings are created, don’t just leave them static. Each week, validate and optimize the data by reviewing and cleaning the strings and adding exclusionary words, as needed, to remove irrelevant mentions.
Here are some examples:
- A client conducted a competitive analysis about new downtown office buildings, but the results included posts about arenas and stadiums, which were not of interest. We then updated the strings to omit stories about sports and concerts.
- Another client developed search strings about local governments in several regions. The results included police stories that were immaterial. We revised the strings to exclude crime-related posts.
- I worked with a different client to develop search strings in 20 markets. When the data was pulled, we discovered that the company’s stock market ticker was also an unrelated word in Spanish. Qualifiers were added to the strings to eliminate that coverage.
- A separate client wanted to find out how many times the word “innovative” was associated with their brands. Our search turned up massive results because the word appeared in the boilerplate of company press releases, which were published by thousands of online distribution sites. Once we excluded press releases (and associated terms) from our strings, the results were considerably smaller but more accurate and realistic to inform future strategy.
4. Translate Keywords
After your search strings are built, translate them for each language you’ve chosen in your priority markets. The best practice is to have a native speaker in each language translate the keywords; your analysts in different countries also have the advantage of being able to interpret the results from a local perspective, with an understanding of the resident issues and cultures.
If you don’t have native speakers in your global command center, translation services can be outsourced to agencies or research companies that have an international presence. Or perhaps your analytics tool provides translation.
Free online translation tools are also available. However, should you take this route, use more than one tool for accuracy. I have seen automatic translations that provided out-of-context information that made no sense.
The best method to find trends and insights is to sort and visualize the data in different ways. Create a mentions chart so you can review spikes and dips in the conversations, and determine what caused the chatter during a specific point in time. Check to see if external factors, such as breaking news, played a role. Did a competitor have an announcement, a product launch or host a conference? Did weekends, holidays or the weather factor in?
Create a word, hashtag and/or emoji cloud to investigate the trending topics and themes in your industry. Set up filters so you can drill into the details that are influencing brand reputation, and also understand attributes such as sentiment, top authors, media-channel breakdown, competitive share of voice, influencers, and heat maps to discover the point of origin of a story, how fast it evolved and why.
In other words, different visualizations of the data give you contextual awareness and the reasons behind the numbers. Ask yourself what happened, why it happened, if it matters, and what to do about it. Look for successes and missed opportunities in relation to your business and communications goals. Evaluate global trends that are generating attention.
This process will unveil ways to develop data-driven recommendations for future action.
If done properly, a global analytics program, led by a cross-functional team in a dedicated command center, can help you to: formulate campaign tactics; listen more closely to the market and your customers; benchmark competitive positioning; refine messages; pinpoint trends in consumer opinions; drive thought leadership; power your content creation; inspire compelling and creative new strategies and tactics; and ultimately, make better business decisions.
My first blog post about command centers gave you a step-by-step guide to set one up. My second post focused on creating a command center for the C-Suite. The third post looked more closely at how to create consensus and support among different groups in your company.
Margot Sinclair Savell is an award-winning writer who has decades of experience crafting and editing content. During 15 years at agencies such as Hill+Knowlton Strategies and Weber Shandwick, she specialized in strategic counsel for measurement, insights and analytics. In 2016, she was inducted into the PR Measurement Hall of Fame.