24 Oct Social Intelligence Optimization (SIO): Deeper Social Media Analytics
How do you take a truly analytical approach to social media? It’s something many brands and agencies have thought about and struggled with since brands started using Facebook, LinkedIn and the other many platforms used to connect with customers and prospects online.
Initially, no one knew how to measure social media’s effectiveness, and it was often what we refer to as ‘vanity metrics’ like followers, impressions and reach that were monitored. These metrics are still relevant when a company’s goal is brand awareness building, particularly to monitor any shifts or trends, but they are backward-looking metrics that don’t give us any clear sense of how to optimize content going forward.
Social Media Goal Setting to Determine KPIs
One of the most important steps when laying out a social media strategy and action plan is setting clear objectives. Social media is great for building brand awareness and creating reciprocal engagement between your brand and prospects or customers. It can be also be effective for fostering positive sentiment and brand advocacy, and encouraging referrals from those who are connected with your brand. In some cases, and for some products and services, social media can help effectively move prospects further down the sales funnel by driving email capture, lead capture, website traffic and even purchase.
One thing we often see is a mismatch between goals and strategy. Some clients are actively pursuing an organic-only social media strategy, but are expecting to drive significant and immediate sales from those efforts. Others are using a heavily promotional strategy, with most posts linking to products on their site, and are expecting this to drive engagement and help build brand affinity.
Our approach to social is always strategy first. We need to determine the brand’s overarching business goals, and then define accompanying digital and social media goals with the client. We then look to complement that with a deep understand of their audience, the competitive landscape, their product or service, and the journey we expect customers to take to make a purchase. Understanding historical performance and tactics can also be helpful to get a sense of things that can be built upon and improved, as well as areas that need an entirely new approach. We also need to understand parameters like budget and desired timeline to achieve the established goals. We use all of these important factors, most of which we typically gather during our upfront diagnostic, to develop the holistic strategy and tactics that will best help us achieve these goals in the client’s specific context.
Working from this starting point, we can develop a strong social media program that is rooted in strategy with a set of Key Performance Indicators (KPIs) that we will focus on. We need to be able to measure the KPIs, so depending on the goal, we sometimes need to get creative, using technology or process solutions to gather the data we need. For example, particularly where the customer journey takes them from online to offline and sometimes back online again, we create an approach to capture this as accurately as possible. We then use historical data, competitive/benchmark data, and expected performance improvement from optimization to set targets for each KPI and layout the trajectory we expect to achieve over the first 12 months.
With clear goals, KPIs and targets, we can begin to develop the content and campaign strategy to drive to these objectives, and start measuring how that content and the account as a whole is performing over time.
Social Media Monitoring and Measurement
With our content and campaign strategy in full execution mode, we start looking at how our audience is responding. The typical approach here is to look at best and poorest performing social posts and to look for any perceived patterns. This is complemented by a look at overall channel metrics like reach, engagement rate, engagements (broken down by type, i.e., likes, shares, comments and clicks), social traffic to the website and, where trackable, sales.
We followed this typical approach for several years, but always felt there was an element of subjectivity, and that conclusions were drawn based on a very limited set of posts.
Monitoring KPIs at the account and platform level on a monthly basis is a great way to monitor overall progress towards our goals. But what happens if something changes? What happens if a particular metric is unexpectedly down during a particular month? Or if performance has drastically improved on one channel? How do you determine what drove that change?
Deeper Analysis through Social Intelligence Optimization
Enter SIO. It was with these questions in mind that we developed Social Intelligence Optimization (SIO). It’s a multi-dimensional analysis technology that allows us to analyze performance drivers based on many variables to determine which factors matter, on which platforms, and for which clients.
Performance here is defined by the social media KPIs that we set to tie back to brand objectives, but almost always also includes impressions and engagements. Organic impressions are a helpful indicator of how the channel algorithm is ‘perceiving’ and ‘deciding’ to show your content to followers. If organic impressions increase drastically during a particular month, period, or even on certain posts, we want to understand why. Was it something about the content of our posts that resulted in the platform serving that piece of content to a higher percentage of our followers? Was one post particularly ‘shareable,’ so much so that it was shared to a larger audience by our followers? Or was it something about the time of day or week that aligned particularly well with our audience’s preferences?
We have certain variables like post length, time of day and third party vs. owned links that we monitor across all accounts, while we have other factors that are specific or customized to a particular brand. For example, we might look at whether the post was part of a particular content series, whether a certain graphic illustration style was used, or whether a certain brand term or hashtag was used.
With all posts categorized according to each of these many variables, we begin to build a more useful and reliable dataset. This dataset continues to get larger and more useful over time, allowing us to develop a more accurate and predictive model for content performance, and also allowing us to pick up on larger trends or algorithm changes that may require us to revisit certain assumptions.
Social media is an ever-evolving landscape, and while we can’t always predict how platforms will change next, or what will be trending a month from now, our Social Intelligence Optimization tool closes the gap when it comes to analyzing content performance among our clients’ many unique audiences.