AMZ Ads Campaign and Bid Strategy, Part 2 - Understanding The Game

“The key is not the will to win… everybody has that. It is the will to prepare to win that is important.” - Bobby Knight

As a preface, this article is part 2 of a 3 part series.  If you haven't read part 1 or you’d like to review it, click HERE to read that as we’ll be building on the foundational elements outlined there.  As a reminder…for the purposes of these articles, I’ll use the terms “keyword(s)”, “keyword phrase(s)”, “term(s)” and “search term(s)” interchangeably because for the purposes of this article, they all mean essentially the same thing.  I hope that all makes sense.

It’s probably worth starting with my take on some of the fundamental terms, acronyms and options we’ll see as we navigate the Campaign Manager dashboard.  I’ll assume for the moment that if you’re reading this, you know how to log in to the Campaign Manager dashboard.  I’ll also stick to “Sponsored Product” campaigns since those should form the basis of your marketing…we'll cover the additional, more advanced campaign types (sponsored brands, sponsored display, lockscreen, etc.) in a future article.

---

Spend - How much has a given campaign or group of campaigns spent within the time period selected.

Sales - How much has a given campaign or group of campaigns generated in sales within the time period selected.

Orders - How many orders has a given campaign or group of campaigns generated within the time period selected.  In this case, a consumer who orders 2 units will count as 1 order.

*A reminder that any sales figures shown above (“Sales” or “Orders”) within the Campaign Manager dashboard do not include organic sales.

Impressions - In AMZ’s Price Per Click (PPC) ad format, you don’t pay for impressions.  Impression(s) are the number of times your ad is loaded on a page, regardless of whether the person clicks on it or even sees it.  If your ad is prominently placed at the top of the page or buried down below, that’s an impression.  If your ad is on page 1, page 5, within the detail page of another listing…so long as that page is loaded on to the consumers device, that’s an impression.  If your ad is shown to a consumer who searched for keywords unrelated to your offering or keywords that are both relevant and specific to your offering, those are impressions.  Although impressions are a valuable metric, not all impressions are created equal.  That said, they're an excellent indicator of whether your ads are being seen.  Low impressions means something’s wrong.  It could be that your listing isn’t optimized for that term, that AMZ doesn’t feel that your listing is appealing to consumers using that term or that you’re not bidding competitively amongst other things.  Either way, impressions are the proverbial “canary in the coal mine”.  If the canary drops dead, other things take on much less importance but a singing canary doesn’t necessarily mean you’ve hit pay dirt.

Clicks - In a PPC ad format, clicks are what you pay for.  Click(s) are the number times a consumer saw your ad, found it appealing and clicked on it…simple as that.  Keeping in mind that before clicking your ad, the consumer only sees the basics of your listing (primary image or book cover, title, review rating and price), a click is the first indicator that consumers who are searching a given keyword might be looking for what you’re selling.  If that weren't the case, they wouldn’t have clicked on it.  AMZ assumes this too and generally speaking, will at the very least lean towards showing your ad to keywords that generate clicks over keywords that don’t.  Click data can be an excellent indicator of effective keywords, ineffective keywords and irrelevant or overly vague keywords…spoiler alert for the next article but most people don’t do a good job of managing this information.

Click Thru Rate (CTR) - This is simply the % of the time your ad is shown and gets clicked on…impressions divided by clicks.  0.5% and above can be considered a good CTR.

Sales or Conversions - Of course, this is where you get paid.  A conversion is when a consumer likes what they see, adds it to their cart and buys it.  This is validated proof that customers who are searching using that keyword are looking for what you’re selling.  Again, AMZ sees this too and generally speaking, will boost your listing’s ranking (sponsored and organic) for that keyword and variations very similar to it.  From an ads perspective, these will be the terms that you will ultimately target as part of an effective manual campaign(s).

Conversion Rate (CR) - This is simply the % of the time your ad is clicked on and then purchased…clicks divided by conversions.  It’s difficult to say what a “good” AMZ conversion rate is because it varies based on the niche you sell in but generally speaking a conversion rate of 10% or above should be the goal.

Cost Per Click (CPC) - This is the average amount paid for each click within a campaign or ad group.

ACoS and/or RoAS - These were outlined in part 1 of this article.  To review that, click HERE and scroll down to the part where we talk about “effective sales” near the end.

Bid Range - This is calculated from a group of recent successful bids, promoting listings that are similar to yours. The range represents the low and high end of those successful bids.  It is dynamic in that it is based on a rolling time period and therefore updates daily.  “Successful” means that the bid was awarded a placement that received an impression.  Since we know that the vast majority of consumer purchases are generated from the coveted “page 1” (again…covered in part 1 of this article) and we also know that not all impressions are created equal (as outlined under “Impressions” above), it’s easiest to think of “successful” impressions as those that occurred on page 1 since they are most likely to be seen and most likely to result in a sales.

Suggested Bid - Think of this as the proverbial “sweet spot”.  It represents what AMZ feels is the best starting point for you to begin bidding based on the data of your listing and the listings of your competitors.

With respect to bid range and suggested bid above, trust the information that AMZ gives you access to and trust me when I say that it’s good information that provides a solid starting point.  If you’re operating under the theory that “I don’t trust the data or the numbers”, you’re probably in the wrong game.

---

Whew!...That may have seemed like a lot of real estate to waste before even getting to strategy but terminology matters…language matters.  If you’ve ever tried to hold an important conversation with someone who isn’t fluent in the language you’re speaking, or vice versa, the risk of a misunderstanding is significant.  There are many other terms, acronyms and options but for the purposes of this article, I’ll limit it to the above and move on to bid strategy.

I’ve always believed that effective ad strategy can be boiled down to data.  The more data you have, the higher our level of confidence that it’s painting an accurate picture. The better the data and the more well understood the data…the better the results.  It’s served me well in each and every aspect of my life and I don’t plan on changing that approach any time soon.  Any time I’ve been pitched a strategy that was based on anything but widely applicable (it may have worked for them but will it work for you), repeatable (it may have worked once but you need it work consistently) data (not theory, anecdotes or gut feel but actual numbers)...I’ve walked away without hesitation.  I’ll always listen to a concept and I’ll often even research it myself but if it doesn’t check those 3 boxes, I’m out and you should be too.

Your ad strategy then, should be based on a 2 stage approach.  Stage 1 is focussed on validating the assumptions that were developed during early keyword research specific to your offering.  If we think of keyword research and development as being analogous to the scientific method, then our initial keyword research is simply developing our hypothesis and like any hypothesis, it needs to be tested through experimentation.  No scientist worth their salt would confuse a hypothesis with proven results.  Stage 1 tests our hypotheses.  There’s a variety of ways to do this but I generally recommend 1 of 2 methods.  For beginners and intermediates, the traditional auto campaign works very well.  AMZ will market your listing to the consumers using search terms that your listing has been indexed for and presuming your listing was developed using the keywords identified during initial keyword research, those keywords will inevitably be validated (or invalidated) by the AMZ algorithm guiding the auto campaign.  Of course, we’ll have one hand on the wheel at all times as we keep an eye on what’s working, what isn’t and what we need more information on.  For advanced and expert users, a manual research campaign can take the place of the auto campaign.  This system should only be used when we have an extremely high level of confidence in the initial keyword research…and I mean extremely high.  “I know a person who’s pretty good at this.” or “So-and-so did mine and they’re a terrific author.” won’t cut it here.  To have any chance, the person(s) doing the original keyword research should be experts in that field…data analysts specializing in ecommerce.  The reason for this is that an auto campaign will not only validate assumptions, but it can identify unforeseen opportunities that were missed during keyword research…handy if that keyword research was anything less than 100%.  A manual research campaign focuses and tests only the assumptions.  It will get you where you want to go more directly but any oversights made during initial keyword research will disappear into the rear view mirror as missed opportunities.  For most people, auto is the way to go.  As mentioned in part 1 last week, stage 1 is about data.  You're essentially buying data that you’ll then use to make more informed business decisions going forward.

Stage 2 then, is about generating “effective sales” and in some cases, propping up organic ranking.  Using the now proven/validated terms and avoiding the ineffective/invalidated terms, we use targeted manual campaigns to reach the consumers using those high potential keywords and ensure that our listing(s) are optimized for those same keyword terms.  There’s a wide range of campaign types that can and will be implemented here but they all share 1 commonality…if stage 1 was done well, stage 2 becomes infinitely more productive…the opposite is also true.

When it comes to our bid strategy within a given campaign, there are 2 aspects that will control how bids are placed on your behalf…”Campaign Bid Strategy” (CBS) and your actual bid.  

CBS offers 4 options you can select from.  The 4 options are Dynamic Bids (down only), Dynamic Bids (up and down), Fixed Bids and Rules Based Bidding.  The majority of people should be using dynamic (down only) bidding.  This allows you to set your desired bid and AMZ will either bid that amount or adjust your bid down in real time when it feels your ad is unlikely to convert to a sale or when a higher bid is not needed.  Essentially, your ad will bid the amount you set or less, but never more.  This method allows you to set a competitive bid but simultaneously “stretch” your daily budget and at least in theory, never overpay for ad placement.  

Eg. - You set your bid at $0.75 and you’ll pay $0.75 or less for each click.

Dynamic (up and down) bidding is a very aggressive strategy in that it allows AMZ to adjust your bidding in real time up to double the bid you've set.  Again, very aggressive but be forewarned, this will spend money quickly and if there’s any weakness in your targeting it will result in a poor return on investment.

Eg. - You set your bid at $0.75 and you’ll pay $1.50 or less for each click.

Fixed bidding is an excellent strategy for users who have performed extensive research on their targets (keywords, ASINs, categories, etc.) to determine exactly what bid will be required to meet their advertising goals.  AMZ will simply use the bid amount set and not adjust it in any way.

Eg. - You set your bid at $0.75 and you’ll always pay $0.75 for each click.

As of this writing, rules based bidding is in beta (trial) mode and is not available for all campaign types or users.  As a general rule, I don’t recommend dynamic (up and down), fixed or rules based bidding for anyone but advance to expert users.  Stick with dynamic (down only) and although you may not achieve perfection, you’re less likely to get burned.

When it come to your actual bid…all the data I’ve ever reviewed, all the research I've performed or read, my experience with my own business and that of my students and private clients tells me that starting at or slightly above ($0.03-$0.05) the suggested bid and then adjusting that within the bid range over time (based on your goals and results) is the right way to go.

Eg. - The suggested bid at the time you set up the campaign is $0.65 with a bid range of $0.50 - $0.90.  You would then set your bid to $0.69 +/-  after which it can be adjusted up or down within the range.  You may wish to adjust it up to say $0.85 if it’s achieving an exceptional ACoS/RoAS and/or you want to be more aggressive towards that target.  You may wish to adjust it down to say $0.55 if it’s achieving a marginal ACoS/RoAS and/or you want to be less aggressive towards that target while still targeting it.

There is anecdotal data to support other strategies such as “low bid” in which a pre-determined bid of say $0.30-$0.40 is put in place (this strategy disregards the bid range and suggested bid).  There’s a great deal of complexity to the “why” of this strategy but generally speaking, the only time it actually works is when the listing and/or the specific search terms are supported by a historically high conversion rate..ie, a listing that already does very well.  In those cases, AMZ will accept a lower bid because it knows there’s a higher likelihood that it will result in conversions, which will in turn make up the lost ad revenue (think cents) with the increased fees it collects as part of the sales (think dollars).  Any time I’ve heard this argument and asked to see data showing success for new or even existing average listings…I’ve been left wanting.  I’ve also heard this strategy promoted as a method to simply generate impressions for new listings.  Although this can theoretically happen, remember that not all impressions are created equal as we discussed above.  Impressions without clicks or conversions or more generally, “low potential” impressions are about as useful as a screen door on a submarine.  Remember that your goal is to win “effective” bid placement, preferably on page 1.  A low bid strategy is like attending a traditional auction for that beautiful antique you’ve had your eye on.  When the auctioneer starts the bidding at say $100, you put your hand up as high as you can and proudly bid $30.  You may walk away with something from the auction but it won’t be that beautiful antique.  The other argument I’ve heard in support of a low bid strategy and the one that I find most difficult to justify is that “it’s all I can afford”.  I understand the need to work within a budget but buying the cheapest option whilst maintaining any hope of generating long term value is a fool's errand.  Buying the cheapest tool, the cheapest pair of shoes, the cheapest umbrella, never ends well.  Instead, find that money somewhere else in your business and invest in an ad strategy that will build long term value.

Another one to keep an eye out for is high bid strategies where bids are set above the top end of the range and this one’s easy to address. There are people out there who have a lot of money to spend on ads (think corporate advertisers) and there are people out there who have no idea what they're doing (this is actually the majority of people).  When it comes to bidding at auctions, stay away from trying to compete with either type…they are dangerous.

Remember that the purpose of ads, boiled down, is to discover the search terms that are most likely to generate consistent, profitable sales and then to use that information to both optimize your listing(s) for organic growth and target future ads for sponsored growth…that’s it.  Again, there are other more advanced…and far less advanced…strategies but for the purposes of limiting this article's length, I’ll simply say that the above will get you 80-90% of the way and leave it at that.

Next week on Purple Patch…Part 3 -  Tips & Tricks To Get You The Final 10-20%.

If you’ve got questions about anything outlined above or specific aspects you’d like me to cover in part 3, add it in the comments and I’ll do my best to address it.  As always, thanks for your time.