Amazon Sponsored Ads ROI


The System For Newly Listed Products

There is a certain approach for the first two weeks of your new product you should consider for your Amazon Sponsored Ads.

The objective is to make money ASAP. That might not always be your main objective (see the section on using Amazon Ads to buy rank in Jason’s Reverse Engineering Guide). If you’re looking for a positive return on your ad spend as quickly as possible, then this is the approach I recommend.

If you can swing it, I highly recommend you also have the following in place before you spend dime 1 on Amazon Ads, however, I also do not delay starting your ads only because it will take you a while to get these things in place either:

1 – At least 10 Reviews

They don’t have to be verified — though that would be ideal. There is something about double digit number of reviews that just seem to make your ads work better.

Plus, havinglO reviews means 10 more pieces of content for Amazon to consider when they determine what you’re relevant for. (Hint: think keywords in the reviews!).

2 – All 9 Images for Your Product

Now, I’ve seen people make money with one image and zero reviews -however since you’re spending real money here, it’s better to play the odds in your favor. Having 9 images – even if they are only slightly different just seems to work best for Amazon ads.

3 – The Ability To Lose It All

No matter what, don’t spend money on ads that you can’t afford to lose. Even if there is a 99.9% chance you feel you’ll be profitable – someone has to be the 0.1% sometimes. Only spend what you can part with. Better than budget is the approach, anyway.

You don’t need to spend a lot to get started. In fact, $10 a day for 14 days is usually more than sufficient. I recommend under ZERO circumstances you ever spend more than $75 a day for a new product on Amazon ads.

Cool? Let’s dig in.

Amazon Sponsored Ads – What Are They?

Before going any further, let’s take a look at where your ads will run.

Let’s say you had a silicone baking mat, and when someone types in “silicone baking mat” you’d like to show up as a sponsored result. Well, if you set up a bid that you’re willing to pay per click for that keyword phrase, and Amazon feels your product is relevant for that keyword phrase, then you might well show up here:

There are typically 4 units of sponsored ads that show up right at the top on the right hand side of a search result per page. If you don’t bid the most or aren’t considered the most relevant, you might show up on the right hand side on page 2, 3 or 10.

Or you might show up here…

In this case, this is the bottom of the page after the natural results Amazon shows. So right before someone can click “next page” they usually are presented two ads.

So you might conclude there are 6 unique spots on each page where an ad can show up and you’d be almost right. Look at the two images above -notice how the Farmers Market Delivered brand is showing up both at the bottom results as well as on the right hand side.

Why? Beats me. If I had to guess, they have set a high daily budget for their ads and Amazon is trying to spend as much of their daily budget as possible. Or they could be duplicating their campaigns in order to try to get more impressions of their ads… or something else. The real point is that is it possible for one product to show up twice on a page through Amazon

Sponsored Ads. Once you get some data and see what keywords are performing this is where you can scale it up even further.

But we’re getting ahead of ourselves.

First, how exactly do we set up our first Amazon Sponsored Ad campaign anyway?

Setting Up Your Campaign

Inside your seller central account

( when you log in, at the top navigation there is a section called Advertising. If you hover over it a drop down menu appears.

Click Campaign Manager:

Next, you’ll be able to create a new campaign:

Now begin setting up your campaign by giving it a name, an average daily budget, a start and end date and select manual targeting. Then click “continue to next step”…

Now you are going to select your product to advertise, the default maximum bid per click you’re willing to pay, and the keywords you’d like to bid on.

We’ll cover keyword picking and bidding here shortly. First let me just show you how to finish setting up this campaign:

Name your Ad group and pick your product.

With this strategy, we’re only going to use one ad group per campaign, so it doesn’t really matter what you name the ad group.

Then you’d put in your default bid, enter your keywords, adjust your bid per keyword (optional) and click Save and finish:

When you’re done, you’ll be able to see your campaign:

Notice how it says it may take an hour for your ads to show? Yeah, best case. It might take longer. In fact, you might not get yours ads to show at all. That’s what we’re going to create two more campaigns… more on that in a second. Click on the campaign name, then click on the ad group name…

And once there, click on the “keywords” tab, you’ll be able to see exactly how each keyword is performing an where you can adjust your keyword bid:

In fact, after you set up each of your three campaigns, you will absolutely want to come and adjust your keyword bid. I’ll explain latter what things like CPC and ACoS are.

Now let’s discuss how to pick keywords.

Picking Keywords

Some definitions first.

Technically a keyword is one word like “silicone”, whereas a keyword phrase is multiple keywords that, together, make up a search phrase. Example: “silicone baking mat.” We often use keyword and keyword phrase interchangeably.

With search engines, there are two ways to interpret search results – broad and exact. If you look at where you enter your keywords into your ad group for your campaign, you’ll see “broad” greyed out. It’s the only type of search you can target.

To understand what broad match means, let’s first talk about what exact match means. Say we were bidding on this: “red silicone non stick baking mat”.

With exact match we’d only show up when people put a search phrase in that had somewhere in that search the exact words in the exact order “red silicone non stick baking mat.”

So if someone searched for “red silicone mat non stick for baking”, you wouldn’t be considered an exact match because even though it has all the same keywords they are in different order plus the additional word “for.”

However, with broad match you could show up. As long as all the words in your phrase are part of someone’s search, it doesn’t matter what order the words are and if there are additional words in the search.

In other words, if you bid on “red mat”, you broad match all of these searches…

•    red silicone mat

•    mat, red

•    non stick mat that is red

•    I want a silicone mat that is not red

•    red door mat

Whereas in an exact match scenario you wouldn’t be an exact match for any of the above searches because the words “red mat” aren’t contained exactly in any of those searches.

Now know this – all of Amazon searches and all Amazon Sponsored Ads are broad match. Most people don’t realize this. We will use this to our advantage.

There are, in my mind, four types of keywords we use:

1.    Primary Keywords

2.    Secondary Keywords

3.    Penny Keywords

4.    Connector Keywords

To come up with the words that will fall into those categories, here is what you do. Go to and type in what you feel is your best keyword phrase and then open up the first 15 listings in a new tab….

Looking at the title of the first listing… you’re going to consider each keyword that you think a person might search for (even if not that often) and you’re going to start making a list with one word per line on it. In this example, here is the list I came up with based on the top 15 listings on page 1 for silicone baking mat:

1.    artisan

2.    pack

3.    non-stick

4.    silicone

5.    baking

6.    mat

7.    set

8.    silchef

9.    pieces

10.    professional

11.    grade

12.    heat

13.    resistant

14.    2

15.    fits

16.    half

17.    sheets

18.    red

19.    cook

20.    silpat

21.    premium

22.    inch

23.    cookie

24.    of

25.    liner

26.    fda

27.    safe

28.    certified

29.    reusable

30.    surface

31.    bakeware

32.    oven

33.    microwave

34.    attmu

35.    pastry

36.    no

37.    bpa

38.    parchment

39.    paper

40.    easy

41.    cleaning

42.    food

43.    linger

44.    heavy

45.    cooking

46.    best

47.    perfect

48.    Wilton

Some of the words you have to eliminate – like if mine is red, I don’t want to bid on someone searching for blue. And if mine an extra large mat I wouldn’t want to show on a small, and so forth.

You’ll notice that while looking at the list there is some weird stuff there: 2, no, for. Those are what we call connector keywords. Remember, our intent is words that people would use to make searches that our product would be relevant for. So if I had a 2 set, just putting “2 mat” as a keyword phrase can be powerful… Why? Because of broad match.

The word “no” is great for the same reason. “no mat” might be a phrase you’d bid on because you could show up as a broad match for words like “silicone baking mat with no bpa” as well as a hundred variations. Even better, since most people are not bidding like this, you can bid super cheap and get some great results. True, there aren’t a ton of searches for these longer tail searches but when you combine them all up together that’s lots of dollar bills you can be picking up with no effort.

Notice we also put brand names in here. It’s unclear exactly if it’s smart or not from a terms of use perspective to use these words to bid on your Amazon Ads. I will tell you this though – someone who works at Amazon higher up in their advertising department one time looked over our ads… and we were not bidding on brand names… and he told us to start bidding on words like nike for our product. So read from that what you will and make your own decisions.

Having 48 keywords seemslike a lot, right? Well, we’re just getting started.

Next, go to – if you don’t have an account, you’ll need to create one. The only reason you need to login to AdWords is to use their keyword planner tool:

Once there, click on “search for new keyword and ad group ideas” and then put in the URL of the best selling Amazon listing that is most similar to your product:

ON the next page, click on keyword ideas and you’ll notice there will be like 800 keyword phrases they suggest for your product:

You’re going to scan through all these phrases and pull out specific words that aren’t currently on your list that you think people might use when they do a search on Amazon. So here are some additional keywords that I think are relevant:

1. cookware
2. trays
3. cupcake
4. pan
5. kids
6. cake
7. tools
8. best
9. teflon
10. bread
11. drying
12. dishwasher
13. rubber

And there are more, but you get the point. When you’re done you should have easily 70+ unique words you can use. Now we’re going to create all the keyword phrases we’ll bid on.

Let’s start with your first list – you’re going to create what you think are your ten best keyword phrases. This is what I refer to us primary keywords. What word or combination of words do you think are most searched for. In my example, here is what I think:

14.    Silicone Baking Mat

15.    Silicone Mat

16.    Baking Mat

17.    Non-Stick Mat

18.    Non-stick baking Mat

19.    Silicone Baking

20.    Cookie Sheet

21.    Best Baking Mat

22.    Baking Mat Set

23.    Cooking Mat

Now this is important… when you create your three advertising campaigns you will use your 10 best keyword phrases in each campaign. Yep, you read that right. It’s weird, but sometimes a campaign doesn’t get as many impressions as it should, so for your best keywords, you should have them in 3 campaigns. Redundant? Sure. Worth it though.

Next I want you to create a list of 30 unique keyword phrases with a mixture of the words you used in your primary keywords as well as words you haven’t yet used that you think are somewhat strong keywords.

For example, here is what I came up with for silicone baking mat:

1.    silicone cooking sheet

2.    teflon baking mat

3.    non stick cooking mat

4.    silicone mat oven

5.    kitchen baking mat

6.    non stick baking sheet

7.    non stick cookware

8.    silicone bakeware

9.    best silicone mat

10.    professional baking mat

11.    non stick tray

12.    silicone cookie sheet

13.    silicone rubber mat

14.    silicone tray

15.    half sheet baking mat

16.    half sheet mat

17.    silicone rubber sheet

18.    silicone liner

19.    silicon mat

20.    2 set baking mat

21.    baking gift

22.    silicone gift

23.    chef mats

24.    fda mat

25.    oven mat

26.    oven baking mat

27.    silicone microwave mat

28.    no silicone mat

29.    inch silicone mat

30.    safe baking mat

You don’t have to stop at 30 – you can do more. Just divide the keyword phrases equally over your three campaigns. So at this point, I’d have 20 keyword phrases in each campaign. Before moving on, take another look at that list I just generated. See that there are things like “fda mat” and “oven mat” – these are phrases that are specifically created to try to trigger all these broad match searches. I’ll break down later on what the recommended bidding strategy for these should is.

Other keyword phrases you see are basically different ways to say silicone baking mat or secondary keywords we hadn’t used yet that I wanted to go after such as tray, Teflon, and cookware.

Now you’re going to create your next list – this is what I call “penny keywords.” You’re going to go after super duper broad one word keywords that you think are worth gambling a few pennies a click to show up for…You will create a list of at least 30 of these keyword. In this case, here is what I came up with:

1.    cooking

2.    silicone

3.    mat

4.    baking

5.    kitchen

6.    oven

7.    cookware

8.    bakeware

9.    sheet

10.    liner

11.    tray

12.    cookie

13.    non stick

14.    stick

15.    rubber

16.    cupcake

17.    bread

18.    pan

19.    silpat

20.    artisan

21.    set

22.    reusable

23.    wilton

24.    food

25.    pastry

26.    cake

27.    molds

28.    pizza

29.    silicon

30.    supplies

You will not bid more than $0.08 for these keywords – in fact I recommend $0.06. Just like we did with the last keyword set, you will divide these across the three campaigns you set up. When you’re done you’ll have 3 campaigns that each have 30 words that you’re bidding on. Technically you’ll have 70 keywords and phrases in total you’re bidding on.

Your Initial Bid Strategy

We just discussed for your last 30 keywords you spread over your 3 campaigns you’ll bid between $0.06 to $0.08. It’s important to know why.

Imagine someone does a search with the work “cooking” in it – a single word we’re bidding $0.06 on. If that was all they searched for, it’s doubtful we’d show up, and if we did, there probably isn’t much buyer intent for a keyword like that. However, do you suppose someone might search for “great cooking gifts for mom”? Perhaps. And since we’re bidding on the word cooking – even if it’s $0.06 – we might show up on page 1.

Let’s say we don’t though. Let’s say it is on page 5. Now think if someone is going five pages deep to find what they are looking for… and they see your ad (already having seen tons of ads before yours) and click on it – that is a great customer to have!

So now you know what your initial bid strategy will be for your last 30 keywords you put in, 10 per campaign.

What about the first 10 keywords you loaded in across all campaigns? In my example it was these:

1.    Silicone Baking Mat

2.    Silicone Mat

3.    Baking Mat

4.    Non-Stick Mat

5.    Non-stick baking Mat

6.    Silicone Baking

7.    Cookie Sheet

8.    Best Baking Mat

9.    Baking Mat Set

10.    Cooking Mat

Which I thought would be the ten most searched for phrases related to my product.

The approach we take with this is: low, mid, high.

In one campaign bid a low amount – less than a dollar, ending with a 1 or 6. Set the bid amount for each keyword – if you have the same exact amount for every keyword you’re doing it wrong. So here is what I’d bid on in one campaign for these 10 keywords:

1.    Silicone Baking Mat – $0.81 CPC

2.    Silicone Mat – $0.66 CPC

3.    Baking Mat – $0.61 CPC

4.    Non-Stick Mat – $0.86 CPC

5.    Non-stick baking Mat – $0.91 CPC

6.    Silicone Baking – $0.51 CPC

7.    Cookie Sheet – $0.41 CPC

8.    Best Baking Mat – $0.86 CPC

9.    Baking Mat Set – $0.86 CPC

10.    Cooking Mat – $0.66 CPC

You might wonder why something like non-stick baking mat I’d bid $0.91 for per click and something like cookie sheet I’d only bid $0.41 per click. The answer is simple – the more specific the keyword is related to what I think the customer wants to purchase, the higher I’m willing to bid.

Cookie sheet can produce buyers but it’s not as targeted and qualified as non-stick baking mat.

This is not an exact science and it’s okay to guess because when we acquire data we can fine-tune. When you are setting your bids, give a brief consideration to how qualified a buyer is to your product, based on the keyword phrase. Then bid accordingly.

For your second campaign with these same keywords, increase the bid price by 50-75% per keyword, again taking into account buyer intent. In this case it might shake out like this:

1.    Silicone Baking Mat – $1.41 CPC

2.    Silicone Mat – $1.16 CPC

3.    Baking Mat – $1.01 CPC

4.    Non-Stick Mat – $1.46 CPC

5.    Non-stick baking Mat – $1.71 CPC

6.    Silicone Baking – $0.81 CPC

7.    Cookie Sheet – $0.71 CPC

8.    Best Baking Mat – $1.26 CPC

9.    Baking Mat Set – $1.46 CPC

10.    Cooking Mat – $1.06 CPC

And then, for the third campaign, do a similar increase:

1.    Silicone Baking Mat – $2.11 CPC

2.    Silicone Mat – $1.86 CPC

3.    Baking Mat – $1.71 CPC

4.    Non-Stick Mat – 2.06 CPC

5.    Non-stick baking Mat – $2.31 CPC

6.    Silicone Baking – $1.11 CPC

7.    Cookie Sheet – $1.01 CPC

8.    Best Baking Mat – $1.86 CPC

9.    Baking Mat Set – $2.06 CPC

10.    Cooking Mat – $1.46 CPC

And there you go. Now for the remaining 30 keywords, you want to bid “slightly lower.” Let’s take a look at a few keywords for this:

1.    silicone cooking sheet – $1.01 CPC

2.    telfon baking mat – $0.86 CPC

3.    non stick cooking mat – $1.01 CPC

4.    silicone mat oven – $0.61 CPC

5.    kitchen baking mat – $1.01 CPC

6.    non stick baking sheet – $1.16 CPC

7.    non stick cookware – $0.41 CPC

In these examples, I’d have no trouble actually bidding more for each of those keywords… but I tapered down my bids to be at slightly less than what I’d normally bid. Again, what we do is compare buying intent based on the keyword and then we take that and compare it to similar keywords we’ve already established a bid price on to come up with our “slightly lower” bid price for these keywords.

A few things to note – doesn’t it seem crazy to pay over a dollar everyone someone just clicks on your ad? Well, that’s the maximum per click price you’re setting. Almost always what you actually pay per click is lower than your maximum bid. So there’s that.

The other thing is this – Amazon is a buyer search engine – not an information search engine. Almost every keyword phrase we are bidding on is going to show to someone who is actively shopping for something that hopefully is related and targeted to the keywords we bid on.

To put it another way – let’s say you have a 50% profit margin on a product that sells for $30. For every sale you produce, you get $15 in net profit for.

If you’re paying $1 per click for Amazon’s sponsored ads, you can buy 15 clicks to break even. If only 1 out of every 15 who click buy, that’s a 6.6% conversion rate, which is not that great. So as long as you can convert at over 6.6%, you’re making money. Awesome!

Of course, if you can’t convert at over 6%, then what can you do? Well, you can decrease your maximum cost per click bid price. Consider this scenario. You reduce the cost for $1 per click to $0.50 per click. Now because you reduced the cost you’re not as prominent with your advertising so you get a lot less clicks on your ads. However, now you only need 1 in 30 people who do click to buy to break even. Anything better than that is profit.

This is why for our main keywords we bid low, mid and high across 3 campaigns – to accelerate the data of what is the sweet spot that gets you the most clicks and has the highest ACoS – which is what Amazon refers to Advertising Cost of Sale.

Let’s analyze some data from a real world ad campaign:

Here is how you can analyze your initial bids to determine if you should lower them, pause them completely or let them ride.

For keyword #1 we have an ACoS of $20% – in other words every time we spend $2 in advertising it produces $10 in sales. Is this good? Depends.

On what? Profit margin. As long as your ACoS doesn’t exceed your profit margin, you’re break even or better. In this case we have a profit margin of 53% so we’re doing damn good!

Look at keyword #13 – it has an ACoS of 134%. Not good. It costs us $1.34 in advertising to produce $1 in sales, of which we only net $0.53. Look closer though at the “keyword bid” column and you can see we are now down to $0.48 per click on a bid. We have lowered it once, and if this doesn’t work we’ll lower it again – and if that doesn’t work then we’ll just pause the campaign. You have to break some eggs though to make an omelet, right?

Look at keyword #2 – seems to be awesome with a 8.7% ACoS. Well, not exactly. It only has 3 clicks. So we will watch this and see how it continues to perform as it gets more clicks. I typically like to get at least 20 clicks before I make a decision on an ad.

Now, with all that said, you’ll be tempted to immediately tinker with your bid price and/or pause keyword phrases. Don’t!!! While your clicks, CPC, spend, Sales, and Impressions data is pretty much in real time… your sales data is not, and therefore your ACoS is not accurate.

Why? Because sales are only directly linked to your ad campaign when the product is shipped. So you’ll see the click today, but the sale that click produced might not show up into two or three or four or more days later.

So it’s important that you do not pause a keyword during your first 14 days of running your tests. You can make snap decisions to lower a bid if you wish to, but most likely the best approach is to let it run untouched for the first 14 days – then you’ll at least be able to make more reasonable decisions about what to do with each individual keyword you’re bidding on.

Finally, there is a chance you do not get impressions for entire keywords -or at least minimal impressions. And furthermore, it might have nothing to do with your bid price either. For whatever reason, Amazon doesn’t see you as “relevant” for that keyword.

It is imperative that every unique word that you’re bidding on ideally appear in either your title or your search term fields.

You can and should edit your search terms for your listing right before you set up your ads to make sure each unique word in any phrase you’re bidding on is either in your listing title or in the search terms fields.

Additionally, once your ads start running – if you don’t get impressions still then edit your bullet points and put those keywords / phrases in there as well. Finally, if all else fails, create another campaign with those exact keywords that aren’t getting impression and try it again. Sometimes things just get “stuck” and this is one of the ways to attempt to “unstick” it.


At the end of the day, here is why this works – by mixing primary, secondary and “penny” keywords together and diversifying them across multiple campaigns we both accelerate our ability to get data as well better uncover those keywords that out of the gate are profitable.

I’ll take 6 losers for 1 winner any day of the week. With 70 keyword phrases, that’s 7 winning keywords pulling in ridiculous returns. And that’s not to say you can’t further fine-tune the other keywords that are profitable, break even or not yet profitable.


Please enter your comment!
Please enter your name here