If there’s any constant in social media, it’s constant evolution. That means that to stay relevant and effective, marketers need to stay on their toes. Take, for example, Facebook’s recent decision to shift its feed algorithm to focus on actual audience engagement, or Google’s recent updates to its search algorithm, or the popularity of new social platforms like SnapChat, Instagram, and Vine. Today, I’d like to talk about the latest piece of the social marketing puzzle – predictive analytics.
Predictive analytics, simply put, use past data and statistics to model and predict the future. Among other things, they can allow brands to locate social media users with purchase intent. To understand how this piece completes the picture, let’s quickly review the evolution of social marketing technology:
Social Media Applications, Dashboards, and Platforms
Following Facebook’s open invite in 2006, organizations began to understand the potential impact of social media on business goals and strategies. As a result, companies like Wildfire and applications like Marketo Social Marketing emerged, providing services to help grow visibility of a brand through increased numbers in a brand’s followers, likes, etc.
Using these applications, Marketers soon realized that social was a place for conversations around their products and services – not a place for one-sided monologues about the virtues of your brand. This led to a shift from broadcasting marketing messages over social, to using social for listening and engagement. You couldn’t just monitor social media engagement; you needed to initiate and respond to conversations as well.
If you take a look at any sales advertisement, marketing campaign, or customer service interaction today, the chances are that social is involved. Today’s best-of-breed social media tools are part of of complete marketing ecosystems, allowing businesses to handle all of their enterprise social needs within one platform – from sales leads to customer service queries – closing the gap between departments and increasing efficiency and conversion rates.
Marketing’s Move to Predictive Social Analytics
Strategic companies now work to identify and participate in conversations on social – whether these conversations are specifically about your brand, or are simply relevant to your brand and/or audience. Marketers can use social to reach out and develop relationships, offer discounts and specials, resolve issues, make sales, and more.
But change is still the only constant in social media. Take the example of Facebook’s algorithm change. A few months ago, if a consumer “liked” a retail brand on Facebook in order to enter a contest, he or she would’ve started to see updates from the brand, possibly about upcoming promotions or sales. If the brand was lucky, that consumer might have made a purchase – so the contest might have directly led to revenue.
Facebook’s new algorithm prioritizes posts that have a higher percentage of engagement. As a result, users are more likely to see posts with high engagement rates within their personal networks (i.e. posts that your friends have engaged with), as opposed to seeing posts from brands they have liked. For marketers, this means a reduced likelihood of a conversion or sale.
This also means that relevant, personalized social engagement is more important to marketers than ever – and for many marketers, that’s where predictive social analytics come in. As I mentioned earlier, predictive analytics can help companies identify social media users with purchase intent.
Here’s a basic example: Let’s say James proposes to Ann. The next day, Ann shares the news with friends and family through her social networks. She’s likely to use hashtags like“#isaidyes” or “#justengaged”, or phrases like “I’m engaged”. Anne is now (or will soon be) in the market for goods and services like honeymoon destinations, wedding attire, stationary, etc. Predictive analytics tie the social behavior (using those hashtags/phrases) with the future buying behavior (purchasing a wedding dress).
If businesses offering these items can identify people like Ann, they can create targeted campaigns for segments (i.e. display ads for wedding dresses shown to everyone using those hashtags) or reach out directly on social channels to congratulate Ann and build a relationship. Obviously, this increases the chances of Ann choosing their business when it’s time to make a purchase.
When to Use Predictive Analytics
So how can predictive analytics can help you meet your marketing goals? Here’s a few examples:
Milestones: James and Ann’s upcoming wedding proposal is just one example of how predictive analytics can help marketers approach their audiences in a timely, relevant way. Moving days, graduations, pregnancies, and new purchases can all be ripe opportunities for marketers. A social search for the term “moving day”, for example, might help utility providers, alarm system installers, and truck rental companies pinpoint users who will need their products/services.
Product Launches: Social data can provide businesses with tons of information to use in launch plans. For instance, Deutsche Telekom-Hosted Business Services (DT-HBS) recently launched a product to provide Voice Over IP (VoIP) phone services to small businesses. DT-HBS was able to use predictive social analytics to locate users in need of VoIP services (by searching for “#VoIP” and “new office” on social), reach out those users, and to identify the influencers in the space.
Sentiment Analysis: Businesses count on social media to provide them with accurate, unfiltered thoughts and conversations around their brand, product, or service – as well as that of their competitors. Sentiment analysis uses sophisticated technology and natural language cues to sort social mentions as “positive”, “negative”, or “neutral”. This can help businesses discover consumer opinion trends about a particular product or service, leading to potentially invaluable insight.
Predictive Analytics Pitfalls
Interested? If predictive analytics sound like a promising addition to your social marketing strategy, here are two common pitfalls to avoid:
Cluttering Your Search Results: Often, less is more. This is especially true when attempting to narrow down or pull actionable results from social media. You might pull thousands (even millions) of mentions or conversations based on your keywords – but how will you know which are relevant? Build your search terms and keywords strategically to ensure you’re bringing in the mentions that matter.
Putting Social in a Silo: Sure, social media is a great conduit for pulling insights and ideas, but it’s not a stand-alone solution. In order to get the results and data you want, you need social media to work in tandem with other technologies and strategies. For example, if you’ve identified that your target audience loves cute animals, you might lead with a pet analogy in a sales email campaign. If your target audience enjoys the outdoors, your display ads might incorporate appealing outdoor activities. Don’t limit the results of your findings to social alone.
Tying Social into Marketing Automation
So how can marketers turn social data into tangible results and converted dollars? By combining predictive analytics with marketing automation, marketers can locate, target, and ultimately convert leads into paying customers.
Here’s a real-life example: mobile user acquisition network Appia recently released a product allowing developers to monetize their own apps. Prior to the launch, Appia’s marketing team applied data gleaned from social media research to their marketing automation campaigns, and they also reached out to my company, Viralheat. Appia’s team first used Viralheat to research competitor wins and fails, sentiment around the developer space, and what terms, ad types, etc. were resonating in their target conversation threads.
Using marketing automation, Appia’s marketing team was able to turn their social insights into targeted marketing programs. From welcome programs and site activation to nurture and support, Appia used the social insights from Viralheat and extensive marketing automation campaigns to produce and launch their product.
If you’re interested in learning more about predictive analytics, social marketing, and marketing automation, join Sally Lowery, VP of Marketing at Appia, and Jeff Revoy, CEO of Viralheat, as they detail Appia’s journey to success at our LaunchPoint webinar: “The Move to Predictive Analytics”. The webinar takes place tomorrow, Wednesday, April 30th at 10:00AM PDT.