download the Amazon Ads calculator
You can only use this calculator to analyze ads that are (1) running for Book 1 or Box Set 1 (e.g., starting from Book 1; Books 1 – 3, Books 1 – 5 etc.) and (2) include only a single book being advertised in the Amazon Ad campaign (e.g., you can’t have all the books in your series being advertised in the same campaign because it skews the sales and reads numbers). Otherwise the analysis will be inaccurate.
Revenue and Profit Per Click Analysis
These are the two main metrics you use to analyze the Amazon Ads. For a new series or other situations where you don’t have reliable revenue per sale and revenue per borrow numbers use cost per unit instead.
You need 20+ clicks on a keyword or ASIN to have enough data for the revenue per click (RPC) and profit per click (as well as conversion and cost per unit) to be stable enough to analyze. If possible, I want 30+ clicks (not always possible, however, in smaller genres). In smaller regions other than the US / UK, I’ll aim for 15+ clicks.
If a keyword / ASIN gets 30+ clicks, and your revenue per sale (RPS) and revenue per borrow (RPB) are reliable, then the revenue per click number you see for a given target will likely be around the breakeven CPC (plus-minus 10 – 15%).
More clicks means you have a larger sample size and the data will be more stable. However, it takes longer to accrue this data and it’s also more expensive to test each target when you’re using a higher click cutoff.
The Three Ad Strategies
Where you choose to set your bids is based on your strategy; this will vary from series to series and can change depending on the situation (e.g., if you’re running a deal you might temporarily be willing to lose money on the ads to drive visibility, even if your normal strategy for that series is breakeven):
- Profitable: want the ads to be directly profitable; can set a target profit per click here and calculator will give you a target bid. How much you want to make per click depends on your personal preference; a higher number (e.g., aiming to making $0.40 per click instead of $0.10) will limit scalability since fewer targets will hit this number. Your target bid is the RPC – your target profit per click. E.g., if “lee child” has a $0.91 revenue per click, and you’re aiming to make $0.20 per click, then you’d set the bid at $0.71 ($0.91 – $0.20 = $0.71 target bid).
- Breakeven: aiming for ads to break even and be profitable from the algos / fanbase building. Set your bids at the revenue per click. E.g., if “jim butcher” has a $0.78 revenue per click, this means setting the bid at $0.78 will be around break even.
- Loss Leader: willing for the ads to lose money with the aim of being profitable from the algos / fanbase building. Can set a target here on the calculator (e.g., -$0.30 = willing to lose $0.30 per click). Your target bid is your RPC + your target profit per click. E.g., if “thriller” came in at $0.42 revenue per click, and you were willing to lose $0.30 a click, you’d set the bid at $0.72 ($0.42 + $0.30 = $0.72 target bid).
Note that in most fiction genres, you’re looking at a bid of $0.70 / 70p or more minimum in the US and UK to get the ads to serve impressions; to scale up competitive targets, you’re likely looking at a bid of $1+. Below this minimum bid floor you’ll get very few clicks / impressions; usually I’ll turn a target off if I’m not willing to bid $0.70, since it’s not worth the time to manage it. There are no tricks to get the bids lower; these are just based on competition in the genre.
The bid floor will vary based on genre / region, but the principle remains consistent: below a certain bid, you won’t get many (or any) impressions / clicks.
When trying to scale, I generally start with a breakeven strategy, as this allows me to target more things and push the bids higher (thus allowing me to spend more). If I’m not seeing the algos kick in or a boost in newsletter subscribers (e.g., I’m just actually breaking even), I’ll switch to a a profitable strategy.
A loss leader strategy is extremely aggressive and not recommended unless you’re getting algorithmic help already (e.g., from a launch or big promo). You can lose a lot of money here, so keep an eye on your numbers. I do not recommend running a loss leader strategy for backlist (as I’m doing for the campaigns in the video); I am testing things using an extremely aggressive approach that I do not recommend for others.
80/20 Analysis
(optional before starting) add Search Term Impression Share report to calculator
- Set the series / ad region / target clicks / target profit per click on the Series sheet.
- Start your analysis with the book spending the most money.
- Export the data for all the campaigns for this book. I usually select year to date or lifetime to get a good sample size of data to analyze, unless I’m just analyzing a specific time period.
- Focus on optimizing the campaign(s) spending the most money first. Generally, 1 – 3 campaigns will account for over 50% of the spend. Make sure you analyze and optimize these before moving on to other campaigns.
- To analyze the performance for the targets (keywords, ASINs, categories etc.) in a campaign, export the data and then paste it into the respective calculator. Optimize the bids and turn off targets in accordance to your strategy for that book (profitable, breakeven, loss leader).
- If time allows, analyze and optimize the other campaigns, and / or go on to the book spending the second highest amount. If you’re going on to the other campaigns for the same book, I would recommend analyzing all the campaigns of a specific ad type (e.g., author keywords) together then going to the next type of ad (e.g., ASINs, Kindle Category ads etc.), rather than going down the remaining list in order of spend. It’s easier to focus on optimizing one ad type at a time since you’ll be moving winners between campaigns, turning off targets, etc.
- Research additional keywords and / or ASINs and add them to your test campaign(s).
How to Scale
There are three primary ways to scale the Amazon Ads:
- Turn off poor performing targets so your budget goes to the better performers and thus generates more sales / reads for the same ad spend.
- Increase bids for the best performing targets so that you get more clicks on those and thus generate more sales / reads from them.
- Test additional targets (keywords, ASINs, categories, auto ad) to get more clicks and thus generate more sales / reads.
Three additional quick ways to scale:
- If you have backlist in the same sub-genre, you can re-use your winning targets for Series A with Series B to scale. Note that the keywords that work for Series A might not work for Series B, even if they’re in the same sub-genre. The winning targets will be unique to each series, so make sure you’re doing the revenue per click analysis on a series by series basis.
- If you have a box set in the same series, you can re-use your winning targets that you tested for the individual book(s) there to scale. Winners won’t always replicate for the box set, but are worth testing.
- If you’re running ads in the UK or another region where English is the main reading language, you can re-use your winning targets in the US (or vice versa) to scale. Winners won’t always replicate across regions, but this gives you a good starting point for testing.
Impression Share Rank and %
These are very general rules of thumb; there are plenty of exceptions. But you can use these metrics in conjunction with your revenue per click and profit per click analysis to determine whether it makes sense to increase your bid.
These metrics are cumulative across the entire account, not measured on an individual book or campaign basis. So, using the guidelines below, if you have 30% impression share for a keyword and it’s being targeted for three of your books, it still might be pretty scalable for the book you’re looking at, which might only be responsible for, say, 7% of the impressions (Amazon doesn’t split this data out on a book-by-book basis, so we don’t know the exact breakdown across our catalog).
- Rank: if you’re already ranked #1 for a keyword / ASIN, then it can be tough or expensive to get additional impressions / clicks. For some keywords, #2 is going to be as far as you can get because the author is bidding on their own name (and will thus have preference in the ad auction).
- Impression Share %: if this is above 25 – 30% for a midlist or larger keyword / ASIN with decent impression / click volume, then it can be tough or expensive to get additional impressions / clicks. For less competitive / smaller targets, then you might be able to capture 50% – 80%. For your own keywords / ASINs, it’s possible to get 90% or even 100% (note: 100% is not the goal for your own targets, I’m just highlighting that it’s possible; it’s often prohibitively expensive to do this).