Data Driven approach to Amazon Ad Optimization

Amazon advertising is changing – all the time. If you were around in 2015 when it all started, you would give me a resounding yes. Amazon Ads has come a long way from just auto-ads to 5 different ad types, placements, bid adjustments, bid modifiers and so much more. It’s hard to navigate complex data, features and functionalities of Amazon Advertising.

Today, the challenges involved in ad optimization go beyond bidding and keyword selection – it involves a detailed understanding of multiple variables, and the subtle art of balancing spends with sales. From choosing the right product skew to advertise to selecting the right campaign types, budgets, bids, placements – sellers and Amazon Agencies juggle through a lot to get their desired results.

But there is a catch. In the pursuit of fine-tuning details, brands and agencies get lost in the minute details, losing the broader picture.

Enter the concept of "Zooming Out". Zooming out is a strategic approach that allows users to step back and look at data at a 10k feet level. Let’s call it a holistic view of the advertising landscape. It allows sellers and agencies to assess their ad spends and performance from a macro perspective before delving into the specifics. It also helps sellers and agencies identify overarching patterns that get masked while focusing on granularities.

Today, we will talk about simplifying metrics, zooming out before zooming in and analyzing before optimizing. I promise, this strategy will help you make informed decisions and more successful outcomes.

Let me lay out the most important things that I would look at every week to find optimization opportunities:

1. Spend profile by products advertised

2. Spend profile by campaign type

3. Spend profile by placement type

4. Spend profile by keyword type

5. Spend profile by search term type

Understanding Spend Profile by Product and Product Band

I always say: when it comes to digital/ performance advertising, the “WHAT” is always more important than the “HOW”.

What are you advertising?

Are you advertising a high performing or high converting product or a low performing one.

Driving traffic to already high converting products will amplify your sales. Driving traffic to low converting products will simply be a money sink.

This is where the concept of Product Bands come in. Simply put, it is an extension of the 80-20 rule. The 80-20 rule states that 80% of your sales come from 20% of your products.

But relying on the 80-20 rule is hard for Amazon advertising. A seller account that has hundreds of products/ SKUs (a.k.a ASINs (Amazon Standard Identification Number)) advertises variety of products for variety of reasons: new launches, inventory liquidation, increasing sales.

Hence, it is better to break down the 80-20 rule into the 50-30-20 rule. This concept offers a structured approach to categorize and analyse your ad spend profile.

So, there are 3 Product Bands – Band A, Band B and Band C. The bands are defined based on sales performance of each product of the ASIN.

Product Band A represents the top echelon, including ASINs that drive the top 50% of your sales. These are your bestsellers, the products that consistently perform well and are often the primary drivers of your revenue.

Product Band B includes ASINs that contribute to the middle 30% of your sales. These products are significant but don’t match the stellar performance of Band A.

Finally, Product Band C consists of the bottom 20% of ASINs in terms of sales. While these products are the least performers in terms of sales volume, they can be important for portfolio diversity or targeting niche markets.

Categorizing ASINs into these bands has several advantages:

1. Prioritization: No points for guessing that Products in Band A should be allocated highest Budgets.

2. Strategy: High-performing ASINs in Band A will have a different advertising strategy compared to those in Bands B and C. Band A ASINs will focus on domination and visibility, Band B need ASINs might need a targeted ad boost to elevate their performance. Band C ASINs might need a conservative spend focusing on ONLY niche targeting. Or maybe even re-evaluation of their market viability (i.e. no ads and product discontinuation in simple language).

3. Impact: Advertising each product band calls for different ways of impact measurement. For Band A, analyze the return on ad spend (ROAS) to ensure that the high investment is justifiable by equally high returns. In Band B, consider experimenting with different ad types or marketing messages to see what resonates with the audience and has the potential to elevate these products. For Band C, the focus should be on cost-effective strategies, perhaps leveraging long-tail keywords or exploring cross-selling opportunities with higher-ranked products.

So, now that we know it is important, how can you identify this KPI. There are 2 ways:

(a) Data Aggregation:

In this method, you can aggregate data from seller central and advertising to define spend by Product Band. If you are already aware of your top sellers, you can also look at your total sales and define this number.

Step 1: Go to seller central> Business reports and download for the given week. If you are on vendor central, you can get this data from Report Analytics > Sales

Step 2: Divide product data into 3 Bands: Top Selling ASINs, Medium Selling ASINs, Bottom Selling ASINs. You can define top selling ASINs at top 50% ASINs by sale or a higher number. You have the flexibility of defining the metrics when manually doing the entire operations.

Step 3: Go to Advertising console. Download Product Dashboard data.

Step 4: Combine the spends and sales in one spreadsheet

Step 5: Add percentage spend on each ASIN. Compare % Spend on Band A, Band B and Band C.

Well, its not voila! But you got the job done!

(b) Ready to use dashboards:

On softwares like atom11, you can get a ready to use ad spend profile. For ex.

(i)You will know your spend by Product Band.

(ii) On clicking the expand button, you will know the performance of each product band

(iii) Within the expanded table, you will know which products are included in each band during that time period.

Understanding your Spend Profile by Product and Product Band is not just about allocating a budget; it's about strategically investing in your products based on their market performance and potential. This approach leads to more informed decisions, ensuring that each dollar spent on advertising contributes effectively to your overall business goals on Amazon.

Spend Profile by Campaign Type

There are three types of campaigns on Amazon: Sponsored Products (SP), Sponsored Brands (SB), and Sponsored Display (SD), each serving unique objectives and targeting different aspects of the buyer's journey.

Sponsored Products (SP): This type of campaign is particularly effective for driving sales for specific items and is best suited for increasing the visibility of individual products. High focus on SP reveals higher focus on ROAS metric for advertising.

Sponsored Brands (SB) campaigns, formerly known as Headline Search Ads, are designed to enhance brand recognition. SB campaigns are ideal for sellers with a range of products and those looking to boost brand awareness.

Sponsored Display (SD) campaigns are focused on reaching customers both on and off Amazon with auto-generated, display-like ads. These campaigns use Amazon's shopper data to retarget customers who viewed your products or similar ones. SD campaigns are crucial for reinforcing brand recall and nudging customers who might be considering a purchase.

Allocating budget across these campaign types is not a one-size-fits-all approach and should be aligned with your business goals. For ex. high ROAS focus brands should spend 60-70% of their budget on SP. For sellers who are trying to build a brand, atleast 30% spends should go to SB.

Another example is, a new product might benefit more from SB for brand awareness, while an established product might need more SP campaigns to stay competitive in search results.

So, now that we know it is important, how can you identify this KPI. There are 2 ways:

(a) Data Aggregation:

Step 1: Export Campaign manager data for a given date range

Step 2: Create a pivot on campaign type and measure spends and ROAS. Add Ad Spend % to find Spend % across campaign types.

(b) Ready to use dashboards:

On softwares like atom11, you can get a ready to use campaign spend profile. For ex.

(i) Right off the bat, you will know which campaign type is dominant (in terms of spends) for your account

(ii) You will know your spend by Campaign type

(iii) On clicking the expand button, you will know the performance of each campaign type

Effective budget allocation requires continuous testing and data analysis. Regularly reviewing campaign performance, experimenting with different ad creatives, keywords, and targeting strategies, and adjusting budgets based on performance trends are essential techniques. Using tools like Amazon’s Campaign Manager and considering historical sales data can provide insights into how best to distribute your advertising budget across these campaign types, ensuring each dollar spent is an investment towards achieving your specific business objectives on Amazon.

Spend Profile by Placement Type

Ad placement (only for SP ads) plays a critical role in ad effectiveness and overall performance. Amazon offers different ad placement options. Each placement has its own advantages and limitations. Understanding these placements – primarily top of search, bottom of search, and product detail pages – is key to crafting a spend profile that maximizes visibility and conversion.

Top of Search placements are arguably the most coveted. They appear on the top of the page when customer types a search query on Amazon.

Top of search is highly competitive due to its high visibility and the propensity to attract clicks from customers who are early in their buying journey. Products featured here get higher click through rates, higher conversions and subsequently higher sales.

Rest of Search placements appear at the end of the search results. While these spots receive less visibility compared to top placements, they capture the attention of customers who scroll through multiple listings, indicating a deeper level of engagement or a more considered purchase decision. Ads in these placements can be effective for products that require more thoughtful consideration or for capturing sales from determined buyers who haven’t yet found what they’re looking for.

Detail Page placements show ads on the product detail pages of similar or complementary products. This placement is effective for cross-selling and upselling, as it targets customers already in a purchasing mindset. Detail page placements can also divert a customer’s attention from a competitor’s product to yours, offering a strategic advantage in competitive categories.

How to evaluating Performance by Placement

So, now that we know it is important, how can you identify this KPI. There are 2 ways:

(a) Data Aggregation:

Step 1: Download bulk sheet for given time range

Step 2: On Sponsored Product tab, use all data to create a pivot placement too see placement wise data

Step 3: Create pivot on placement type and measure ad spend and ad spend% by placement.

(b) Ready to use dashboards:

On softwares like atom11, you can get a ready to use placement spend profile. For ex.

(i) Right off the bat, you will know which placement is dominant (in terms of spends) on your account

(ii) You will know your spend by Placement type

(iii) On clicking the expand button, you will know the performance of each placement type.

In summary, a nuanced understanding and strategic application of different ad placements on Amazon can significantly enhance the performance of your advertising efforts. By analyzing each placement’s performance and adapting your strategy accordingly, you can optimize your ad spend for better visibility and higher conversion rates.

Spend Profile by Search Terms

After peeling all the layers, we come to the crux of all Spend Profiles. Spend Distribution by type of search terms. Broadly, search terms can be categorized into branded, competitor, and generic. Let’s dive into it.

Branded Search Terms refer to keywords that include a “your”/ seller’s brand name. For example, “Sony headphones” or “Adidas running shoes”. These terms are highly targeted and often yield high conversion rates as they attract customers already familiar with and interested in your brand.

The primary strategy for branded search terms is to defend brand territory and capture high-intent buyers.

It is essential that we watch out on our spend distribution on branded search terms. Too high a spend means you are not focusing on new customers and too low a spend means you are not defending your turf.

Competitor Search Terms involve using competitors’ brand names in your keywords. For instance, a small electronic company might use “Bose speakers” as a search term to attract customers looking for similar products. The strategy here is to capture traffic from competitors, which can be effective but often comes with higher costs and lower conversion rates compared to branded search terms.

Generic Search Terms are broad and do not include any brand names, such as “wireless headphones” or “running shoes”. These terms are essential for reaching a wider audience, especially those who are not brand conscious or are in the early stages of the buying process. While generic terms can drive a significant volume of traffic, they often have lower conversion rates and higher competition, resulting in higher advertising costs.

Typical Spend Profile by Search Terms:

Spend Profile by type of search term depends on 2 things:

Brand’s sub-category: Each sub-cat is different. For ex. Skincare, electronics (phones/ laptops) relies a lot on branded search terms and hence brands in this sub cat should focus more on branded and competition search terms rather than generic search terms. On the other hand, commodity sub categories like floor cleaners or head phones are more generic. It is wise to measure your spend profile

Brand size: Larger brands with established recognition might benefit more from focusing on branded and competitor terms to defend their market position and capture traffic from rivals. Smaller or emerging brands, however, might find more value in allocating a significant portion of their budget towards generic terms to build brand awareness and customer base.

How to measure your search term spend distribution:

So, now that we know it is important, how can you identify this KPI. There are 2 ways:

(a) Data Aggregation:

Step 1: Download search term report for SP and SB

Step 2: Divide product data into 3 parts and Branded, Competition and Generic

Step 3: Create Pivot on Search term type and measure % Spend.

(b) Ready to use dashboards:

On softwares like atom11, you can get a ready to use search term as well as Keyword spend profile. For ex.

(i) Right off the bat, you will know which search term and KW type is dominant (in terms of spends) on your account

(ii) You will know your spend by different search term types and KW types

(iii) On clicking the expand button, you will know the performance of each search term type and KW type

Tip: We like to cap branded spends to 30% for large brands and <10% for smaller brands.

Conclusion:

In conclusion, creating a spend profile by search term type on Amazon involves a nuanced approach that considers the strengths and limitations of branded, competitor, and generic terms. By strategically allocating budget, dynamically adjusting bids, and continually analyzing performance data, sellers can optimize their ad spend for maximum ROI and effectively reach their target audience.

Summary

Regularly "zooming out" gives you a broad perspective on ad spends and helps you reprioritize your operations. It opens your eyes to understanding patterns rather than just upping and downing bids.

But it involves stepping back from granular details of day to day campaign management to evaluate the overall effectiveness of your ad strategy. Periodically zooming out helps in avoiding tunnel vision, adapting to changes in the marketplace, and efficiently allocating resources for maximum impact and return on investment.

Set your processes up to ensure this data is available at the drop of a hat. Because availability of data solves half the problem. Below is the ad spend portfolio graph that atom11 provides to its users.

Fun fact: As seen from above, this 1 view is a culmination of 7 different reports on Amazon Seller central and Amazon Advertising.

FAQ's

Q1: What is a data-driven approach to Amazon ad optimization?

A data-driven approach to Amazon ad optimization involves using specific metrics and data analysis to guide decisions on how to adjust and improve your advertising campaigns. This method relies on performance data, such as click-through rates (CTR), conversion rates, Advertising Cost of Sale (ACOS), and Total Advertising Cost of Sale (TACOS), to understand what works and what doesn't, enabling targeted adjustments for better results.

Q2: Which metrics are most important for Amazon ad optimization?

Key metrics include ACOS, TACOS, CTR, conversion rate, cost per click (CPC), and return on ad spend (ROAS). These metrics provide insights into the efficiency, effectiveness, and overall impact of your ad campaigns on sales and revenue.

Q3: How can I start implementing a data-driven approach for my Amazon ads?

Begin by setting clear goals for your campaigns, such as increasing brand awareness or boosting sales for specific products. Next, track and analyze the relevant metrics regularly. Use tools provided by Amazon, like the Campaign Manager and Advertising Reports, to collect data. Then, apply insights from this data to refine your ad strategies, such as adjusting bids, optimizing ad copy, and targeting more precise keywords or audiences.

Q4: Can data-driven optimization help reduce ACOS and TACOS?

Yes, by focusing on data and analytics, sellers can identify areas where ad spend is not yielding sufficient returns and adjust their campaigns accordingly. This might involve pausing underperforming ads, optimizing bids, or refining target keywords, all of which can help lower ACOS and TACOS while maintaining or increasing sales.

Q5: How often should I analyze my Amazon ad performance data?

Regular analysis is key to a successful data-driven approach. Weekly checks can help you stay on top of performance trends and make timely adjustments. However, for more strategic changes, a monthly review provides a broader view of what's working and what isn't, allowing for deeper analysis and more informed decision-making.