Amazon PPC – ACoS, Suggested Bid Manipulation, Automation

[et_bloom_inline optin_id=optin_3]

Sean Smith comes back to update us all on Amazon PPC and sponsored ads along with some magic on manipulating suggested bids…

  • New reporting and a graph has been released
  • Sean saw them and they are really cool
  • Gives get a Lifetime view
  • You can manipulate the date ranges
  • Can compare two metrics
  • You can look at trends
  • It saves so much time
  • Click into an ad group and you can see the graph too so can view at campaign and ad group level

Data go-tos

  • Always go to ACoS as it’s the one people are most concerned with
  • The reverse of the Return on ad Spend ROAS
  • Start off at ACoS and Sales and have targets for those
  • Impressions are good for real-time data
  • Sales don’t roll in immediately but impressions, clicks and spend show up closer to realtime

Campaign performance

  • These indicators give you view on whether you’re using good keywords
  • Visibility is clicks and impressions which are important to get to the bottom of the funnel
  • Looking at the graphs and seeing an increase or decrease ten you can dive in to diagnose and take action
  • Conversion is king and drives sales
  • Conversion rate is a key metric for Sean
  • Top influencer of PPC is the difference between the price point and the average cost per click
  • Also measure:
    • Price point
    • Margin
    • Average cost of click in the niche
  • You can say to yourself then: I have a low price point item, my average cost per click is high and I have issues
  • So you need to increase your conversions to decrease that ACoS

Benchmarking ACoS

  • What’s an acceptable ACoS across the board
  • Is 20-30% a good happy level
  • Around 20-40% is the  median trend to target
  • Looking for roughly a breakeven-ish point on cost of running the campaigns
  • You want to cover those fees for ads ideally
  • Really big sellers value their time rather than the money
  • People get obsessed with ACoS
  • Overarching goal is to achieve Sales velocity on that product not just the current ACoS number
  • Go beyond ACoS and look more broadly at the product horizon

Why people use a PPC service

  1. Sellers don’t have the time to manage it themselves
  2. They may not be able to manage it themselves
  3. They want someone to do it but cheaply (not necessarily the best)
  4. Want access to an expert to run it for them with tools
  • Sean aims at the higher end of the market with bigger brands


  • Sean was on the phone with Amazon last week
  • Learned that when you’re looking at suggested bids you are given three values:
    • Low-end – 25% percentile
    • Actual suggested bid – 50% percentile
    • High-end – 75% percentile
  • The algorithm for each value is driven by percentiles
  • So for the low-end you have a 25% chance of winning the bid
  • Sean has always based their bidding on this model anyways
  • You can apply the suggested bid to all you keywords, then select them again and increase the bids by an additional % above the suggested bid

Automation of Amazon PPC

  • Sean has just enabled the auto-creation of launching auto campaigns, just copy and paste your SKU and hit create campaigns
  • Naming system that Names the ad group, the campaign and puts in the budget
  • Only the bids have to be changed as the API does not provide this
  • This is way faster to do now
  • Setups up three auto campaigns with waterfalled bids

Data and Key Dates

  • No real uplift through Q4 that Sean has seen so far
  • Pretty slow across the board currently and quite flat
  • Did Valentine’s Day have any effect? No
  • Golf space is trending up in March so Sean will be on that one
  • Summer items and getting used to the new reports is the focus

Manipulate the Cost Per Click

  • Trying to manipulate the cost per click at the account level for the bigger accounts
  • Look at your price and your cost per click to tell you how much you can afford to spend on your ads
  • If you have a big portfolio with a lot of campaigns then it gets a bit hectic with 2,000 keywords being tested across 10 products
  • Sean successfully brought amazon ppc cost per click down from $1.78 to $1.23 within 24 hours across a couple of hundred campaigns nad 40 products
  • Price point was $18 and the only way to bring down ACoS was by bringing down average cost per click
  • Had to go tight on cutting keyword bids

Technical Approach

  • Found weighted average price using the business reports and then the weighted average cost per click on month to date data
  • Weighted average price is $20 and weighted average cost per click is $1
  • You can only do a 20% ACoS
  • So they only had $4 to spend on clicks
  • So if they can bring it down to 75 cents allows more clicks
  • Sean was slicing and dicing the campaigns to ensure no bid was above a dollar
  • Bid is always higher than the cost per click
  • Rule: Nothing at portfolio keyword level should be above weighted average cost per click (a dollar plus 20%) UNLESS it is within our target ACoS then we leave it alone
  • If it has zero sales then you cut it
  • This worked really well and managed to get daily spend down and ACoS down for a couple of his accounts
  • Took a lot of time to work it out and make it happen

Day Parting

  • Day parting tested with 3 accounts but not seen any results doing this
  • Changing bids nad day-parting bids (lower at night, higher at day time)
  • No conclusion on this yet from Sean

Contact Sean