Many dev marketers and devrels get this question from the leadership or CEO. A fair question to ask but often extremely loaded. Especially when extended to a single activity like a blog post or a video. 

So here is my answer. 

You treat it as a product and measure ROI at a program level. You optimize activities within the programs but don’t report ROI on them. 

Ask yourself (or your CEO) if they measure product ROI at a feature level.

Do they know what is the ROI of that new dashboard or integration?

Most likely, no. 

Unless you work at a super mature org people wouldn’t know. Getting that information would be very expensive and it would slow down development. 

The vast majority of orgs understands that building a product is important and they look at ROI at a department or product line level. What I see most often is that people don’t measure “product ROI” it at all. 

So...

What should you answer your CEO?

First off, what is the question, really. What your CEO is actually asking (most of the time):

  • Do you deploy your marketing budget optimally across programs? 
  • Could you grow the marketing metrics if you have more budget?
  • Do you know how to optimize activities within your programs to get better ROI from program?

And to these, you should have answers. 

You should know:

  • What is the spend per program and blended for the entire dev marketing org?
  • What are the core metrics, leading and lagging, per program.? You should understand how those programs support blended results (brand, acquisition, activation, conversion, expansion).
  • Do you see improvements in leading activity metrics within programs to spend more? 

Once you have that you can easily answer the ROI question at the program level and say what is your plan for optimizing activities within programs or scaling spend within them.

Then....

Why not go ahead and measure ROI for each activity, like a blog post?

If you were to run brand ads to improve brand recall a single ad would likely not push the needle. Multiple ads likely would.

But you probably don’t have the budget to measure that brand recall lift. But you know this is important and that it would improve conversions across the ads you are running. 

So could you run this ad and show ROI for it? Probably not, even if you knew it would help at a program level.  

ROI and metrics should help you optimize your marketing spend to get more signups/downloads/trials. Not push you to do directly attributable activities. 

Measure at a program level, and optimize activities within programs.  

Ok, then…

How to calculate program-level ROI?

What I will do here is do a lower-bound estimation of the ROI.

It counts only the impact you directly see from the program. Not all the impact that the program might have had in other channels for example. There are tricks for that too but to keep things simple if you prove ROI on lower-bound the rest is a bonus.  

So let’s take the blog as a program we want to estimate the lower-bound ROI of and go at it. 

First, let’s think of the high-level output that this program or “product” called blog produces. Generally, the outputs that marketing programs produce are:

  • new customer ARR
  • ARR pipeline
  • Trials 
  • Signups
  • Demos
  • High-intent pageviews (pricing, terms of service, contact us, etc). 

Each of these should have some $ value to the business. You’d be happy to pay X$ for a Qualified demo if you could. 

You can estimate those $ values through conversion rates from those various events to down-funnel events like pipeline/paying/trial where you have a better feel of a $ value of it. 

Obviously, the event of people viewing pricing has a lower value than people booking a demo but there is some $ value to it. 

So for example, if:

  • your signup -> paying conversion is at 1%, 
  • paying customer is paying 5000$ a year
  • and customers stay for at least 2 year

Then the estimate for signup $ value is 1% x 5000$ x 2 = 100$.

And with say 300 signups / month the estimate of how much value your blog generates would be 300 x 12 x 100$ = 360 000$.

That is just based on signups. You should do the same for other important events (trials, demos, high-intent pageviews).

Now, you can go through all those different “product” outputs from the blog and estimate their value. 

The cost side of ROI is typically somewhat obvious to calculate. And then you divide the $value by $ cost and here you have it. ROI of your dev marketing program. 

Measuring ROI like this at a program level gives you clarity on investment without shooting yourself (and your content team) in the foot by pushing for reporting and attribution over impact. 

Even if you "waste" half of your budget on the unattributed impact you spend the other half with a better ROI cause you focus on measuring the impact of the program/campaign that you believe in rather than doing things that you feel are not the highest impact just easiest to report on.

Resources