The 90-Day Social Media ROI Test Any Agency Should Pass

Most small business owners cannot answer one simple question: "Did your last $5,000 in social media spend generate any revenue?" They can tell you their follower count. They can tell you that one Reel "did really well." They cannot tell you what came back in the door.

That is not an accident. It is how most agencies have been allowed to operate for a decade. The fix is a structured 90-day test that anyone, including your current agency, should be willing to pass.

Ninety days is enough time to see real movement, short enough to recover the spend if the answer is no, and clear enough to give you an honest line in the sand: this either produced leads, or it did not. Print this. Hand it to your agency. If they push back on any step, you have your answer before the test even starts.

Step 1: Set Your Baseline (Before You Spend Another Dollar)

You cannot prove growth without a starting line. Most agency tests fail before day one because nobody wrote down what "normal" looked like. Spend one week pulling baseline numbers for the 90 days before the test starts.

Here is the minimum you need to record:

• Total inbound calls per month (from your phone bill, CRM, or call tracking)
• Total form submissions per month from your website
• Total booked jobs or paid invoices per month
• Average revenue per new customer
• Foot traffic or walk-ins, if you have a physical location
• Website sessions from social channels (Google Analytics 4, Acquisition report)

Write the baseline numbers down on one page. Tape it to the wall. This is the document the entire 90-day test gets graded against.

Step 2: Pick ONE Primary Goal

The number one reason social media spend disappears without a trace is that the goal was never defined. A good test has one primary goal, written as a number, with a deadline.

Not "grow the brand." Not "build community." Not "post more consistently." Those are activities, not goals. Pick one of these instead:

• Leads: "Generate 25 qualified phone calls or form fills attributed to social by day 90."
• Sales: "Close $15,000 in new revenue traceable to social by day 90."
• Bookings: "Add 40 new appointments to the calendar with a social-source tag by day 90."
• Foot traffic: "Drive 100 redemptions of a social-only coupon by day 90."

There is no universal return target. When I set one in a digital strategy plan, I start from four inputs: your gross margin on a typical job, your close rate on leads, the length of your sales cycle, and a comparable baseline period. A business with high margins and a short sales cycle can justify spend that a low-margin business cannot. Work out how much trackable revenue the channel needs to produce to cover its ad spend and management fee at your margin, set your target above that line, and write it down before day one.

Step 3: Track the Right Channels (So Nothing Hides)

You cannot grade what you cannot see. Before day one of the test, install three pieces of plumbing. None of these are optional.

UTM-Tagged Links on Every Post

Every link you push from social, paid or organic, gets a UTM tag. Use Google's free Campaign URL Builder. A typical tag looks like ?utm_source=facebook&utm_medium=social&utm_campaign=spring-promo. Now every click that lands on your site shows up in Google Analytics 4 under the exact campaign that sent it. No more "direct" or "(not set)" mystery traffic.

Call Tracking Numbers

Use a service like CallRail, CallTrackingMetrics, or a low-cost Twilio number. Assign one tracking number to your social bio links and one to your paid ads. When the phone rings, you know which platform earned it. Recording calls also lets you grade call quality, not just quantity.

Unique Offers, Codes, or Landing Pages

If you are running social-only promotions, give them a coupon code (FB10, INSTA20) or their own landing page (/spring-special). Brick-and-mortar businesses can print "Show this post at the counter" coupons. Every redemption is a vote that social drove the visit.

Set this up in week zero. If your agency is not willing to install attribution before the test starts, the test is over. They are telling you they do not want to be measured.

Step 4: The 90-Day Week-by-Week Schedule

Here is the actual schedule. Anyone, in-house or agency, can run this. Hold your manager (or yourself) to it.

Weeks 1 to 2: Setup & Foundation

• Lock in baseline numbers from the prior 90 days
• Install GA4 conversion events for calls, forms, and bookings
• Provision call tracking numbers and swap into bios and ads
• Build UTM template and tag every outbound link
• Pick the one primary goal and write the target number on the wall
• Audit profiles: bio, link in bio, profile photo, contact info, hours

Weeks 3 to 6: Volume & Variety

• Post a minimum of 3 to 5 times per week per active channel
• Mix content types: short video, customer photos, before/after, behind the scenes, offer-driven
• Run at least one paid test per week, $20 to $50 minimum
• Track which post format drives the most link clicks (not likes)
• End of week 6: pull a midpoint report against baseline

Weeks 7 to 10: Double Down on What Works

• Identify the top 2 content formats and top 1 paid creative from weeks 3 to 6
• Cut spend on anything that produced zero attributed clicks
• Increase paid test budget on winning ads to $100+ per week
• Add a clear call-to-action to every post: call, book, message, visit
• Push the social-only offer hard with coupon codes and exclusive promos

Weeks 11 to 13: Final Push & Scoreboard

• Maintain the proven cadence. Do not change strategy in the last 3 weeks
• Pull final attribution report from GA4, call tracking, and CRM
• Tally redeemed coupons, booked jobs, and closed revenue
• Compare against baseline and against the original goal number
• Schedule the 90-day review meeting

Step 5: What Counts As Success

Day 90 arrives. You sit down with the numbers. Here is exactly how to grade the test, with no wiggle room.

Clear Pass

The test passes if all three of these are true:

• You hit or exceeded the primary goal you wrote on the wall in week one
• Trackable revenue from social met the return target you set from your margins, close rate, and sales cycle, counting the total spend (ad budget + management fee)
• Lead volume from social-tagged sources beat your comparable baseline period by the margin you set in week one

If you pass, keep doing what you are doing and start scaling spend on the winning formats. Add a second channel only after the first one is producing reliably.

Conditional Pass

You hit the lead goal but revenue is below your return target. If your sales cycle is longer than 90 days, some of that revenue may simply not have closed yet. Otherwise, this usually means the leads are real but not qualified, or your sales process is dropping the ball. Do not blame social yet. Audit your intake. How fast are you returning calls, are you following up on form fills, is your booking process clean? Run another 30 days with tightened sales follow-up before judging.

Clear Fail

Step 6: When to Fire the Agency (Or Change Strategy)

Many owners hold on to a non-working agency far too long. Here are the four moments that should end the relationship, no matter how nice the account manager is:

1. They refused to install attribution before the test started. They told you it was "not necessary" or "too complicated." That is a tell. Real agencies install measurement first.
2. The day-30 report is all about reach, impressions, and follower growth. Not one mention of calls, leads, or revenue. They are training you to grade them on the wrong scoreboard.
3. They cannot answer "which post drove the most leads this month?" If they do not know, they are not paying attention to the only metric that matters.
4. Day 90 missed the goal and the proposed fix is "more posts, more spend." Doing more of something that did not work is not a strategy.

Step 7: A Few Industry-Specific Notes

The protocol works the same way for everyone, but the right primary goal shifts by industry:

• Restaurants: Track coupon redemptions, reservations, and direction taps from social, not menu views.
• Spas and salons: Track booked appointments with a social-source field in your scheduler.
• Cleaning services: Track quote requests by source. Most leads come in by phone, so call tracking is non-negotiable.
• Fitness studios: Track free-trial sign-ups and intro-offer redemptions, not class photo likes.
• Chiropractors and clinics: Track new-patient form fills tagged from social, plus first-visit conversions.
• Contractors: Track quote calls. A single signed roof or kitchen job can pay for the entire 90-day test several times over.
• Real estate: Track buyer-guide downloads and showing requests, not listing-photo engagement.

If you serve a specific Maine market, geo-tag your paid posts and target the towns you actually serve, whether that's Portland, Bangor, Ellsworth, or Augusta. A tight local audience keeps your budget on people who can actually become customers, instead of spreading it across the whole state.

Step 8: Trust the Numbers, Not the Story

The hardest part of this test is not running it. It is being honest about what the scoreboard says at day 90. Account managers will tell you a great story, share the one Reel that "went viral," and point to follower growth. None of that pays your rent.

The test you wrote on the wall in week one is the only thing that matters. Either trackable revenue hit the return target you set, or it did not. Either call volume beat your comparable baseline, or it did not. The numbers are the verdict. If you have been running social for a year without ever doing this exercise, you do not know whether it is working. You only know that you are still paying for it.

This same approach is how I think about every channel, from Google rankings to email to ads. I have written more about how to spot when paid efforts are not landing in why your competitors are ranking higher and how to plan a smarter mix in the small business marketing budget for 2026. If reviews are part of your social mix, my piece on getting more online reviews ties in directly. If you are still working on local search fundamentals, read the seven local SEO mistakes first. Fix those before testing social, or your traffic will leak out the back door.

Bottom Line

Ninety days is enough time to know whether your social media is a real lead source or an expensive habit. Set the baseline, pick one goal, install attribution, run the schedule, score it honestly. If your current agency is willing to be measured this way, keep them. If they are not, you have your answer.

If you would rather have someone run this test for you, I do exactly that. My AI integration and social media help can include the full 90-day protocol: baseline pull, attribution setup, content production, and a real scoreboard at day 90. No vague "engagement" reports. No "brand awareness" excuses. Just calls, leads, and revenue, tagged to the source.

Send me your social accounts (email me or text me at (207) 598-1219) and I’ll do a free audit of your current social presence. I’ll tell you whether the 90-day test is worth running, what realistic targets look like, and what I would do differently in week one. No pressure, no jargon. Just a straight read on whether your social media is earning its keep.