How to optimize campaigns for profit?
Updated: Jun 26, 2020
In most definitions of business is mentioned the goal of maximizing profits. Most marketers are guilty of ignoring this goal and focus on ROAS (Return on Ad Spend).
What is ROAS? It is simply calculated as Revenue / Media Cost = ROAS
This is not in line with the business goal and business owners learned to accept this view.
There is a way to align these views and optimize marketing towards maximizing profits.
What numbers can you see with ROAS and revenues?
As you can see, the higher the ROAS, the better the result. It is quite easy to understand, right?
We need to include also absolute revenue, so we can see the revenue impact of each channel.
"We should kill off ROAS, as it is really misleading."
It takes into account only two inputs (Media Cost & Revenues) and makes our marketing activities look better.
How about discounts, shipping cost, other media costs like TV, returns, cost of goods sold? Now you can start seeing the issues with this metric.
Cost of Goods Sold
If you consider all the cost of producing an item (or purchasing it from the manufacturer), you get the cost of goods sold.
Fully Loaded Media Costs
Marketing campaigns have media cost but there are fees for a marketing agency, cost of all the tools you use, cost of producing banners or videos, etc.
We still did not include costs tied to operating the business. Everything that is needed in order to keep the light on is included in these costs.
Add everything up
Now that we have all the needed data, we can actually calculate profits.
In the table below you can see Costs assigned to each channel and calculated Profits.
Profit = Revenue - Media Cost - Other Costs
There is one issue. Not all costs can be assigned to a specific channel. What should we do in this case?
Evaluate products, categories, and manufacturers, not channels
This may be controversial, but we should change the way marketing is evaluated. Reporting on the channel level is not showing us enough insights.
What information do business owners need more? Which channels work better or which products work better?
Consider this. You use the channels mentioned in the above tables. Can you tell which products are stars and which are losing money? Not really.
How do you evaluate this level of profitability?
This level of reporting is not trivial but the value of this data is well worth it! Here are some basic steps you can do:
Start with Google Analytics (or other reporting tools you use) and include profits instead of revenues, Expert hint: You can create a second Property for Google Analytics and create a separate event instead of Purchase, during which you use dataLayer.push in order to change dataLayer Revenue data to Profits.
Gather all costs for each item. This comes to all the products which have some costs. They include Media Costs, Shippment, Warehouse Costs, etc.
Divide all other costs like Operation Costs, General Media Costs (TV ads, Brand Building Ads), etc. between all the items you have available. Calculate the Profits of each item, category, manufacturer or just overall Profits.
Now you are able to calculate the profits of each item, category, manufacturer or just overall profits.
What data do you get as a result?
Now we have way more useful data than before. Here are some insights you can get:
ROAS is misleading. Item #4 has a better ROAS than Item #2. Yet this item is losing money. All thanks to the Other Costs that are now considered in the calculation of Profits.
Some items are losing money. Item #6 is losing money and there seems not to be a demand. As you can see, there are still some Other Costs associated with this item, yet there is no Revenue. This means we should reconsider this product. Maybe there is a strong seasonality, this item is outdated or some other reason.
Some items have no Media Costs, yet they have Revenue. This may be a case where this specific item does not have direct Media Costs, just like Item #3. On the other hand, purchases of this product could be a result of the Branding Campaign, which is included in Other Costs. Or simply, this Revenue was not affected by Media Costs at all. This would show us a great opportunity to make this product a star of our next campaign, as there is demand even without direct media costs.
This is possible on a level of products, categories or manufacturers. Having this data will allow portfolio managers to make informed decisions.
Consider adding order returns to the mix and multiply the profit of products by the average return rate of this product or category.
Let's say you would enhance this data by adding seasonal trends of each product. This way your portfolio manager will have a solid prediction of demand for each product.
Doing this allows you to predict not only the demand for the following months but also to predict potential profits.
What are the pitfalls?
There are always some cons to a performance-oriented system. You can see we are considering Branding Campaigns in our calculations, so this is slightly covered.
Users buy different items than those that brought them to your store. This a common behavior. In some businesses, it can affect around 20 % of Revenue,
Upsells are not included. If an ad for a specific item results in the purchase of multiple items, we do not cover this case in this system (for now),
Some items are required in your portfolio. We can't change human nature. Even though some products have such a low margin they will be never profitable, we simply have to include them in our portfolio. A great example is a product with a recommended reseller price.
Consider moving towards measuring profits and showing marketing results this way. Not only will be marketing managers more accountable for the results, but business owners will also see more accurate data.
What are your thoughts? Do you use profit reporting today or plan to move this way? Let us know in the comments below.