Consider this thought experiment: how many hours and workforce would you need to research and analyse 1.000+ competitive products, along with 4 billion rows of historical data, while considering 20+ variables and their interconnectedness to conclude the optimal bid per product and keyword on a daily basis?

Our algorithm needs just a few seconds.

In the last couple of years, we’ve all seen a substantial shift in the digital marketing landscape. The digital revolution has shown a direct impact on the retail industry, leading to a decline in sales in physical stores and events such as the global pandemic of COVID-19 have accelerated the process to a great extent. One of the forefathers leading to such revolution is Amazon. The significant impact of Amazon on the transition of traditional brick-and-mortar shops to eCommerce is called the “Amazon Effect”. It is usually associated with the variety of products on the platform and the large number of customer-centric services (same-day deliveries, 30-day exchange policy etc.), which attract consumers to move towards an online shopping experience. This change in consumer behaviour led to increasing competition among manufacturers on the platform, whose aim is for their products to be featured among Sponsored Products. Only 0.1% of products account for around 80% of the sales on e-commerce marketplaces.

By utilising Machine Learning (ML) techniques like reinforcement learning and neural networks in an End-to-End (E2E) daily automated way, our solution can improve the performance of your Amazon Sponsored Products by adapting frequently and competing where it makes the most sense.

The Challenge

Win your key battles with the help of AI

Currently, SP is managed together with other digital campaigns and is often handled the same way. However, SP is unique in the sense that it is a standalone math problem. Indeed, there is no space for creativity in that ad format, as it takes all its components from the PDP (product detailed page).  Agencies usually manage multiple products in one campaign and review once a week, at best! An Amazon marketplace seller has to daily review keyword bids and make adjustments to meet the requirements of the auction in order to stay competitive. They need to go through a tiresome process of using different keyword research tools and reviewing competitors’ products to compile a list of the most relevant search queries. Also, this includes the daily task of reviewing the performance of campaigns and allocating the budget accordingly to maximise ROI.

Unfortunately, this is neither scalable nor efficient enough to couple with the agility, flexibility & reactivity that Amazon marketplace requires.

The Solution

If you want to stay ahead of the Amazon curve, you should adapt daily !

Get a new teammate in our AI buddy.

Over the last five years, we at Amplify Analytix, have continuously honed our approach to helping companies make better commercial decisions by implementing advanced data science solutions within their Marketing & Sales space. We’ve clearly seen the power of the Amazon marketplace for our customers and their competitors, which is why we developed a relevance enabling AI solution to build competitiveness vital in the dynamic Amazon space.

We ‘translate’ Amazon SP into a purely mathematical and logical problem, which our algorithm continuously works to solve.

Our Approach

Utilising advanced data science techniques like reinforcement learning and neural networks in a E2E daily automated way promises more won opportunities.

Amazon works with Ad Exchange to offer a real-time bidding environment where advertisers can participate in order to select Sponsored Products. Manufacturers (advertisers) choose to select the frequency of bids to be placed in a day. The bid price in this scenario is placed on the search keywords, which means advertisers have the option to bid for multiple keywords in a day.

The decision on the required keywords for which a bid needs to be placed is made by competitors’ analysis. The list of search terms on which competitors gain conversions is identified and is considered to be the target keywords to bid on. However, these keywords are quite dynamic in nature and are expected to change because of how customer search evolves and changes over time. Using these keywords as a feature for the model would ask for updating the model architecture every time.  As an alternative, we built a model architecture that is independent of the number of keywords by ensuring that:

  1. The number of keywords enters as a “batch” dimension instead of a “feature” dimension. This means that predictions on one keyword are treated independently from predictions on the other keywords.
  2. The keywords are represented as continuous vectors in space. The model outputs a vector in that exact space, and we find the K-nearest neighbours to that prediction.

Upon identification of target keywords, the optimal bid value of each keyword is identified using a dual-agent reinforcement learning model. One agent predicts the bid amount to be placed for a given keyword and the other predicts the long-term performance in the form of conversion for the suggested bid amount. The two agents are dependent on each other such that one agent helps to win the real-time auction and the other ensures that the bid results in maximum Return of Ad Spend (ROAS).

We used the open-source AuctionGym – a simulated environment created by developers from Amazon to enable a close replica of the actual bidding environment at Amazon for researchers and professionals. It helped us fine-tune our reinforcement learning models and strategise on deciding the target keywords, which could lead to better conversion.

The open-source simulated environment presents an impression opportunity to the bidders, on which the bidders decide a bid price and the ad to be shown. The environment then decides the winning ad and price, which is then shown as a Sponsored Product, possibly leading to conversion. The internal system in AuctionGym environment consists of Bernoulli process to simulate whether the allocation decision leads to conversion, enabling an end-to-end auction scenario.

The model shows the critical parameters used to select the most optimum bid and the estimate of the performance based on the bid price set. As time progresses, the agent learns the bidding strategy better and starts outperforming manually suggested bids for the keywords. For a completely trained model, RL agent’s bid suggestion provided approximately 30%  than a human agent (a manual process).

Already showing great tangible results

We initially implemented our approach within Google Ads, which resulted in a 30% improvement in Click-through rate (CTR) within 6 months. The increased visibility led to an 8% higher ROI, and the optimisation suggestions generated by our algorithms enabled the Account manager to allocate 35% of his time to more strategic tasks previously devoted to manual, operational work.

A friction-less implementation

Our solution can be an additional feature to what you currently use and an upgrade of your dashboard approach, so the investment in terms of time and resources would be minimal. Alternatively, we can deliver a stand-alone front end to manage input & output.

dashboard AI Amazon SP

The Success

Maximise efficiencies & business impact from your sponsored product investments

Consumers’ shift to digital media for shopping, entertainment, education, and other purposes has resulted in the accumulation of mountains of data that can be used for in-depth data analysis to depict complex consumer behaviour. Meanwhile, businesses around the world are recovering from the effects of COVID-19, and marketing budgets are limited. In such a case, implementing an intelligent bidding strategy that extracts meaningful information from vast amounts of data as efficiently and cost-effectively as possible aids in making the best use of the marketing cost constraints.

By automating performance optimisation, you or your Amazon ads partner can transform execution and operation time into strategic thinking to make better decisions. You get more actionable insights, which would help improve your KPIs, like CTR, ROAS and Sales Volume. Replacing the manual daily work with our reinforcement learning model saves time and generates more sales using fewer resources while being a source of extra intelligence for your media and marketing teams Finally, adding this unique AI layer to your Amazon SP provides a competitive advantage.

Our model is not only tested and approved, but it also is Amazon’s algorithm update-proof. It is change-resistance and is less dependent on historical data and trends vs other, widely used techniques. It perfectly mimics the search algorithms in the digital advertising space.

It’s not a black-box solution. Taking the user on the journey is our speciality! We can explain any context and actionable insights that come out of the data model, which we believe is also key to building trust in the solution and using it optimally.

Amplify Analytix AI Buddy

Generates more business impact & Intelligence

Requires less resources

Frictionless implementation

Unique and transparent approach

We are happy to answer any questions regarding our services, your data, and how we could work together

Contact us