How brand manufacturers get more sales from Amazon – eCommerce Magazin

In e-commerce, large retailers and online marketplaces such as Amazon, Ebay, Zalando and specialized platforms such as AboutYou lead the field in the use of data-driven technologies. Customers receive purchase recommendations tailored to their interests. Prices for products from brand manufacturers are continuously adjusted to the competition and Product descriptions are often automatically internationalized. The manufacturers receive digitally generated purchase orders (POs), i.e. reorders of products. These often contain incomprehensible quantities and come in atypical frequencies, as they are based on forecasts, which are not transparent for the recipient, of how many products the AI ​​from Amazon and Co. believes it can sell. Systems for prescriptive analytics can help here.

Prescriptive analytics rarely used by brand manufacturers

So the dealers have upgraded the AI. Nevertheless, the use of AI in the everyday life of brand manufacturers is still in its infancy. A current survey by the digital association Bitkom among 600 companies shows that the proportion of companies in which the use of AI is discussed has jumped from nine percent in 2019 to 30 percent in 2021. At the same time, however, only eight percent of companies actually use AI-based technologies such as prescriptive analytics (the proportion has quadrupled since 2019). The value of AI is recognized, but the technology of the future has not yet arrived in practice.

Optimize report and planning processes with AI-supported software

There are many areas in which AI applications can support brand manufacturers along the entire value chain. One of the most common fields of application today is the display of personalized advertising, mostly using behavioral data from the aforementioned online shops. Amazon has developed into the market leader in the new form of advertising digital retail media. The online giant competes with its sponsored ads in search and programmatic advertising using its own DSP (Demand Side Platform) on an equal footing with Google and Facebook for the advertisers’ budgets.

AI is mainly used in process optimization and customer care

However, the artificial intelligence is still on the part of Amazon and brand decision-makers only use the retailer’s algorithms. According to Bitkom, companies use AI in the process optimization of production and in customer care, such as through automatic chat and e-mail systems. The use of AI-supported reporting and analysis software, which allows demand and sales forecasts, as well as scenario analyzes with a view to the return on ad spend, is only available in a few companies.

Such advanced analytics, i.e. systems that try to derive future business development, can represent significant added value. The management consultancy BCG has analyzed that brand manufacturers achieve at least ten percent more sales if they make AI-based forecasts (predictive analytics, i.e. “What will happen?”) And suggestions for action (prescriptive analytics, i.e. “what do I have to do to make it happen”) ) in your own business planning as well as in marketing and sales.

Elaborate and complex: the reports from Amazon

Sufficiently reliable data is crucial for this. And here is a problem that comes to light especially when dealing with online retailers. Retailers like Amazon do not even share a lot of relevant data and the data that is available are not easily suitable for AI-based predictions in strategy and investment planning. Amazon makes helpful reports available to brand manufacturers, for example on the margin, the inventory or the performance of the advertising campaigns, but these are limited to the recent past and are also specified in the data cut-off (daily, monthly, quarterly).

In addition, the access to the reports is distributed in different tools, most of the data can be manually downloaded as Excel files and saved again. Nevertheless, the effort is worth it. Just the preparation of the historical data that Amazon makes available can show exactly what is happening in your own business, for example that sales are falling (descriptive analytics) and why that happened, for example because the advertising campaigns have ended (diagnostic analytics) .

Prescriptive Analytics emax

The analysts approaches descriptive analytics and diagnostic analytics in comparison. (Graphic: emax digital)

Brand manufacturers should make reports available

Every brand manufacturer who trades with Amazon and other platforms should make the effort to consistently download the reports provided and make them available centrally. The head of e-commerce does not have to take on the tedious downloading process, but the junior can do it as a standard task. In the simplest case, a proper data model in Excel or, one step further, business intelligence software such as Power BI is sufficient for visualization.

Deep data is required for prescriptive analytics

For AI-based predictive and prescriptive analytics, a significantly higher effort is necessary, which many companies shy away from due to a lack of money, the right staff and time, as the Bitkom study confirms. It is not enough just to collect data and produce big data; companies have to upgrade their own data to so-called deep data. This requires appropriate IT expertise, such as data scientists and software developers, as well as very high data quality. Only then can an algorithm calculate how much sales a manufacturer is likely to generate in the next few weeks. For example, they can recommend investing 5,000 euros more in advertising in order to increase their own product visibility on Amazon by ten percent and generate between 13 and 15 percent more sales.

Prescriptive Analytics
The analysts approaches predictive analytics and prescriptive analytics in comparison. (Graphic: emax digital)

Manufacturers must supplement the historical report data with additional information on when sales events such as Prime Day took place, how demand on Amazon developed and which competitive products are displayed next to mine in the search. In addition, all data must be prepared and stored in secure database systems in order to be able to provide the results of the AI-based calculations reliably and quickly via various interfaces. The introduction of AI-based forecasts and recommendations for action is therefore a huge effort. Compared to the management consultancy Accenture, 82 percent of German top managers admitted that their AI projects would not get beyond the pilot phase.

Prescriptive Analytics emax digital
The founders of emax digital GmbH: Dominik Pietrowski, Andreas Kleofas and Dimitri Dumonet (from left to right). (Image: emax digital)

Open the Amazon Black Box with Prescriptive Analytics

Specialized service providers like emax digital This is exactly where they start and enable the use of AI-based advanced analytics on Amazon without having to set up your own AI department. Thanks to the Amazon Analytics Hub developed by emax digital, manufacturers can plan and control their business more precisely with the support of AI. The web-based tool also gives recommendations for action to increase sales and optimize advertising investments. Customers benefit from the company’s many years of expertise in e-commerce and the concentrated Amazon know-how. Two of the three founders are former executives of the world’s largest online retailer, while the third founder already had his own start-up in the e-commerce environment. You finally want to open the Amazon black box to others – thanks to prescriptive analytics.

About the author: Andreas Kleofas is co-founder and managing director of the emax digital GmbH, an e-commerce analytics hub based in Munich. The 39-year-old is responsible for product development there. Kleofas has more than eleven years of experience in the media industry, the last eight of which at Amazon Advertising. At the shopping platform Amazon he was in charge of account management and was responsible for the development of analysis tools for large advertising customers in a wide variety of industries.
After completing his bachelor’s degree in popular music and media at the University of Paderborn in 2005, the native Westphalian first gained experience in the development of digital advertising marketing as the online editor-in-chief of the music magazine Visions. He then studied international economics in the master’s program at ISM Dortmund and INSEEC Business School in Paris with a focus on data-driven business models. In addition to his founding activities, Kleofas also acts as a mentor for other startups and works as a lecturer at the University of Munich. (sg)

Also read: AI in e-commerce: what is already working today?