How Auto‑Tagging Helps Retailers and Publishers Get More Value From Marigold Engage by Sailthru, Faster
February 29, 2016
When a client fires up Marigold Engage by Sailthru, we want them to get value out of our customer retention solution right away. We’ve built this whole company around personalization as a path to happier, more-effective customer relationships, and nothing is more gratifying to us than to see it produce measurable lift for our clients.
But for marketers who have made the decision to move from static marketing to dynamic, customer-driven messaging, the work needed to develop and implement a content tagging strategy can slow implementation. That’s right, tagging. The seemingly innocuous but time-consuming task of attaching keyword descriptions and tags that connote audience interest in a web page, so that software knows instantly what that particular content is actually about and can serve it to the right customer at the right time.
Sailthru now has an answer: Auto-tagging.
Why Tagging is Crucial to Personalization
When we’re talking about content, we mean it in the broadest sense: For a publisher, it could be anything that appears on their site. For a retailer, that might mean a product description, the increasingly creative and persuasive copy we’re seeing from retailers pursuing a dual commerce/content strategy, or anything in between. In all cases, for client to achieve results like Business Insider’s 200% lift in pageviews and Country Outfitter’s 72% lift in revenue, accurate profiles must serve as the foundation. Tags are critical to this.
Without tags, or with lousy ones, the ability to understand consumer interests diminishes, and delivering true email, web, mobile and cross-channel personalization becomes a challenge.
How Auto-Tagging Works
Auto-tagging involves very little work on the client’s part because, you guessed it, it’s automated. The tagging gets done in real time, as customers make their way through a site. When a customer pulls up a web page, Marigold Engage by Sailthru performs a textual analysis of the page and generates three sets of tags:
Keywords: The important words that we discover on a document as written. We rank them by relevance so only the best keywords are included (up to the maximum number you specify).
Taxonomy: Topic categories, matched from a predefined, Marigold Engage by Sailthru-provided database.
Concept: Brief descriptions of the content. Concepts might not include any of the text on the page but will identify what that content is about. For example an article discussing the President’s reelection chances may not mention the word “politics,” but “presidential-politics” may be identified as a concept tag.
Clients can choose how many of each type of tag they’d like based on what’s most relevant to their specific content. To ensure fit and accuracy, we help our customers run a test on a segment of URLs to make sure the results are hitting the right balance.
Tagging, Made Easy
Manual tagging that is done well can outperform auto-tagging. But for a large media site with hundreds of thousands of articles in their content library, manual tagging is a significant project requiring valuable time and resources. And even after tagging is “finished,” editors have to be make sure new content is tagged as it’s created.
For clients who want to streamline and scale the utilization of untagged pages, auto-tagging provides a simple, easy-to-implement solution. It lets Sailthru start building interest profiles immediately, greatly shortening the time-to-value for new clients, and allowing them to begin on the path toward more meaningful consumer interactions that create retention.
Are you a current Sailthru customer who wants to learn more about Auto-tagging? Get in touch with your Customer Success Manager or the Sailthru Support team. If you’re not a Sailthru customer but want to learn more, please contact our Sales team.
–David Studinski, Director of Product Management at Saitlhru
The State of Brand Loyalty in the U.S. in 2023
4 Signs That It’s Time to Upgrade Your Messaging Platform
Is it possible to scale exponentially while getting more personalized with each and every customer? It is with machine learning and the power of...
Data & AI
Thanksgiving Week Shows Retailers Collecting Zero-Party Data While Loyalty Programs Continue To Provide Solid Revenue Streams
The 2022 holiday shopping season has begun and forward-thinking brands are seizing the opportunity to collect zero-party data via interactive experiences in the busiest...