On Monday, Teikametrics has announced the addition of an hourly bidding solution for Amazon advertisers using Teikametrics Flywheel campaign optimization platform.
According to the company, the bidding algorithm is powered by machine learning and analyzes ad data, transaction data, and importantly, cost of goods sold data. The algorithm analyzes bids hourly to make adjustments when needed. “Amazon is continually providing more and more access to data and API opportunities, along with more ad units,” said Alasdair McLean-Foreman, CEO of Teikametrics.
Teikametrics refers to Flywheel as a retail optimization platform (ROP). “We consider this a new category,” said McLean-Foreman. According to him, it’s a different use case, since the transactions from ad engagements happen right on Amazon. The platform uses machine learning to optimize pricing, operation and inventory performance for advertisers on Amazon.
“We can see the transactions and profitability. The brands don’t give Amazon cost of goods sold data, but they give it to us,” explained McLean-Foreman. This combination of data (transaction, ad and cost of goods sold) allows the algorithm to optimize bid for profitability.
McLean-Foreman made the comparison between the hourly bidding algorithm and a digital thermostat. It can regularly check and react to climate changes to maintain a steady temperature. “The bids don’t change every hour by default,” he said. It basically checks the bid status more regularly in order to deliver more constant optimization.
“The longer you have an under-optimized bid, the more it’s going to cost you,” said McLean Foreman. The company claims $6 billion in retail sales has gone through the platform. The scale of data and optimization frequency means its hourly algorithm can deliver millions in increased profitability for sellers, said McLean Foreman.