FAQ
What is Nimble Ads?
Nimble Ads is an AI-Powered advertising bid and budget optimization platform for sellers on Amazon, Instacart and Walmart marketplaces. Our mission is to simplify the management of advertising campaigns while increasing their effectiveness at the same time. We are focused on advertising optimization because we believe it can have the most impact on sales volume and profitability.
What does Nimble do?
We help sellers and brands optimally allocate their advertising spend across multiple campaigns and dynamically maintain their bids for targeted keywords, products, and audiences. Our unique approach helps substantially increase sale velocity and reduce costs by consistently improving the bids and optimization.
What is the Nimble approach to advertising products?
We are fast, agile, and customizable for our clients to drive the best possible desired outcome. We have a holistic solution which leverages multiple data points including manufacturing and marketing costs, your budget, goals, inventory availability, lead times, and past performance and helps you maximize your returns.
How does Nimble make decisions for you?
Our proprietary machine learning algorithms constantly learn and predict performance for each ad for each targeted keyword, SKU/ASIN, ad type, placement type at a granular level.
How does Nimble account for seasonal trends for your products without very much seasonal information?
Once we connect with your Amazon advertising account, we are able to download past performance data to learn seasonal patterns. Amazon and other platforms allow us to look at the past 60 days (or longer) which is enough to learn day-of-week patterns. Additionally, we use insights from our overall customer base, feedback from Amazon and Instacart, as well as historical traffic trends for important keywords to estimate recurring seasonal patterns and help customers plan bids and budgets (e.g., for holiday season). Our AI-Powered system will learn very quickly and adapt to seasonal trends rapidly across 1000’s of products.
What is Nimble’s system optimizing on? Sales? Traffic? Conversions? ROAS?
At their core, our algorithms optimize on revenue per click. We then use configured portfolio and campaign goals to translate these revenue estimates to bids and constantly improve our models with feedback as it arrives. We are also able to optimize on profit (with some additional configuration), and for a given portfolio-level overall sales target (using data from the Orders report and some periodic high-level configuration changes). We are also able to support several other goal types, if desired, as long as the data to compute the desired metrics is available.
What does Nimble support on the Amazon platform?
We support Sponsored Products (search & detail pages), Sponsored Brand Ads (Video, Product Collect) and Sponsored Display (both on-site and offsite retargeting and audience-targeted Ads)
What does Nimble support on the Instacart platform?
We support Sponsored Product and Display Ads on Instacart Premium Placement Advertising.
What data inputs are put into Nimble’s machine learning system in order for it to have an make decisions on budget allocation between products, campaigns and goals?
We primarily use ACOS / ROAS goals at portfolio (and campaign) level to evaluate campaigns for budget allocation. These goals are computed considering COGS, organic performance (evaluated using data in the overall ASIN/UPC/SKU orders report), inventory, and the current stage of the product in its lifecycle in close partnership with the customer. Available bandwidth is automatically determined, and the system experiments with incrementally increasing the budget for high performing campaigns to find the sweet spot where it can spend the budget while meeting performance goals.
Is Nimble Ads maximizing true incremental return or measuring attributable sales?
We collaborate with customer on overall goals, periodically analyze both Ad-Attributed and overall sales, and configure the system accordingly. For example, if the main goal is to hit $10M in overall sales for a given brand, we may configure campaign-level goals in a way that reward campaigns for ASIN/UPC/SKU with better Ad-based to Organic sales ratios, even though their Ad-based performance may be similar to other ASIN/UPC/SKU in the portfolio. The system then takes that input and makes up-to thousands of bids changes every hour to meet the configured goals.
How is Nimble factoring things like non-brand, conquesting, and brand keywords? Equal value?
The Nimble system is flexible and supports several ways of organizing campaigns and portfolios to fit a variety of scenarios in this area. We can put all keywords in a single campaign and just bid based on performance, or create separate portfolios focused on branded, unbranded and conquesting campaigns with different goals and budgets. We are flexible and agile, so we work with you to address preferences in the set-up process.
How does an auto campaign fit into the campaign?
To obtain additional keywords and targets. Once we have sufficient data for a given keyword and target, it’s blocked from the auto campaign so the manual campaign could bid more precisely.
What reports are available through Nimble Ads that are unique?
Campaign reports, keyword reports, product / ASIN reports, Ad-Group reports and Ad-Group reports use direct advertising data but provides several interactive ways of displaying the reports that help our customers with new and innovative insights. Amazon: One report that is unique is our overall orders report, portfolio report and BSR / category rank reports. We can configure these for you.
Is there any human element that is optimizing once loaded into Nimble Ads?
Our goal is to create a very productive human-machine partnership where human help is used to configure goals and guide the system with minimal effort, and let the machine learn and manage the vast majority of actual bid and budget changes. For a typical account, human involvement is somewhat substantial in the first few weeks, and it reduces substantially upon stabilization. Some of our largest stable accounts average a handful of changes per quarter (typical to align advertising with changes in business priorities) and the system is able to maintain excellent performance at its own. We are open to human input to machine learning settings and set up regular calls as needed to incorporate this valuable input data.