E-commerce Enhancements and Opportunities: Optimizing to Profitability

When measuring the success of E-commerce campaigns, it’s easy to lose site of your end goal by focusing on the myriad secondary and tertiary metrics on offer for assessing sales performance. This is a shame because the goal (and end key performance indicator) of most e-commerce businesses is simple: profit.

While this may seem obvious, it can be hard to stay focused on this end goal when you consider how many different options you have for gauging success in AdWords and Analytics — optimizing to pure profit.


CPA, Conversion Rate, Conversion Volume, ROAS = The Wrong Metrics


  • Cost Per Conversion –While CPA is an important tool for optimizing your campaigns, it doesn’t tell the whole story if you sell various items at different price points. For example, a CPA of $10 on an item that sells for $20 is not preferable to a CPA of $20 on an item that sells for $50. Your CPA doubled but you made more profit on the sale of the higher value item.
  • Conversion Rate — It’s a great metrics for optimizing ads, keywords, landing pages, etc., but it doesn’t tell you anything about how much revenue you brought in from the sale, or how many sales were made from the element you’re optimizing. For example, a conversion rate of 50%, at a CPA of $50, on an item which sells for $40, lost you money.
  • Conversion Volume – We’re moving in the right direction, but this still doesn’t help us, is we’re selling items with varying values. For example, selling 100 units of item A (valued at $5 each) brings in $500 in revenue. Selling 10 units of item B (valued at $100 each) brings in $1,000 in revenue.
  • ROAS (Return on Ad Spend) – We’re getting very close, but still not there. ROAS = Revenue / Ad spend. This metric actually uses revenue data to give you a ratio for how much you made, relative to how much you spent. The problem is that it’s a ratio. If you make $40,000 in revenue from $10,000 in ad spend, that’s an ROAS of 4:1. However, $200,000 on $100,000 in ad spend is an ROAS of 2:1. Which one would you prefer?



Profit (as a whole number) = The Right Metric

Now let’s look at a simple calculation: Revenue – Cost = Profit

This will simply give you the real amount of money you made from your advertising spend. You can’t deposit a ratio into a bank account. Assuming your goal is to maximize real profit, you should be using that as the end KPI for measuring success.


Use Secondary Metrics in the Right Context


So, does this mean you should ignore all the secondary metrics mentioned above? No. The trick is not focus on a single metric, but rather to understand that these metrics are only a means to a larger end: profit. They are all extremely useful in the right context, but remember that they’re all just indicators.


Taking Action on Profitability


Take a look in your AdWords account. Test some assumptions. Look at your keywords with the highest ROAS and consider this question: (Be sure to have revenue and cost data columns viewable)

How much money are you making on each keyword?

Getting this information is simple. It come be done manually by just subtracting cost from revenue, or better yet, do it in bulk:

  1. Download your data into Excel
  2. Create a new column and title it “profit”
  3. In the first field of the new column, subtract the revenue field from the cost field (=SUM)
  4. Drag corner of the field down entire entire column to get this calculation for each keyword


Now, compare the ROAS field to the profit field. Changes are, your top ten ROAS keywords won’t be the match your top ten “profitability” keywords. This is due to three things:


  • ROAS is a ratio, so it doesn’t look at how much money was actually made. You might have great ROAS on keywords with barely any conversion volume and representing a minimal  amount of revenue.
  • Your products likely sell for different prices. Because specific keywords in your account likely correspond to different products on your website, some low value sales may fetch a high ROAS without actually bringing in much total profit.
  • There usually isn’t a 1-to-1 relationship between keywords and products. Much of the time, the keywords that drive a large amount of your conversions are general and can’t be tied to a specific product. This is where and average order value comes in handy. It’s harder to predict performance of general keywords because they generally point to less specific pages, from which the user navigates to more specific product pages.Therefore, it’s common for these keywords to have ROAS stats that don’t always correspond to profit stats.


What To Do With This Information


  • Test Lower ROAS against Higher Conversion Volume –If you that certain keywords reliably correspond to specific products (low funnel terms), you can set uptests that push your ROAS. Raise bids to find the sweet spot between ROAS and conversion volume which maximizes total profit.
  • Use Average Order Value for Higher Funnel Terms –If you have a lot of data for your higher funnel terms, you can use average order value and test in the same way. This takes into consideration the fact that many different products may be purchased as a result of a user clicking on this keyword set, but still gives you a reliable number to use as a guide.
  • Use Profit Per Impression (PPI) for Ad Testing Profit per impression is a way of optimizing to profit at the ad level. It forgoes all of the other secondary metrics such as CTR, impressions, clicks, etc., and focuses on the end goal, which allows you to make better ad optimization decisions. See my recent post on optimizing to PPI.



  • Optimize to profit using a simple formula: Revenue – Ad Spend = Profit
  • Optimizing to profit means optimizing to your real business goal  
  • Don’t forget about secondary sales metrics, but use them in the right context
  • Don’t assume success based on the wrong metrics
  • Take action: test lower ROAS against higher Conversion volume to find Profit sweetspot

References and Resources: