Showing posts with label Data Science. Show all posts
Showing posts with label Data Science. Show all posts

Sunday, April 29, 2018

YOUR Customer Aquisition Social Media?


Customer acquisition on social media -- with your own data..

At a time when the use of third-party data is under increased scrutiny, Brian Handly touts the benefits of using your own.



In the battle for customer acquisition, data plays an important role in marketing strategy, along with a desired product and excellent creative.

There’s also the challenge of reaching a target audience where they spend most of their time, which today is within mobile apps and browsing social media.
When we look at Google, Facebook and Amazon from the perspective of an advertiser, we see that they utilize much more data for their own benefit than they make available for audience segmentation.

Amazon’s data has always been a walled garden. Their incredibly deep historical data on buying behaviors and patterns gives them a sizable advantage, leading to what many argue are cutthroat product decisions and incredibly targeted product recommendations.



I expect Facebook will increasingly become a walled garden after overexposing and ineffectively monitoring third-party data use. By shutting down their Partner Categories program, they’re reinforcing to their advertisers that Facebook audience data is the primary source for campaign segmentation.

How to cope in such an environment?

While numerous data sources are available for targeting across most digital properties, one of the most effective ways brands can target is by bringing their existing opted-in datasets to social media.

This frequently provides a competitive advantage over the “walled gardens” of the major technology players, as your own data is typically much more relevant to your marketing efforts.
The four major sites -- Facebook, Instagram, Twitter and Snapchat -- all provide advertisers the ability to create custom audiences using their own data, and in some cases to use third-party data sets.



The workflow is similar across all sites:

--Prepare your data.
--Upload it.
--The social media sites hash and de-identify the data.
--Your data is then matched to the social media site’s user base.
--Your custom audience is created.
--And your original data file is deleted.

Typically, the most utilized datasets to match against are email addresses, identifiers/tags provided by the social media sites themselves and mobile advertising IDs. Most sites require a minimum of 1,000 records in order to create a custom audience. This is for privacy reasons (to ensure data is aggregated and no individual could be identified), and to ensure that the segment is large enough to deliver appropriately.



The perks of using your own data

The ability to create custom audiences on social media allows advertisers to reframe many of their existing marketing tactics. They can encourage repeat visits, whether in-store or online, from existing customers, or try to win shoppers from competitive locations.

Brands without physical locations that seek to go directly to the consumer can use custom audiences to reach their market on social media as well. Most sites also allow advertisers to create “lookalike” audiences to help increase the scale of the campaign. They look for common characteristics from the audience you’ve uploaded and find similar consumers for your campaign to reach.

One final example of how you can use your own data is to drive mobile app acquisition. Building a custom audience from existing customers creates a segment with a much higher propensity to download and use a mobile app, especially when paired with appropriate incentives.



A key component of such strategies has always been, and will continue to be, ensuring that the datasets you’re using have opted in to marketing communication and advertising. Expect to see more transparency required on behalf of the end user, especially as the General Data Protection Regulation (GDPR) goes into effect next month.

Aside from being able to reach a relevant audience, bringing your own data to a social media site can also result in performance improvements and cost savings. The cost savings stem from being more relevant -- in Facebook parlance, this is having a higher relevance score -- which can result in lower cost-per-click fees because you can potentially win the auction for a given impression at a lower price.

Guest Authored By Brian Handly. Brian is CEO of Reveal Mobile, possesses more than 20 years of technical, operational and executive management experience, with 18 years of that in advertising technology. Brian was co-founder and CEO of Accipiter, which was acquired by Atlas in December of 2006 followed by the $6.1 billion acquisition of Atlas by Microsoft in 2007. Before their recent acquisition, Handly served on the Board of Directors for WebAssign, and currently serves as an Operating Partner for Frontier Capital. Brian also has extensive experience as an angel investor and is an active advisor for several North Carolina technology companies. Follow Brian on Twitter.





"Brands, advertisers and the agencies they work with have been hungry for the right data to help them reach the holy trinity of right time, right place, right person.

Using their own opted-in data sets will become an increasingly important tactic for the marketer’s overall customer acquisition strategy to achieve that goal.." -BrianHandly


    • Post Crafted By:
      Fred Hansen Pied Piper of Social Media Marketing at YourWorldBrand.com & CEO of Millennium 7 Publishing Co. in Loveland, Colorado. I work deep in the trenches of social media strategy, community management and trends.  My interests include; online business educator, social media marketing, new marketing technology, skiing, hunting, fishing and The Rolling Stones..-Not necessarily in that order ;)

    Monday, January 29, 2018

    Applying YOUR 2018 Marketing Data Science?


    Complex relationship patterns and groupings can become clearer through visualizations..

    Harvard Business Review famously described data science as the “Sexiest Job of the 21st Century” in 2012, causing a massive explosion in opportunities in this space.



    Today, data science has spread its hold over the digital marketing landscape.

    Particularly for social media marketing, data science promises a lot. From advanced analysis of social media activity on branded content campaigns to create insightful user personas via social media listening, to complex data patterns made easy to understand via visualizations, to overcoming the perennial problem of ad fraud in advertising ecosystems, data science has potential applications that significantly improve social media for brands.

    In this article, I’ll cover four ways in which brands can leverage data science for better social media marketing results in 2018.

    It's disappointingly common for people to use data science when they actually mean data analysis or analytics, and that's not exactly right. Data science is not even business intelligence. It's way broader in scope, and it involves exploration of multisource data to understand unseen underlying pattern that bring out important insights and relationships, which can be expressed through visualizations.



    Moving Beyond Word Clouds with Data-Science-Powered Tools

    Word clouds have been trusted tools for social media marketers to analyze social conversations and understand what's being discussed.

    Although brands could often stumble upon an important pattern, word clouds are, in reality, pretty blunt tools. Unless you have a high volume of activity, word clouds can be misrepresentative, requiring marketers to carefully guard against irrelevant words.

    Thankfully, marketers have access to tools that leverage the power of data science along with natural language processing algorithms in order to contextualize word usage and deliver meaningful insights.

    BuzzGraphs, for instance, show you how words are linked, and which words are most frequently used. Entity analysis also helps, associating words and small word groups with their semantic types, such as a brand, a person, a website, etc. Deep diving into BuzzGraphs and entity analysis is possible in order to gather more insight.



    Data Science for Community Groupings

    Social media marketing campaign results need to be measured and improved continually. Targeting strongly connected groups, naturally, amplifies campaign effectiveness.

    First, identify topic areas that receive good responses as a starting point for your community grouping campaign.

    Data science has tremendous applications here. Based on the frequency of keywords observed, marketers can identify the most commonly discussed topics in social conversations. The topics can then be analyzed across social platforms to classify them.

    In 2015, research journals published a lot of content about the use of machine learning in social media message classification. Today, marketers can use tools to execute the same.

    Next, leverage cluster analysis to identify how people participating, for instance, in a Twitter conversation are associated with each other. Such analysis can then group people together, separating weakly connected groups.



    Visualizations for Greater Insights

    Social media explosion has been one of the reasons why the volume of global data is surging every year. Each regular social media user's timeline is potentially the story of his or her life. Visualizations make it practical for marketers to understand these stories and generate insights that can massively improve social media marketing.

    Social graph visualizations, for instance, showcase the social dynamics playing out around us. SociLab, for example, lets you visualize your LinkedIn network and evaluate its “quality.”

    Complex relationship patterns and social groupings can become clearer than ever through visualizations. Data-science-powered social media tools help you by creating visualizations such as scatter plots to present correlations, pie charts to show proportions, line graphs to show trends and tables to show exact values. Hootsuite Analytics, for instance, can take your social media metrics and transform them into visualizations that make them much more insightful.



    Advanced Persona Research Backed by Social Media Listening

    Customer personas are much more effective than broad demographic descriptors. Personas are meant to humanize, although they’ve traditionally been filled with marketing jargon that eventually kills the effectiveness of targeting campaigns.

    Data-science-backed tools can transform how brands conduct market research using social media data. Social media listening platforms can allow marketers access to global conversations, bringing together large data volumes, capturing customer opinions and trends and feeding the data to a brand's specific market research campaign:

    --Begin with social media listening for researching a central topic.
    --From the general data, build maps of the most crucial consumer conversations.
    --Export the data to a spreadsheet and clean it.
    --Develop a listening dashboard to monitor discussions.
    --Study the natural language of the market and build it into your customer personas, helping copywriters create social content that converts more often.

    Guest Authored By Guy Sheetrit. Guy is CEO of digital marketing agency Over the Top SEO. Follow Guy on Twitter.





    "Data is being called the fuel of the present and future.

    Your social media analytics need to hit overdrive, powered by data science.

    Trust the methods explained in this guide to get started.." -GuySheetrit


      • Authored by:
        Fred Hansen Pied Piper of Social Media Marketing at YourWorldBr@nd.com & CEO of Millennium 7 Publishing Co. in Loveland, CO  where I work deep in the trenches of social media strategy, community management and trends.  My interests include; online business educator, social media marketing, new marketing technology, skiing, hunting, fishing and The Rolling Stones..-Not necessarily in that order ;)