To understand why natural language generation (NLG) is on the cusp of becoming the new standard in ecommerce merchandising, you need to understand the concept of natural language. In the simplest of terms, NLG is a form of Artificial Intelligence; a sub-field of computer science that draws raw data, interprets it, and then presents it back to us not in specs, statistics or graphs but in remarkably articulate human language.
While this in itself is extraordinary, it’s the high rate of speed and staggering volume of product data that NLG systems can process that holds tremendous potential for ecommerce merchandising. We started working on our NLG software (Kopigin: ginnie) with the goal of producing rich, enhanced product content at scale. The results have been impressive.
What Does NLG Mean for Ecommerce?
For ecommerce, the benefits of NLG are exponential. It’s the difference between onboarding 1,000 or 100,000 products in time for a critical retail period. In a highly competitive environment where brands continually compete for wallet share through richer customer experiences and channels that now include mobile, online, and social media, the ability to create and present consistent content can be make-or-break. While one would naturally expect this to impact the more visible touch-points such as branding, imagery and usability, it also trickles down to those all-important product page narratives that are intended to catch the consumer’s eye and drive conversion. The product story has become critical not only for search engines but for the buyer who’s try to determine if this “pair jeans” is perfect for them. The time and expense an organization must dedicate to producing creative content for literally thousands of items and then continually updating this content as new products launch or details changes is considerable.
This is where NLG technology holds its greatest potential. Its ability to quickly scale up to extraordinary amounts of content including catalogues that may contain millions of SKU numbers is remarkable. Once in the system, structured item spec data can be converted to high-quality, SEO rich content that delivers millions of unique product descriptions. The speed at which this information can be generated not only makes NLG substantially more cost effective than manual writing but it ensures the brand voice remains strong and consistent throughout.
Getting More From NLG
For NLG, the rapid production of unique content is just the beginning. This technology also offers enormous additional content production capabilities including:
• Development of themed landing pages. This is ideal for reaching niche markets and driving SEO rankings. Whether its BMX bikes with R7 alloy frame and press-fit BB-86, or sleeveless pink summer tank tops, the level of product attributes an NLG system can be designed to draw from and reproduce as reader- and SEO-friendly content is broad.
• Customization of individual product pages to highlight specific SKU information (such as price points, sizing, colors, special features and more). NLG not only delivers unique, creative content that speaks specifically to an item feature, but can also change the tone and style of writing for each product category.
• Generation of customized communication tailored to the specific purchasing patterns of the individual consumer. NLG can produce a conversational message that incorporates item details without the typical “product pitch” language. It doesn’t just recommend a product, it explains in simple language, why it would be a great fit for the consumer.
As the number and rate of products onboarded in the collective “ecommerce” ecosystem grows, the need for intelligent systems that can provide high quality and natural-language content will be even more pronounced. As content experts, here at geekspeak, we’re building and piloting NLG technologies to provide a fast, high quality, and budget-friendly solution to developing high volumes of ecommerce content for our customers. We can’t wait to see what the future holds.