I started optimizing Amazon PPC Campaigns almost 10 years ago. Here’s how Amazon Advertising really got started.
Some History on Amazon Advertising
When Amazon first rolled out its advertising platform, it was segmented as it is today. Meaning, they had Amazon Product Ads which let people direct traffic to their Shopify or standalone eCommerce stores. I was able to tap into Amazon traffic without having a full blown Amazon store. I was able to upload data in a spreadsheet, that would create a listing, but when people would click “check out,” they were actually taken to my product page.
As far as I know, this is and was the only way to get 10%+ conversion rates on a standalone eCom site. Amazon viewed this as a way to collect some revenue for products from sellers that weren’t on their site.
Over the past 12 months, Amazon has made $240 billion in revenue, according to macrotrends.net. But way back in 2010, it was more like $30 billion. Believe it or not, there was a time where Amazon sales didn’t take up half of all online transactions.
Amazon shut down Product Ads in 2015, and I remember all of my ecommerce clients taking a massive hit. For many, it was their number-one converting channel. Amazon knew this, and around this time of 2015, started prioritizing its own channel.
Now today, we have ad types like Product Targeting where you can target individual ASIN’s, we have Category Targeting, Brand Targeting, stack those with price and star rating filters, keywords with broad, phrase, exact and modified broad match types, Sponsored Brand Ads, Display Ads, different bidding options for Top of Search, rest of search, product page impressions, close match, loose match, substitutions, and the list goes on.
However, there was a time before all of this complexity.
Amazon actually introduced exact match and phrase match in mid 2015. Yup, at one point in time there was just one match type: broad. In fact, if you do some googling, you can see people talking about the addition of match types and negative keywords as completely NEW in 2014 and 2015.
Amazon Advertising started off very simple.
Then, over time, they took cues from Google Ads.
Match types were introduced.
Auto and Manual Campaigns as we know them today.
Negative Keywords added.
Additional ad types.
More data, better data.
And it got more complex, FAST.
And this brings us to 5 years after the arrival of automatic campaigns. We’ve got more features, rolling out faster. Vendor Central Ads are still separated from Seller Central Ads. There is still a big divide between the two.
There is also a massive divide between Seller Central and Vendor Central campaigns within their own platforms.
For example, in Sponsored Brand Ads, you have some metrics, and some features unavailable in Sponsored Product Ads, and vice versa. For example, still no search term report for Sponsored Brand Ads. There is also “New-to-Brand” metrics for Sponsored Brands, but not for Sponsored Products. There are plenty of advertisers that still don’t have self-serve access to Amazon’s DSP (Demand Side Platform) or simple retargeting or simple video ads.
Which brings me to my first prediction.
Prediction #1: There will be a better unification between all ad types, but it will get worse before it gets better.
I think, eventually, we’ll start to see familiar metrics and features get unified with Seller Central and Vendor Central ads. Now my hope is that they actually unify everything, and have the exact same interface regardless if you’re a 1st party or 3rd party seller.
Prediction #2: There will be a sharp rise in dimensions.
Let me give you an example of dimensions to help you best understand it.
Let’s take 24 hours in a day. Then 7 days a week. Then male/female genders. Then let’s take age range buckets from 18 to 80 and divide it up into 6 buckets. Let’s take household income, and divide that up into 6 buckets too. And for fun, let’s also add top of search, rest of search, and product page placements. And just because we’re all crazy PPC’ers here, let’s throw in desktop, tablet and mobile.
Now let’s meet all those dimensions with one keyword. Now currently, there is no way to measure a keyword’s performance across hours, days, genders, age ranges and income. You get only one bid, that when optimized, is based off the average of all those dimensions.
Now, let’s say you had the ability to study data on different times of day, days of the week, and all those other dimensions. Imagine you were able to pull up an ACOS report on a keyword based off of gender or income. Then you were given the ability to set bid modifications for each of those dimensions.
All of a sudden, we go from bidding on just one keyword to bidding on:
- 7 days
- 24 hours
- 2 genders
- 6 age buckets
- 6 income buckets
- 3 placements
- 3 devices
Now we have 108 thousand different dimensions all for ONE simple keyword. You might want to bid more on Fridays between 5-8pm to men aged 25 with above-average income who are searching on their phones. But bid a little bit lower on other dimensions. Now multiply this by 100 keywords. That over 10 million individual slices from what is currently just 100 bids.
I think the data availability, audiences, and the ability to set different bids will improve over time, and this is great news for everyone reading this post, because it will allow us to outpace and outcompete our competition that isn’t leveling up their Amazon PPC skills.
Prediction #3: Amazon will want to pick your options for you.
With 10 million slices of data points for just 100 keywords, you can easily see an “auto-optimized” version. We see this in Google Ads today, we see it in Facebook Ads as well. I think we’ll see more of these “auto-optimized” features in Amazon.
However, just like with Google and Facebook, it’s incredibly easy to spend money with reckless abandon. It’s too easy to have a campaign go off course. It’s just too easy for these kinds of runaway spend areas take us away.
It’s important to remember that there are goals that Amazon has, and goals that you have, and more often than not, they’re not the same. As Ad Spend increases on Amazon, and as advertising becomes more valuable to Amazon, Amazon will be under more and more pressure to keep increasing Ad Spend, just like Google does. Which means that “auto-optimized” may mean “higher ACOS.”
Does Google or Amazon or Facebook care if your ACOS is 10%, 15%, or 20%? Not exactly. This is where I believe 3rd party tools like Ad Badger can help keep campaigns well-optimized according to your goals, not Amazon’s.
Prediction #4: Targeting and retargeting audiences will become a lot better.
Here’s how I see it going.
I think in the future you’ll be able to target runners, or weight lifters, or gardeners.
And here’s the big thing: It won’t be based off the posts they engage with, or the sites they visit, like it is on Facebook. It will be based off what they buy.
Imagine being able to target people that you know spend hundreds or thousands of dollars in your niche. Imagine if you know you can target an ad towards men who always buy items for their wives or girlfriends on Valentine’s Day.
Imagine if you were able to even target customers who are likely to switch brands? Perhaps bid less on customers who are less likely to switch? Imagine being able to directly target customers of competitors’ brands?
Imagine being able to bid more aggressively on previous visitors or previous purchasers, and less aggressively on cold traffic? Or vice versa.
I think all of these features are coming down the pike. More dimensions, more audiences, more complexity.
My fifth and final prediction is a hard one to swallow. But I have to say it…
Prediction #5: CPC’s will rise, Conversion Rates will fall, and ACOS’s will go up.
And I think it will impact smaller budgets more.
In general, you see a trend, that the campaigns with higher Ad Spend also have higher Conversion Rates and lower ACOS. So I think large spenders with massive budgets and huge sales will be insulated from a lot of this. Their star ratings and history will carry their momentum.
However, even large companies will feel a rise in CPC’s. The way this generally works, as platforms get more competitive, the CPC’s go up. That’s fairly standard and it’s already happening, and I think it will continue to happen.
Another thing, that I don’t believe has started happening yet, is an artificial standardization of CPC’s.
Here’s what I mean by this.
If I go into Google and bid on a keyword that NOBODY else is bidding on, let’s say “jump ropes for my pet dinosaurs” – I can’t bid 3 cents and win the auction. Instead, this search will be picked up by phrase match jump ropes. That’s fine, but not what i’m talking about here.
But what about terms that nobody is even advertising on. What if I’m the only advertiser for the auction? Can I win the auction with a 3 cent bid?
No.
What happened on Google was they ended up setting minimum bids for all auctions. On the Facebook side, you have minimum CPM’s. If there was an audience or market that ran wild and was able to get loads of clicks for cheap, Facebook realizes they can artificially inflate the CPM’s and still get traffic.
I know, this sounds like tinfoil-hat conspiracy, but it’s what I’ve seen through time.
So today on Amazon, there are loads of keywords and people still getting clicks for under 25 cents. Those days are numbered. Over time, CPCs will increase either by real competition or artificially. Let’s call it “CPC inflation.”
The Bottom Line
To recap, I predict that in the next few years there will be better unification among ad types, more complexity with dimensions, better audience & retargeting options, Amazon suggesting their auto-optimizations by default, and a rise in CPC’s resulting in a rise in ACOS.
To cope with that, we’ll have to continue to stay on the cutting edge, do things our competitors won’t, and remember to always be optimizing.
Good luck everybody. I’m looking forward to going on the ride with all of you!