Product descriptions are among the most important texts in eCommerce. The quality of a product description contributes as much to sales success as prices, pictures, or positive customer feedback.
A good product description increases the likelihood of a purchase being made. But they have many other advantages for online retailers.
All online retailers know the importance of placement in Google search rankings. The more optimal the interaction between technical and content factors of a page, the higher the search engine places an entry in the result lists.
Mobile usability, loading time or internal linking are important criteria for the placement of a page in the search engine—structure and quality of content are also decisive factors. Texts such as product descriptions are so central to Google because the algorithms that compile rankings can read and evaluate text better than other information, such as images.
Keyword-optimised, well-written texts have a better chance of appearing at the top of an internet search. Google also takes into account the length of a text and users appreciate comprehensively-written product descriptions.
The better and longer a product description is, the longer a user stays on the site. There are two aspects to this: With increasing length of stay, the probability that an interaction will take place increases, such as a purchase or a subscription.
The amount of time that people spend on a website is also important for SEO. The longer a consumer stays on the site, the stronger the signal to Google that the content on that site is relevant. This reads to a long-term reward for the site against poorly-performing competitors.
There is no model for good product descriptions. Instead of following a template, it is advisable to consider certain aspects. Depending on the product, one factor or another may carry more weight.
Continuously creating product descriptions is a headache for online retailers. When a new collection comes into an online fashion shop, numerous product texts for jackets, trousers or pullovers have to be created and entered within a very short time. Updating existing products, such as the annual versions of software, also means that a product description must be adapted.
In many industries, the manufacturer provides a description text in addition to product images and structured data such as UVP, size or colour. The shop operator can make it easy for himself by adopting the manufacturer’s product descriptions word-for-word.
These ready-made texts and structured data can be automatically imported into shop systems such as Magento, Shopware or xt-Commerce. The advantage is that these descriptions go ‘live’ immediately without much investment in time or money. The disadvantage is that these descriptions are not the ‘unique content’ that Google gravitates towards, which significantly reduces the chances of SEO.
In practice, the manufacturer’s text appears on the websites of many different online merchants and marketplaces, including Amazon. Instead of being ‘unique content’, this is ‘duplicate content’. Not having unique content does not necessarily have to harm a retailer, but the use of product descriptions from the manufacturers will not have a positive effect under any circumstances.
It is not common for online retailers to employ people to produce content, generally farming this out to freelance authors or agencies. This work is charged using a variety of payment models, such as package prices for large numbers of texts or billing per word.
Leaving the quality of the text to one side, conventional copywriting often involves additional work apart from the fees. Coordination, briefing, proofreading, and integrating the descriptions is usually carried out by an employee.
Some of these steps can be eliminated by using text generation. The case for using a SaaS solution is intuitive. The basic form of such a solution is for gaps within templates to be filled with data. This data could encompass such things as brand, colour, size, condition, or price, and come from structured data.
This structured data is a prerequisite for automatic text generation. As a rule, this should come from the merchandise management system or from manufacturers. It is rendered as a CSV or XML file into the online system. From there, it can be uploaded, manually or through an API, to a generator. The system then automatically creates content using intelligent linguistic analysis.
The structure of the respective product description is defined by templates and conditions. Templates are essentially varied gap texts with a multitude of synonyms, adverbs and other lexicon entries. Conditions are the circumstances that must be fulfilled for the template to be used.
The process is done via a software engine that arranges the templates in a certain order. The given order in which the templates are put together is called a ‘story plot’ or ‘narrative’. Templates and conditions must be defined by human editors within the framework of an initial setup; the solution then works independently.
Depending on the orientation of the shop, the copywriters can create the product descriptions in different tones. It is possible to provide casual, emotional, or technical-sounding messages to customers.
Amazon, eBay, and other online retailers are a double-edged sword for online merchants: high reach and significant sales on the one hand, but competitive pressure and price war on the other. Only those who achieve a high level of visibility on platforms and marketplaces, perfectly presenting their own offers in text form, can survive in these channels in the long term.
Automatic text generation via a SaaS solution allows the production of optimized, unique texts per channel in a matter of seconds. At the push of a button, software creates multiple variations of a source text. Via an interface, one of these versions is imported into the system of the respective platform.
Special promotions and campaigns, especially on the occasion of the high-turnover ‘Black Friday’ and ‘Cyber Week’ on Amazon, can also be targeted more specifically with the help of NLG.
Online retailers have more issues outside of controlling their products within their own sales areas. Many retailers work across borders and consequently need to produce product descriptions in multiple languages.
A multilingual SaaS text generation solution is a powerful tool for a consistent brand presence. Automated translation between languages is important across Europe, meaning that systems should be able to produce in English, French, Italian, German, Dutch, and Spanish. Between certain language pairs, such and English and German, machine translation is now so accurate as to pass muster with an human editor.
If a generator is producing text and language in multiple languages, there is a significant savings potential for international companies. Translating interfaces such as Google Translate and DeepL are implemented in offerings such as textengine.io.