Best Practice: Advertising / PPC / Sponsored Products

This is a condensed guide to creating and managing your Amazon Sponsored Products Campaigns.

This guide will explain some Best Practice to follow as well as how to use Trendle Analytics to optimise and automate your campaigns.

You do not need to follow our best practice if you don’t want to. There are many ways to manage successful campaigns, you are free to choose your own way.

#0: General Knowledge

What is “Sponsored Products”?

Amazon’s Sponsored Products (SP) is a way to advertise your products to customers at the top, the side or the bottom of a page.

When you run a campaign, Amazon charges you a fee per click (PPC – Pay Per Click), not by impressions (how often your ad appears on search result page).

Why are Sponsored Products important?

To get your product on the page 1 or 2 of results requires sales. And, in order to get sales you need to be on pages 1 or 2.

To resolve this chicken and egg scenario you need to perform some marketing activities.

You can promote your product externally on promotional website to drive discounted sales as well as create promotions and lightning deals within the Amazon ecosystem.

One of the most effective method is to advertise your product on Amazon search results. This will generate your first sales much sooner that just creating your listing and hoping. It does mean you have to be comfortable with the idea of not making any profit on the first few sales until your product starts to generate enough organic sales.

How are Sponsored Products structured?

Campaigns: Highest level. Here you set the overall daily budget for the campaign and the start and end date. There are 2 types of campaigns:

(1) Automatic: easy to set up and run but hard to optimise because you don’t have much control. Amazon automatically guesses which keywords are best.

(2) Manual: You select which keywords to advertise on, and the individual bid for each keyword. You can create Ad Groups so that you can segment your keywords and products. Manual campaigns give you complete control, so they take more effort to set up and optimise but are where the majority of your profits will come from over the long run.

Ad Groups: Middle level. Here you can create different Advertising Groups. This allows you to segment your adverts. For example, in one Ad Group you might focus on certain keywords, or on a specific product.

Keywords: This is what you bid on. If a customer types this word in the search bar in Amazon, then your product has a chance of appearing on the page. Whether your product appears at the top, middle, bottom, right or first page or later page(s) is depending on your bid and your competitors bid for that keyword

Match Type: Match type determine how accurate a customers search needs to be in order for your keywords to trigger an ad impression.

There are 3 Match Types:

(1) Broad: This is the widest net you can cast. If your keyword appears anywhere in the customer search term, then your ad has a chance of appearing. For example your keyword is “swim shorts” and the customer searches for: “yellow awesome summer swim shorts” then your ad could appear (if your bid is higher than your competitors on this keyword).

(2) Phrase: If your keyword is actually more than one word then your ad will only appear if the customer searches for it in the way you have specified. For example, your keyword is “swim shorts” and the customer searches for “swim shorts yellow ” or “yellow swim shorts”. In these cases your ad could appear. Your ad will not appear if the customer searches for “swim yellow shorts” or “yellow shorts” for example.

(3) Exact: This is the most precise option. Your ad will only appear if the customer searches only for this. For example, if the customer searches for “swim shorts” then your ad could appear. But, if the customer searches for “yellow swim shorts” then your ad will not appear.

Each Match Type accounts for some misspellings, plurals and adverbial (e.g. Swim — Swimming)

Bids: When you set a bid, you tell Amazon what is the maximum you are willing to pay to get a click. It therefore does not mean each click will cost that amount. For example, if you place a bid of $1 it could well be that your actual CPC (Cost per Click) is $0.40. This means that your competitors on this keyword at that point in time were only willing to bid up to $0.39 per click. As you were willing to spend up to $1, you won by outbidding your competitor by 1cent.

Daily Budget: Each campaign has a daily budget. Once your daily budget is spent your ads will no longer appear.

Good to Know:

Data: Amazon reports do not show Sales data until 2-3 days later. However you do see clicks and impressions etc data. So when you’re looking at your campaign results, don’t worry if you see 0 sales in the last 2-3 days. This is completely normal. We recommend you always filter out the latest 2-3 days results when looking at your data. Otherwise your data will not be accurate. Our algorithm automatically do this.

#1: Know your SKUs true profitability

Before you perform any marketing activities, you need to know what your true profitability is per product. Once you have established this you can set your campaign spending targets down to the keyword level by managing each keywords ACoS (Average Cost of Sales) carefully.

If you don’t know your true profitability then you will not be able to determine if your campaigns are performing well or not.

Within Trendle Analytics you can add your product purchasing cost. We automatically input the current selling price of your product and the estimated FBA Fees for that SKU at that price point. The result is your true estimated profit margin for that product.

Then, set a target ACoS which generates you more sales within your desired margin. This means that your ACoS should not exceed the estimated profit margin if you are aiming for.

If your product costs change, for example you receive better unit costs rate from your supplier, you can easily update the cost price and the application will re-calculate your new estimated profit margin.

#2:   How to structure you Sponsored Product Campaigns 

Golden Rule: Only advertise one SKU/ASIN/Product per Campaign

There are several reasons to do follow this second best practice rule:

  • Each product has different keywords that customers will search for. Putting ‘walking shoes’ and ‘dancing shoes’ in the same Advertising Campaign will return very different keywords and customer search terms.
  • In addition, filtering through keywords to understand which are relevant and which are not, is easier when you know to which product they refer to. The data we gather from Amazon does not allow us to determine down to keyword level which SKU it relates to. We can only do this at Campaign and Ad Group level. Also, one keyword may be relevant for two slightly different products in your portfolio. By splitting your SKUs into individual campaigns will help you gauge for which product the keyword works best.
  • Each product has a different cost and selling price. If you set yourself a target ACoS for each keyword, you can only achieve this if you know the exact spend per product. And this is only possible if you include only one SKU per Advertising Campaign or Ad Group.
  • More importantly, the more granular you are, the more data you have and the more control you have. You can allocate your budget with much more ease and flexibility across your products. With one product per campaign, you can increase or decrease your daily spend budget for specific products. For example, if your walking shoes have had a big increase in sales and you will run out of stock, you may want to immediately lower your daily budget until you get more stock in, but without impacting your bid price. But if you have your walking shoes and your dancing shoes in the same campaign, you will have to go through each keyword (note: you could do this at the Ad Group level, but we strongly recommend against doing this).

Note: If you have size variations or colour variations, you may include these SKUs into the same Advertising Campaign as the search terms the customer will use will be very similar. However we still recommend creating a different Ad Group per colour/size variation, or even a different Campaign per colour/size variation. You need to decide how closely related the items are.
Alternatively, you can choose to advertise only the SKU that has the most sales from your variation as the goal is to drive traffic to your listing.

#3.1:  Best practice Variant 1 – Process – From Broad to Exact

Step 1. (optional) You can create an Automatic Campaign if you do not have enough keywords to enter in a manual campaign. This is a good option if you’re lazy or want to see what Amazon’s creative algorithms think your product is.
In general we recommend you NOT to do an automatic campaign as this is often waster ad spend. Instead, perform good keyword research and add the relevant keywords straight to your manual campaign.

Step 2. Create a Manual Campaign

2.1 Add Keywords

Add any keywords which you think are relevant. Look at what is in your title, description and backend keyword. You can also use tools such as Keyword Inspector or merchant words or google trends and others to get other related keywords. If you are running an automatic campaign, add any keywords from the ‘customer search terms’ which are relevant.

Add these keywords in an AdGroup with default Match Type set to ‘Broad’.

2.2 Negative Exact bad Search Terms

From Day 2 of the campaign, log in to see which search terms Amazon is choosing to display your ads based on your broad keywords. You can already Negative Exact Match the obvious Customer Search Terms which you know are irrelevant for your product.

For example if your broad keyword for your dancing shoes is “Light Shoes”, and a search term is “Nike running light shoes”, then you can Negative Exact that search term before you even occur a click and therefore a wasted spend. Log in daily for the first week to be prevent wasted spend.

After that you can just log in weekly to see if there’s anything new customers have been typing (this is good to spot trends early)

You can also start creating rules such as “If no sales after 5 or 8 clicks then list as negative”

2.3 Move best Search Terms to Phrase and/or Exact Ad Groups

Wait 14 days before proceeding with this step! Why?

(a) Amazon reporting only shows sales following from a click 48-72hours (2-3 days) after the event. This means with 14 days data, you only really have 11-12 days of usable data.

(b) You need to avoid seasonality. By that we mean: your products will probably not sell at the same rate on Sunday as they will on Wednesday. So trying to have the weekend data blended into the week data will give you more accurate insights.

(c) You need to give your campaign time to work and for Amazon to find the relevant Search Terms.

Ok, so you’ve now patiently waited 14 days and have already filtered out obvious bad search terms to avoid incurring pointless costs.

What now? See Step 3!

Now you can take the best performing search terms and move them to a Phrase and/or Exact Ad Group.

2.4 Optimise Phrase and Exact Ad Groups

As before, continue to Negative Exact Match the obvious wrong search terms in the “Phrase” Ad Group.


#3.2 Best practice Variant 2 – Process – From Exact to Broad

Step 1: Find out your most relevant keywords


Step 2: Add these in ad groups of up to 250 keywords, place the keywords you team most relevant in the first few ad groups. Place higher ad group bid for these ones, and lower for the other keywords you consider long tail

Step 3: Set-up your recommendation and/or automation rules specific to that campaign.

A classic rule would be: If 0 orders after 8 clicks, list as negative exact

Step 4: Take your best performing keywords and add them in to a new ad group in Phrase and Broad matches AND negative exact. This will allow you to do 2 things: First, find long tail keywords for a keyword you know works well for you. So theoretically there should be some long tail related search terms that would work well as well, which you can then add as keywords. Second, by adding it as negative exact in this new ad group, you avoid canibalising your already well perfoming Exact match in the original ad group.

#4:    Recommendation for setting up and managing Automatic Campaigns, Manual Campaigns and HP Manual Campaigns

Important Notice: The recommendations below are no guarantees for success. Each niche is different and Amazon is a very dynamic marketplace. Therefore you may need to adjust the metrics presented below to fit your products niche better.

4.1 Create an Automatic Campaign to discover what customers are typing in the search bar in order to find and buy your product. We recommend not to do this, but instead to do detailed keyword research. If you still want to launch an Automatic Campaign then read on, if not go to 4.2

This is an optional step. An Automatic Campaign is a great way to just let Amazon do the leg work and find what search terms are the most relevant. You can then blacklist (Negative Match) the search terms which are not relevant/profitable and move the good keywords to a manual campaign where you have more control and flexibility.

If you already have a good list of keywords and you don’t want to spend budget on an Automatic Campaign which may be hit and miss, you can go straight to creating a Manual Campaign with one Broad Ad Group (see next step)

  1. Create one automatic campaign per Product
  2. If you have Product Variations, then create an Ad Group per variation.
  3. Set a daily budget of $10-20 and a bid price of $0.50-1.20
  4. Wait at least 10-14 days before making any changes to your campaign.
    Then, do the following once a week:
    4.1 If the search term has 0 orders, 1000+ impressions, 5+ clicks and lower than 0.70% CTR –> list as Negative Exact.4.2 If the search term has at least 1 order –> move to manual campaign (see next step)4.3 Search term has 0 orders, 1000+ impressions and greater than 0.70% CTR –> move to manual campaign (see next step)

    Tip: You can set and edit the recommendation settings by going to the “Recommendation Engine”. You can also delete and create as many custom recommendations as you want.

  5. After 30 days you will have discovered nearly all the search terms relevant to your product and variations.
    We recommend you reduce your bid price and/or your daily budget.
    Don’t stop your campaign, there may always be some new search terms appearing due to seasonality, new trends, competition changes etc.

4.2 Manual Campaign – 1 ad group only (Broad)

After patiently waiting at least 14 days, you can start to take the best search terms from the Automatic Campaign and move them to a manual campaign (see step above in Automatic Campaign).

If you did not do an Automatic Campaign, you will start here with the keywords you have found yourself.

  1. Create one manual campaign per product
  2. Create 1 ad group per variation called Broad” and with Match Type Broad
    Hint: Make sure to always name your ad groups in a way which will be easy for you to identify them. This will help a lot you later.
  3. Take the best search terms from the automatic campaign and move them to the Ad Group (see step above in Automatic Campaign)
    For example, if the term “summer cotton shirt” is a great search term in the automatic campaign, then copy that as a keyword to the Broad ad group.
    Note: When moving a search term from the automatic campaign to the manual campaign, make sure to mark it as “Negative Exact” in the automatic campaign. This will optimise your manual campaign performance.
  4. Manually add your own keywords. You can do this directly in Trendle Analytics app.
  5. Set the daily budget to $20-30
  6. Set the keyword bid to $0.75-1.50.
  7. Wait at least 10-14 days before making any changes.
    6.1 If search term associated to a keyword has 0 orders, 1000+ impressions and 5+ clicks –> add as Negative Exact Match keyword in that Ad group only6.2 If keyword has 3+ orders and ACoS is within your target and CTR is over 0.40% –> This is a High Performer (HP). Increase the bid over the next few weeks to drive even more sales. Make sure the bid remains within your profitability targets.
    Note: Increasing your bid may not necessarily increase your sales. Use the graphs to understand how bid price changes impacted performance until you find the sweet spot.
    Note: The sweet spot may be temporary (competition, seasonality etc are all external factors over which you have little information and control), make sure to continuously evaluate your keywords.
    Tip: The application will recommend the maximum bid you can set in order to remain within your profitability target. 
    6.3 If keyword has 3+ orders, ACoS is within your target and CTR is lower than 0.40% –> This is also a HP. You may want to check what within the title, image and price could increase the CTR.

    6.4 If the search term has 0 orders, 1000+ impressions, 5+ clicks and lower than 0.70% CTR –> list as Negative Exact in that Ad group

Tip: You can set and edit the recommendation settings by going to the “Recommendation Engine”. You can also delete and create as many custom recommendations as you want.

4.3 Add Exact and Phrase ad groups to your manual campaign

  1. Create 2 ad groups called “Phrase” and “Exact” with the associated Match Types.
    Make the Match Type for keywords in each ad group match the name. For example, all keywords in the “Exact” campaign should have a Match Type of “Exact”.

    Hint: Make sure to name your ad groups in a way which will be easy for you to identify them. This will help a lot you later.
    Here’s why:

    • Guiding principle: The more granular you are with your data, the better the quality of that data, and the better your campaigns will perform as a result. However, if you mix too many products per campaign then your data will not be as rich. In addition, the more granular you are, the more control you have.
    • By including only one SKU per campaign and using these 2 Ad Groups, you can be much more granular, flexible and quick in your keyword selection and bid adjustments.
    • It also allows you to have the same keyword under the 2 Match Types running in parallel which will tell you under which Match Type the keyword offers best profitability.
    • Like with all data, the more granular you can be the better. Doing one ad campaign for shoes and then splitting all your shoes by Ad Group will give you less detailed information and flexibility.
    • You allocate a daily spend budget per campaign. So if you have several products in the same campaign, it might be that some keywords for one SKU end up using up all your daily budget, leaving nothing for other products to be advertised. By splitting your products into individual campaigns you can easily adjust your daily budget spends on a per product basis and be in maximum control of your spending.


  2. Take the best search terms from the Broad ad group and move them to each Phrase and Exact Ad Groups
    For example, if the term “summer cotton shirt” is a high performing keyword in the manual campaign, then copy that as a keyword to the Broad ad group, the Phrase ad group and the Exact ad group.
    Note: When moving a search term from the broad ad group to the Exact and Phrase ad groups, make sure to mark it as “Negative Exact” in the Broad ad group. This will optimise your campaign performance and budget. 
  3. After a few weeks look to see if a keyword is performing significantly better or worse in any of the Ad Groups.
    For example, if Keyword “cotton shirt” is doing great in Exact but not so great in Phrase and Broad, then lower the bids progressively over several weeks until you get them to be profitable (even if this means you only get 1 click every few months). We don’t recommend you remove or pause that keyword.
  4. Keep managing each keyword bid to increase your sales whilst remaining within your ACoS target

Was this article helpful?

Related Articles

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.