This guide is for the marketing lead at a small company, or the founder who does the marketing between everything else, who has just run an email promotion, a paid social push or a webinar and been asked a blunt question: did it work? If your honest answer involves impressions, likes and a feeling that it went well, the problem usually started before launch, not after.
Measurement is mostly a planning exercise. Decide what success looks like before the campaign goes out, set up the tracking that will show it, and the analysis afterwards takes an hour instead of a week.
Decide the one outcome that matters
Every campaign should have a single primary goal that connects to money or a clear step towards it. Secondary metrics are fine, but one number should settle the question.
Pick from the bottom of the funnel where possible
A sale, a booked demo, a trial sign-up, a quote request or a newsletter subscriber who later buys. These are harder to move than clicks, which is exactly why they are more honest.
Write the target down before launch
“Fifty demo bookings from the October email series, against an average of twenty a month” is a target. “Raise awareness” is not. Writing it beforehand stops anyone redefining success once the results are in.
Name the secondary signals
Click-through rate, cost per click, landing page conversion rate and unsubscribes help explain why the main number moved. List two or three, no more.
Set a baseline
A result means little without something to compare it with. Before launch, record what normally happens over a comparable period without the campaign.
| Metric | Typical month | Campaign month | Change |
|---|---|---|---|
| Demo bookings | 20 | ? | ? |
| Website sessions from email | 1,200 | ? | ? |
| Landing page conversion | 2.5% | ? | ? |
| Cost per booking | n/a | ? | ? |
The figures above are placeholders; fill them from your own analytics. Seasonality matters, so compare with the same period last year where you can. A December campaign for a retailer will look brilliant against a quiet October.
Tag every link
Most analytics confusion comes from traffic arriving without labels. Google Analytics reads campaign parameters, known as UTM parameters, that you add to the end of a URL. Google’s help article on URL parameters describes the main ones:
- utm_source: the referrer, such as google, newsletter4 or billboard
- utm_medium: the marketing medium, such as cpc, banner or email
- utm_campaign: the product, slogan or promo code, such as spring_sale
The same article says that when you add parameters to a URL you should always use utm_source, utm_medium and utm_campaign, and recommends using only one unique utm_source for each distinct platform and one unique utm_medium for each distinct channel.
Keep a naming sheet
Create a shared spreadsheet that lists approved values: lowercase only, hyphens or underscores but not both, one name per platform. “Facebook”, “facebook” and “fb” will appear as three different sources in reports, which splits your data.
Build links in one place
Use a single URL builder, whether Google’s Campaign URL Builder or a tab in your spreadsheet, so nobody hand-types parameters.
Test before sending
Click every tagged link in a test email or ad preview and check that the visit appears in your analytics real-time report with the right labels.
Decide what success looks like before launch. Afterwards, everyone will want to redefine it.
Track the conversion, not just the visit
Tagged links tell you where visitors came from. You also need to record what they did.
Set up a conversion event
In Google Analytics or your ad platform, mark the action that matters, such as a form submission or a purchase, as a key event or conversion. Test it with a dummy submission.
Connect to the CRM if you sell to businesses
For longer sales cycles, the real result arrives weeks later, when a lead becomes a customer. Make sure the campaign source is saved against the lead record so you can trace revenue back to it.
Use unique offers for offline channels
Print, radio, events and podcasts do not carry clickable links. A dedicated short URL, a discount code or a phone number used only in that channel lets you attribute responses.
Read the results honestly
When the campaign ends, give it time to finish working. Email responses often trail off within a few days, but considered purchases can take weeks.
Compare with the baseline
Did the primary metric rise beyond normal variation? A jump from 20 to 23 bookings may be noise; a jump to 50 is not.
Calculate cost per result
Divide total spend, including design, tools and staff time where you can estimate it, by the number of primary conversions. Compare with what a customer is worth to you.
Look for side effects
Did unsubscribes spike? Did organic or direct traffic also rise, suggesting people searched for you after seeing an ad? Did sales from another channel dip, suggesting the campaign simply moved existing demand?
Run a holdout if you can
For email campaigns especially, sending to most of the list and holding back a random 10 percent gives a clean comparison. The difference between the two groups is the effect of the campaign itself.
Attribution without the headache
Customers often see several touches before they buy: an ad, an email, a search. Analytics tools use attribution models to divide credit among them, and different models give different answers. For a small team, two habits keep this manageable:
- Use the same attribution setting every time, so campaigns are comparable with each other.
- Ask new customers how they heard about you, with a simple dropdown or free-text field on your form. Self-reported answers are imperfect but often reveal word of mouth and podcasts that tracking misses.
Common measurement mistakes
Changing several things at once
If a campaign launched with a new landing page, a new offer and a new audience, a good result cannot be pinned on any one of them. Change one major variable at a time where you can, and note everything that changed.
Ending the campaign too early
Paid social and search campaigns often need a learning period before results settle. Judging after two days leads to switching off things that would have worked.
Celebrating vanity metrics
Reach, impressions and follower counts are easy to grow and hard to connect to revenue. Report them as context, never as the headline.
Ignoring small samples
Ten conversions against eight is not a trend. Before declaring a winner in an A/B test, wait until each version has enough results; many testing tools show a confidence level, and it is worth waiting for it.
Forgetting the cost of time
A webinar that took three people two weeks to prepare had a real cost even if the platform was free. Estimate staff hours, at least roughly, so cheap-looking channels are judged fairly.
Report on one page
Executives and clients rarely read long decks. A single page that answers four questions is enough.
- Goal: what we set out to achieve, with the number
- Result: what happened, against the baseline
- Cost: total spend and cost per result
- Next: what we will repeat, change or stop
Attach the detail for anyone who wants it, but lead with these four lines.
Campaign measurement checklist
- One primary goal with a written target
- Two or three secondary metrics named
- Baseline recorded for a comparable period
- UTM naming sheet agreed
- Every link tagged with source, medium and campaign
- Links tested in real-time analytics
- Conversion event set up and tested
- Offline channels given unique codes or URLs
- Holdout group considered
- One-page report template ready
Takeaway
A campaign is easy to judge when the judging is planned in advance. Pick one outcome, record a baseline, tag every link consistently using the UTM conventions Google recommends, track the conversion itself, and report on a single page. The next campaign will be better because you will know, rather than guess, what the last one did.
