Spending on social media continues to soar, but measuring its impact remains a challenge for most companies. When The CMO Survey asked marketers how they show the impact of social media on their business, only 15% cited they have been able to prove the impact quantitatively. This low percentage is not completely surprising given that social media is a recent innovation that companies are quickly trying to understand and direct to the most profitable ends.
Additionally, The CMO Survey asked marketing leaders to report on the metrics they are using to track and analyze their social media activities (see Table). The most common metric is “hits/visits/page views” which represents the beginning of the funnel—awareness—but is not very diagnostic of purchase. The metrics that show the largest increases over time, are “engagement metrics,” such as number of friends/followers (+88%), net promoter score (+71%), buzz indicators (+54%), product/service ratings (+71%), and other types of text analysis such as sentiment analysis or keyword analysis on Twitter or anywhere customers post text about companies (+77%). Although abandoned shopping carts also increased, in general, we observe fewer companies using actual purchase activities or financial outcomes, such as profits or revenues, as metrics to evaluate their social media programs which was discussed in a prior post.
Table. Frequency of Social Media Metrics Used by Companies
Given these findings from The CMO Survey, we interviewed social media experts to better understand the challenges of demonstrating the impact of social media and the types of metrics used to do so. Here are eleven insights gained from these interviews.
- Use goal-driven metrics. Set specific goals for each social media campaign and then develop metrics based on those goals. If a social media campaign is designed to generate brand awareness, then engagement is an appropriate metric. However, if a social media campaign is intended to drive purchase, then the conversion rate from visitor to buyer might be a more suitable metric. This insight may seem obvious but there is often a disconnect between goals and metrics.
- Demonstrate metric validity. Metrics must be vetted to ensure they are valid—meaning they measure what they are designed to measure. For example, at what point should marketers classify a consumer’s interaction with a company on social media as “engagement”? Is it when a consumer likes or shares a post? Proving metrics requires linkages to key outcomes, customer interviews, and managerial judgment.
- Uncover and verify leading indicators. Social media engagement, measured by the number of page views, click-throughs, comments, shares, and likes, is often used as a leading indicator of downstream sales outcomes. Identifying and tracking such leading indicators is valuable as companies can gain an early sense of how well their strategies will pay off.
- Create dashboards. Most companies with a social media presence track metrics from multiple sources. As a result, it is helpful to create a social media dashboard that aggregates these different sources and shows a comprehensive view of the company’s or brand’s performance. A dashboard saves monitoring time and ensures that marketers have real-time access to how important metrics are trending.
- Develop meaningful benchmarks. Comparing results to meaningful benchmarks provides important context when assessing the impact of social media campaigns. Building a database of social media campaigns and their corresponding outcomes enables your company to develop these benchmarks. Your agency may also be able to provide a broader view of these benchmarks if they have access to a range of campaigns from various companies.
- Conduct experiments. To truly understand the impact of social media, companies must be willing to conduct experiments. Small experiments such as pre- and post-tests that measure consumer activity before and after a social media campaign are a useful way of assessing performance. Even better, include a control group that is matched on observable characteristics for comparison to the treatment group. For example, use geo-targeted social media in one city and compare results to a control-group city in which the campaign did not run.
- Allocate funds to measurement. According to The CMO Survey, companies spend only 2.3% of their marketing budgets on measuring ROI. Measuring the impact of social media requires investing in metrics. This investment might include dedicated staff, agency partnerships, tools and technology, models, or customer databases.
- Consider the cost of ignoring social media. One social media expert we interviewed offered the insight that the inability to perfectly measure social media’s return on investment (ROI) should not limit investments in it. Instead, he encourages his organization to also consider the Cost of Ignoring (COI) social media – “What is the cost to our business of ignoring this new platform?”
- Build predictive models. Metrics are often used to analyze what has happened, but they also can be used to predict what is likely to happen depending on the tactics employed, such as spending levels and media placement. To gain the most out of your metrics, leverage them to build predictive models and then plug in different inputs to simulate possible outcomes.
- Guide future actions. Measures should ideally be designed to offer developmental feedback. Ask yourself, if your social media campaign is not working, what information do you need to know in order to improve? Build your metrics or add additional metrics to capture this information so you know how to course correct and do better in the future.
- Stick with your metrics. Vendors are constantly developing new tools to measure the impact of social media. Nevertheless, marketers should focus on utilizing a handful of tracking tools that fit their goals and have passed important validity hurdles. Be careful not to flit between different metrics as doing so will hinder learning and waste valuable resources.
Although metrics are out there, debates about what these metrics mean and how they should be used can become both statistically and philosophically complex. At a recent conference, Tony Fagan, Google’s Director of Quantitative Research, noted that his staff was implementing propensity score matching in order to improve their ability to make causal inferences from observational data. His comment made it clear that marketers are not in Kansas anymore when it comes to social media measurement. We hope these insights offer a few signposts to marketers on this path.
source: forbes.com
