TMAI #247: 💰 Your Outcome Metric is Profit, not Revenue!

The Marketing < > Analytics Intersect, by Avinash Kaushik

This edition of TMAI Premium is free for Standard Edition subscribers.

We've had a spectacular start to 2021 in Premiumlandia. 243 was a fun NFL data viz exercise (with a Nest Home as a prize!). 244 was a 90-day prescription to re-launch your career (new or current job). 245 I felt was my best newsletter ever to save you tons of money (why does so much advertising fail). 246 you told me was the best one I'd ever written (!) (a playbook for small business success - which applies all-sized companies).

You can upgrade to TMAI Premium here - it comes with a money-back guarantee. All Premium revenue is donated to charity.

As a regular reader of my blog and this newsletter, you are (likely) not someone who gives much credence to metrics like Views, Impressions, Followers, “Social Halo” (whatever that is).

Furthermore, I feel confident saying you have moved your organization beyond the silliness of last-click attribution, you are really obsessive about Cart Abandonment Rate and Checkout Abandonment Rate, and your dashboards are reporting (segmented!) Orders, Average Order Value, and Revenue.

Some of you even measure Economic Value, one of my most favorite KPIs.

This is wonderful.

It makes me happy that as a leader, or practicing analyst, you are giving your organization an opportunity to shine by obsessing about metrics that actually matter. #dievanitymetricsdie

I think we can take your business success to a whole new level.

[For optimal learning, I recommend printing out the report below and reading it with a pen handy to take notes. It’ll be fun, I promise.]

A typical custom report that serves as a starting point for your analysis possibly looks like this one from Google Analytics....

Google Analytics | Acquisition Channel Report

It has a lovely end-to-end view across Acquisition, Behavior, Outcomes (you learned in TMAI #246 why I highly recommend this!). The report also has all the good for your health metrics.

[Pet peeve: It has always irritated me that Google Analytics chooses not to eliminate false precision and displays double decimal points. As in 49.18%. Or $42,805.34. Learn why it is important to not make this mistake in your reports: TMAI #127: Smarter Reporting, Deeper Influence. If for no other reason than that you won’t irritate me!]

The report above, lovely as it is, is not the destination.

Ending on Revenue can be a very misleading destination for an Analyst. It holds the potential to make suboptimal - or even hugely wrong - decisions.

The simplest reason:

The same amount of company effort did not go into each row to earn the revenue in that row. Hence, a direct comparison of performance across rows is imprecise - even foolish.

Let’s dive deeper by comparing the two Search rows, Organic and Paid.

You had to pay for every single one of the 44,575 people who showed up via Paid Search. (33% of whom left without doing anything, ouch!)

You had to pay nothing for the 116,278 people who showed up via Organic Search.

Ergo, it is not a stretch to say that it was cheaper to earn $126k via Organic Search than $337k via Paid Search.

[Note: Organic Search is not free. The folks in your team working on SEO cost money. The SEO tools you might have purchased, like Semrush, deservedly, cost money. You might even have an SEO agency, which, well, costs you money. These people, tools, agency costs also exist for Paid Search and tend to be higher. Tacked on top of that are costs per click.]

As things stand in the report above, an Analyst might recommend exploring how much further you should take Paid Search since it clearly looks amazingly wonderful.

What is the analyst missing? Accounting for the cost of Acquisition!

Let’s say, 44,575 people arrived from Bing Paid Search ads, at an average cost of $5 per click. That could compute to a total cost of Acquisition of $223k.

Let’s redo our revenue calculations:

Revenue from Organic Search: $126k
Revenue from Paid Search: (337-223) = $114k

Is Paid Search still your first obsession?

Maybe not.

At the very minimum, you need to not rush your chq to Bing/Google/Baidu/Seznam.

I suspect you are beginning to see why I’m so deeply obsessive about getting as close as one possibly can to computing Profit. It helps ensure you are getting the most intelligent assessment of value. Stopping at Revenue is a big mistake (that we all make every single day of the year!).

There is a lot that goes into computing real Profit for the company - think of what your Accounting team does and reports to the SEC once a quarter. It can be hard to get to that level of detail. Some of that data might even be sensitive (only available to some teams in your org).

You don’t need to go that far for the concept of Profit to help us be smarter Marketers.

If I simplify things a small bunch, here’s a workable formula that’ll fuel smarter decisions:

Profit =
[Revenue] -
[(Creative Costs) + (Campaign Media Cost + Humans + Fees) + (Cost of Goods Sold)]

Let us look at each of the components of this sexy formula.

Creative Costs.

For many of the channels, the cost of the creative can be as much as the media costs.

A Super Bowl ad costs $1 mil to air but when so many advertisers go all out, it is not uncommon for the creative costs to be multiple times that sum. As an Analyst, you need to account for that.

Ex: One of my all-time fav Super bowl ads, Coca-Cola’s America is Beautiful cost $8 mil to make. Another one of my favs, Chrysler’s Imported from Detroit, cost $12.4 mil.

When you calculate the return on investment, it would be a dereliction of duty to compute ROI of the $1 mil in media cost. It is important to take creative costs into account ($1 + $8 for Coca-Cola).

Creative Costs can contain these elements:

Talent Costs + Production Costs + Agency Fees + Reshooting

If you are using Smart Creative options available on many platforms, like Google Ads, then it is possible that your creative costs are lower.

Still, even there you have to supply logos, headlines, images, descriptions, promotions, etc. Count up the costs of all that, and don’t forget the cost of the humans churning these out, and subtract total from the campaign Revenue.

What’s more likely is that you have a team of 2 - 14 people working with 1 - 3 different agencies and 1 dedicated analyst (for creative testing, and redoing creatives after pre-testing results) to produce creative for digital and non-digital campaigns. This is a massive cost that often gets ignored.

Social ads often requires a bunch of custom display and video creative. For a small company the Talent + Production + Agency + Reshooting costs = $20,000 per month.

In our Google Analytics report above, the revenue from our Social Campaign was $1,285.

Let’s compute Profit:

Social Campaign Profit: $1,285 - $20,000 = - $18,715

Not that Social was looking spectacular in our report, but now I dare say you are a little freaked out about your ROI. That's good for your health.

Creative costs are just the first step of the journey, hence the “Profit” above is preliminary.

We still have to subtract: Campaign Cost & COGS. Let’s continue our journey.

Campaign Media Cost.

Getting the Campaign Media Costs you paid to the Acquisition Channel, as I did above for Paid Search, is less difficult than you might imagine.

For example, if you link your AdWords account to your Google Analytics account, these costs flow into your reports automatically. If you’ve done so already, in Google Analytics go to your Acquisition > Google Ads and Acquisition > Google Marketing Platform sections. You’ll see Cost and Cost Per click waiting for you there.

To make the reporting efficient, you can easily create a custom report and add to it a Revenue minus Campaign Media Cost calculated metric in GA. The resulting number is a little bit closer to Profit, and it takes five minutes.

Do this.

Not all your campaign costs can flow into Google Analytics as easily, ex. Bing or Facebook ads or Billboards or Television. Your entire business is likely not digital either. Hence, you need to move to Google Data Studio, or your company’s internal Business Intelligence tool, or, if you are feeling particularly adventurous, good old Excel to ensure you have Campaign Cost being accounted for when you look at marketing performance.

Start your Campaign Cost with a metric like Cost Per Click, but you don’t have to stop there.

You can also include the cost of the humans involved.

Ex: Paid Search: 7 FTEs. Organic Search 2 FTEs. Email: 0.1 FTE. Social Media: 46 FTEs. etc.

You don’t need to go for deep precision.

Let’s say the average cost of each FTE (full-time employee) is $100k. Now do the simple math.

Paid Search Campaign Acquisition Cost now is:

(44,575 * $5) + (7 * $100,000) = $922,875

Your Paid Search Revenue was: $337,280.

Your preliminary at this stage Paid Search Profit is: - $585,595

Obviously, this is a dramatically different reality than you saw in your Google Analytics report. Your heart probably feels like someone put a heavy boulder on it. :)

Now, I do realize that you are probably not losing that much money and still continuing to spend on Bing. But, if you’ve not done the computation, how do you know?

One final element that applies to many medium and large-sized companies is Additional Fees.

This bucket can include the fees that your SEO agency, Paid Search agency, Social agency, Television agency, and other Agencies are charging you. It can also include the licensing fees for tools you might have purchased or built internally.

Your Paid Search agency charges you a flat fee of $40,000/month. You are also billed $80k/year for using a proprietary tool, which you can amortize to $6,667/month.

Your new Paid Search Campaign Acquisition Cost is:

(44,575 * $5) + (7 * $100,000) + ($40,000) + ($6,667) = $969,542

Your preliminary step two Paid Search Profit is Revenue minus Cost:

$337,280 - $969,542 = - $632,262

A new, incredibly revelatory, view of Paid Search Marketing.

Isn’t the Profit lens amazing?

Do this for every single row of data that shows your online, offline, non-line marketing across Owned, Earned, and Paid Media.

I guarantee, even if you do this computation at an aggregate channel level and only do it once a quarter, it will freaking blow your mind.

And. Your boss will love you.

[Note: I’ve ignored the creative costs for the Paid Search campaign above. If these costs are material in your case, as they very well might be, please include them in the Profit computation.]

Cost of Goods Sold.

By now you should have a very strong sense that Revenue is not all Profit. The most obvious cost still remains to be computed.

Cost of Goods Sold (COGS) is what it costs you to produce the product or service that you are selling.

In this case, let’s say, we sell a product for $50 and the cost to produce it is $25.

To earn the revenue of $337,280 from our Paid Search campaigns, we sold 6,746 units.

The cost for you to produce those units was 6,746 * 25 = $168,640.

With this number, and all others we’ve so lovingly calculated above, we are ready to compute the total cost of our Paid Search campaign:

Creative cost + Campaign cost + COGS = Total Paid Search Costs

$0 + $969,542 + $168,640 = $1,138,182

With this number, we are almost there.

[And, I want you to pause for a moment and ponder just how much accountability is missing when you only look at Revenue in the standard Salesforce or Google Analytics or internal business intelligence reports.]

Final Marketing Profit.

I bet you are giddy with excitement at this point. Me too!

Here’s the final Profit for Paid Search:

Total Costs - Total Revenue = Final Paid Search Profit

$1,138,182 - $337,280 = - $800,902

Now, the company can have an intelligent decision about its marketing.

In the numbers used in this newsletter, obviously I’m trying to make a point.

It is unlikely that you are losing that much money and still doing Paid Search with Bing.

But, how do you know until you know?

Compute Profit. Report Profit Everywhere.

I recognize that our Accounting department has other nuances it has to account for like office buildings, corporate debt, etc.

In our computations above, we’ve skipped little things like taking into consideration discounts, shipping, and fulfillment costs, which can have a material impact on Profit.

You can add all this stuff in as well to get an even more precise picture.

To get going though, to freaking blow your mind, you don’t have to. You can use the formula

I’ve recommended above and get roughly close to Profit.

Here’s our original report, with Profit included. See if you notice the different view...y.

Bottom line.

Simple: Give me Profit, or give me (metaphorical) death.

Go!

-Avinash.

PS: If you don’t have the above level of precision in your analysis thus far, don’t feel too bad. This past weekend, I had a little bicycle accident. Cuts and scrapes everywhere. The big hit was to my left shoulder. My wife insisted the next day we go get an x-ray. Imagine how scared I was when I saw the controls on the x-ray machine were marked on paper with a sharpie!