Digital Marketing

Partner Q&A with Teikametrics – A Retail Optimization Platform

February 7, 2020
By BJ Cook,

Finding and testing the best tools to give our clients an edge is part of our day-to-day at Digital Operative. As our Amazon practice has grown, we’ve tested several ad optimization platforms. We’ve chosen Teikametrics as the newest member of our partner network.

Combining our deep understanding of Amazon and emerging brands with Teikametrics’ sophisticated machine-learning technology is sure to drive sales for our direct-to-consumer clients. We recently had a call with their new Chief Product Officer, Srini Guddanti, and Mike Indigaro, Director of Agency Development, to learn how their machine-learning algorithm works, common keyword bidding mistakes, Mike’s ‘aha moment’ teaming up with brands, and much more.

Chief Product Officer, Srini GuddantiMike Indigaro, Director of Agency Development

Q: Srini, you’re the newest staffer at Teikametrics, after a 14-year stint at Amazon. What about Teikametrics convinced you to join them?

Srini: For me, the biggest thing was what I saw in Teikametrics CEO, Alasdair, and the rest of the leadership team at Teikametrics. The leadership talent and vision which is at par/ exceeds that of the caliber seen at Amazon.

I was also impressed by Teikametrics’ ability to evolve their product offering. I met the Teikametrics team for the first time almost three years ago as one of the Amazon Advertising integrators. The way that the product has been developed from its core, which is driven off of Machine Learning and Data Science, makes it a solid foundation for building future capabilities that are in line with Sellers’ business growth objectives.

Q: One thing we respect is Teikametrics’ proprietary data science practice. How can brands see greater profits by leveraging it?

Mike: Our sellers typically have KPI’s revolving around ACoS and TACoS. Some of them run against strict budget, others run against this at scale. The way in which we drive performance outcomes is through our bidder. On Amazon, you need to be analyzing at the most granular level, the keyword to ASIN relationship, and bid to the true value of each click. Our technology analyzes the keyword activity every single hour, and adjusts your bids up to hourly to place your bids at the exact value bid that aligns with your marketing goal.

Q: Like Teikametrics, one of the first things we do at DO with emerging consumer brands is learn product profitability. How does Teikametrics optimize advertising profitability?

Mike: Sellers express to the Teikametrics software or bidder an objective at the campaign level that we call MACS, or Maximum Advertising Cost of Sale.

Before we optimize campaigns for these products, we need to get an understanding of product level economics and historical trends – what is the pre-ad GM%? What are the targets for these products – are we trying to grow sales or maintain efficiency? From there we will come up with an initial strategy to inform our value based bidding, and refine as our objectives or scale goals change throughout the year. Our bidding technology will update bids on up to an hourly basis as performance data for each keyword changes in these campaigns.

We will always want to start with understanding profitability at the product level, which we’ll ask our brand partners to share their COGS in order to determine margin potential from the Amazon channel. From there, we can group products based on their margin attributes, as well as their product attributes. It will also be important to understand the lifecycle stage or business objective for a particular product to help guide what an appropriate MACS (Maximum Advertising Cost of Sale) is. For instance, a product that is being launched on Amazon for the first time will likely have less strict profitability constraints than an item that is mature and highly desired by your customers – this will translate into a proportionally higher MACS for the new launch for this particular product, but it will still be anchored to the margin of that item.

Q: We hear your machine-learning algorithm crunches billions of data points on Amazon. How does it help prevent advertising dollar waste?

Mike: Across the > 12.5M Keywords that we manage for our customers, we find that only about 5% of them generate more than two clicks per week. Although testing new keywords is important, we feel there is larger value in optimizing as much profit as possible from KWs that are able to generate meaningful click volumes.

Teikametrics Flywheel has developed mechanisms to track spend by search term, with actionable insights to negate keywords directly from the platform. Additionally, because we know a seller’s willingness to spend or your MACS target, our bidding algorithm will analyze bids each hour to determine where bid adjustments need to be made to stay within the limits of the goals that you set. Each hour we’re understanding the likelihood that your product ads will convert; thus using this data to ensure your ads are strategically placed before shoppers that is most aligned to your goals and outcomes.

Q: Through automation, what are the daily tasks of retailers and sellers that Teikametrics handles?

Mike:

  • Hourly bid changes to maximize keyword efficiency and visibility
  • Margin-aware algorithmic bidding
  • Product-level profitability metrics and reporting
  • Holistic reporting on total Amazon advertising performance (TACoS)
  • Automated keyword discovery and negation
  • Proactive account notifications and alerts

Without this automation, sellers are required to rely on making these changes manually through Amazon’s Advertising Platform. These changes are often not well informed or have statistical significance to drive meaningful outcomes.

Teikametrics Flywheel platform takes a methodical approach to breaking down all areas of Amazon Advertising into a user-friendly dashboard.

Q: Mike, what are two tips you can share from the most successful brands using Teikametrics?

Mike:

1) Segmenting brand vs. non brand spending. You never want to blend the performance of brand advertising and non-brand advertising when making bidding and budgeting decisions. We suggest targeting by generic, brand and competitor keywords. Each of these buckets of audience search terms have varying levels of intent and will be measured accordingly. It is much harder to get a customer to search for your category of products rather than your brand term. It is also important to purchase all ad types, from Sponsored Products to Sponsored Brands.

2) Don’t be afraid to spend and do not let budgets constrain growth. One of the most common pitfalls we see are brands not having enough budget to support the demand for their products. Budgets are often capped out during the day, giving consumers just a small window of time to find your products. With over half of all e-commerce searches beginning on Amazon.com, brands should consider an “always on” strategy to capture sales opportunities. This requires some delicate budget planning and bidding to ensure that you are pacing through your budget as efficiently as possible.