I have sat in a lot of meetings where a slide said reach was up and everyone nodded. Nobody asked what it bought. I run Robbp Media, a social media marketing agency in Los Angeles, and the way I keep myself honest is to tie every account to one number the client would recognise on a bank statement. For Dead Soxy that number was return on ad spend, and it moved from 2.23x to 4.64x after we rebuilt the creative and the audiences. For a Beverly Hills medical client it was 1,233% ROAS on paid social we built and managed. For Dr. Zoumalan it was consultations booked from an Instagram account that grew from 6,000 followers to 109,000 without a dollar of paid growth. Those are social media ROI numbers. A reach chart is not.
Here is how to measure social media ROI without fooling yourself: the math, the mistakes I see most often, and the way we report it to clients at Robbp Media.
Start with what a customer is worth
You cannot measure return without knowing the value of the thing you are returning. The first question I ask on a call is what a customer is worth to you over a year. A plastic surgeon knows what a procedure is worth. A pet food brand knows the average order and reorder rate. A nonprofit knows what a monthly donor gives over a year. If you do not have that number, estimate it and refine it later. Every ROI calculation downstream depends on it.
Then add up what social actually costs you: the retainer or salary, the ad spend, the creator fees, the software, and a fair share of your own time if you are still in the comments every night. People leave out their own time and then wonder why the ROI looked better on paper than it felt.
The math is simple, the inputs are not
Social media ROI is revenue attributable to social, minus the cost of social, divided by the cost of social. For paid social we use ROAS instead, which is simply revenue divided by ad spend, because the management fee is fixed and the spend is the lever. Dead Soxy at 4.64x means every dollar of ad spend came back as four dollars and sixty four cents in sales.
The hard part is the word attributable. Platform dashboards claim credit for every purchase from anyone who saw an ad that week. Google Analytics credits whichever channel the person touched last. Your CRM may say something else again. None of them are lying, they are answering different questions. Pick one source of truth, state the attribution window, and keep it consistent month to month. Consistently approximate beats randomly precise.
Three ways people fool themselves
The first is counting vanity metrics as results. Followers, likes, impressions and reach are inputs. They are early indicators that the content is working, and I watch them closely, but they are not the return. Dr. Andrew Brunswick going from 9,590 followers to 115,000 is a result I am proud of, and I still would not call it ROI until I can connect it to patients walking through the door.
The second is ignoring organic because it is harder to track. Paid social hands you a ROAS number every morning. Organic does not, so people assume it is free and therefore infinite ROI, or that it is unmeasurable and therefore zero. Neither is true. Put a tracked link in the bio, use a booking page that asks how people heard about you, run DM automation through ManyChat so inbox conversations leave a record, and give organic a fair accounting.
The third is reading a single month. Social compounds. The Zoumalan account did not go from 6,000 to 109,000 in a month, and the consultations followed the audience with a lag. If you judge organic on a thirty day window you will kill it right before it pays. I ask for ninety days to prove the system works and I judge the trend across quarters, not weeks.
How we report it
Every client gets a weekly note from me that reads like a text from a friend rather than a dashboard. It names the one number we agreed on, where it sits, what moved it, and what we are changing. For paid clients that is ROAS against the target we set before anything went live. For organic clients it is leads or bookings traced to social, with follower growth and engagement reported underneath as leading indicators. Posh Peanut scaled to $41M in revenue with our social, creator and paid work underneath it, and that relationship lasted because the reporting was always about revenue, not how pretty the grid looked.
When the number is not moving, I say so. Every engagement here is month to month after the first ninety days precisely because I think an agency should have to keep earning the retainer with results you can read. If leading indicators climb and revenue stays flat for two quarters, the conversion path is broken, which is fixable. If both are flat, change the strategy. If both climb, spend more.
Questions people ask
What is a good social media ROI?
It depends on your margin and your customer value. For paid social I set a ROAS target with each client before launch, and anything from 3x upward is usually healthy for ecommerce. Dead Soxy at 4.64x is strong. For organic, the goal is a cost per lead that beats your other channels.
How do you track ROI from organic social?
Tracked links in bios and stories, a booking or checkout page that asks how people found you, and DM automation through ManyChat so inbox conversations leave a record. None of it is perfect, but it gives organic a fair accounting instead of guessing that it is either free or worthless.
Are followers a meaningless metric?
No, they are a leading indicator, not a result. Dr. Zoumalan growing from 6,000 to 109,000 followers organically mattered because consultations followed the audience. Watch followers, reach and saves closely, but report them underneath the revenue number, never in place of it, and never as the headline.
How long should I wait before judging social?
Paid social can be read within weeks because ROAS shows up fast. Organic needs ninety days to prove the system and a few quarters to show its real return. Judging organic on a thirty day window is the most common way I see good strategies get cancelled early.
Which attribution model should I use?
Pick one source of truth and one window and keep them consistent. Platform dashboards overclaim, last click underclaims, and your CRM sits somewhere between. Consistency matters more than precision because what you need is a trend you can trust month to month, not a perfect number nobody agrees on.
Let's talk
If you want social media reporting that leads with a number you would recognise on a bank statement, email me at info@robbp.com. Tell me what you sell, what a customer is worth, and what your current reports look like, and I will give you an honest read on whether the return is there.