Measurement Before Spend

Set the baseline before you spend, so you can tell whether a campaign worked.

By James Schramko · Updated October 2026

The Diagnosis That Almost Happened

For years I made short social videos without knowing what they were worth. They looked like the softest activity in the business. The kind of thing you stop doing the moment a quarter gets busy.

Then I got clean attribution running through Wicked Reports. Those little videos showed up driving about $38,000 a month in tracked, attributed revenue. A channel I could not have valued from inside the dashboard I had before.

Run that forward. If a busy season had hit before I had that data, cutting the videos would have looked like discipline. I would have fixed a time problem that was really a revenue channel. The numbers would have looked like progress. Then the sales those videos were quietly producing would have disappeared with them.

That is the risk in every business running without measurement. The gap hides what is working. Worse, it points you at the wrong fix, with total confidence. There is nothing in front of you to contradict the guess.

The Principle

Spending without measurement is guessing with a budget. Once you can measure, the same money becomes a test, and you learn something whether it wins or loses.

Before any paid traffic, organic push or campaign launch, you need three things: a working measurement layer, a baseline conversion rate from existing traffic, and clear source attribution for new traffic. Build those first. Then spend.

The Non-Negotiable Checklist

Before running paid or major organic traffic to an offer, check each of the following.

1. GA4 or equivalent analytics is firing correctly

Pageviews tracked. Key conversion events firing: lead, opt-in, purchase, booking. Ecommerce events set up if they apply. If you cannot see the traffic, you cannot diagnose it. Fix this before you spend.

2. Email click traffic is attributable to a specific source

UTM parameters on every email link. Source tags separating lists, sequences and campaigns. Without this, email could be doing all the work or none of it. You would have no way to tell which.

3. Opt-in sources are tagged at capture

Form fields or hidden inputs capturing source, medium and campaign. Records in your CRM showing where each subscriber came from. If every lead reads "website opt-in," you have no source data to act on at all.

4. Baseline conversion rate is established

At least two weeks of data on the existing offer or page. Write the number down. A guess does not count. You cannot tell whether a campaign improved things if you never knew where they stood before.

5. Diagnostic-stage conversion is tracked

If the offer is a paid diagnostic or entry product, track the conversion from diagnostic to main offer on its own. That rate is the real unit economics. The lead-to-diagnostic rate tells you less. Pricing a diagnostic without it is a guess with a number attached.

Common Measurement Gaps

These show up on repeat.

The unattributed click. Real traffic from email or social, no source tag. You can see the clicks. You cannot see which email, post or campaign sent them.

The orphan opt-in. Leads arrive with no source information. The form never captured it and the CRM never filled it in. The team guesses.

The dashboard with no signal. Plenty of metrics, none of them tied to a decision. The numbers exist. They do not move anything.

The pre-campaign blind spot. A campaign launches before anyone checks that tracking works. The campaign runs. The data comes back empty. Nothing about it can be judged.

The diagnostic without downstream tracking. A paid diagnostic priced on feel, with zero visibility into how many buyers go on to the main offer.

The Fix Sequence

Fix a measurement gap in this order. Skip a step and the next one has nothing solid to stand on.

Get the events firing. Analytics, conversion events, source tagging, checked with a real test event.
Set the baseline. Two weeks of normal traffic. Record conversion rate, source breakdown and lead quality.
Tag every new traffic source. UTM discipline on every email, ad, social post and link.
Then spend. Now there is a measurement layer that can read the result.

Then Spend: Boost Only What Already Converted

Most owners boost the post that looked good. High views, lots of likes, a comment thread. Money goes behind it and nothing comes back.

The views were the wrong number. Here is the sequence I use instead.

Post one short a day. No budget. Let each one run on its own.

Track actions. Views can wait. An action is an inquiry, an opt-in, a booking or a paid step. A view does nothing for your bank account.

Boost only a piece that already produced an action. One sale, one lead, one booking. That piece has proven it can move someone. Put about $5 a day behind it and nothing else.

Keep going while the actions keep coming. Check weekly. Stop the first week that produces none.

One of my clients had a short sitting at about 3,500 views. Modest by any feed standard. It produced a paid diagnostic sign-up. That is the one that gets money behind it. The one with 40,000 views and no action gets nothing.

Small budget, proven piece, weekly check. The spend is a test that already passed once. You are paying to see if it passes again.

The Reframe

When a campaign underperforms, look at the measurement layer before you touch the campaign itself. Did the events fire correctly the whole way through. Can the traffic be tied to a specific source. Is there a real baseline to compare against. Is the conversion data complete, or is someone guessing from half of it.

Often a campaign called a failure ran within a normal range. A gap in the measurement created the look of failure or success, more or less at random.

It is hard to spot which number in your own dashboard is quietly lying to you. You already trust the dashboard you built. Working out which figure to distrust this month is a real chunk of what I do with Mentor clients. It happens before they spend another dollar on anything.

When You Cannot Get the Data

Sometimes the measurement layer is hard to build. The platform will not support it, the integration is broken, or the tracking is too complex to land in time.

In that case, say so plainly. You are running this campaign without full attribution. You will treat the result as a pointer rather than a verdict. You will make the next decision with that limit named up front. That beats running the campaign as if the data exists and then deciding on numbers you never had.

The Standard

Set the baseline before the algorithm gets any spend. Check attribution before you scale anything.

Spending before measuring costs more, teaches less, and hands you decisions built on noise. It feels like the bold move the whole time. The discipline is dull. It is also what separates marketing you can learn from and marketing you are gambling on. Get the measurement layer right first. Then the harder call, which offer or model the business should run on, gets easier to see. See Product Client Matrix for that next decision.