To measure campaign ROI across several brands, make every brand tag and cost its campaigns the same way. One UTM naming rule, one cost sheet and one comparison table let you put a bakery and a software brand in the same row without lying about either.
TLDR: Write a UTM standard with four fields, log every cost in one sheet (media, creative and time), pull revenue from the same campaign tag, and compare brands on return per dollar. Fix the tagging before you analyse anything.
The usual failure is not the maths. It is that Brand A tags links ig_spring and Brand B tags them Instagram-Spring-Sale, so a roll-up shows two campaigns that are the same one. ROI is only as good as the labels.
Write a four-field UTM standard once
Four fields cover almost every case. Put the rule in a one-page doc and make it the only way links get built.
| Field | Rule | Example |
|---|---|---|
| utm_source | The platform, lowercase | |
| utm_medium | organic or paid |
organic |
| utm_campaign | brand-initiative-yyyymm | acme-springlaunch-202603 |
| utm_content | Format and variant | carousel-a |
The brand sits inside the campaign name, so one filter splits a roll-up by brand. Lowercase everything. Never add a field per brand.
Log cost in one sheet, with the same three columns
ROI needs a cost, and cost is usually the part people skip. Use one sheet for all brands with these columns: campaign tag, media spend, creative cost and team hours.
Turn team hours into money with one hourly rate for all brands. If the number is rough, say so in the sheet. A rough, consistent cost still gives a fair ranking between brands. An accurate cost for one brand and none for another does not.
Pull revenue by the campaign tag, not by the platform
Take revenue, leads or sign-ups from your analytics or CRM, grouped by utm_campaign. If the sale happens offline, give the sales team a field in the CRM for the campaign tag and ask them to fill it at first contact.
Expect gaps. Some visits come from dark social and arrive with no tag. Do not force them into a campaign. Show them as "unattributed" in the table so the reader sees the size of the blind spot.
Compare brands on return per dollar, one table
Build one table with the same columns for every brand:
- Campaign and brand.
- Total cost.
- Revenue or leads attributed.
- Return per dollar (or cost per lead).
- Share of results unattributed.
Sort by return per dollar. Then add a column that says what you will do next: reuse, revise, pause or escalate. A table that ends in a decision gets used. One that ends in a number gets filed.
Compare like with like. A brand selling a $20 product and a brand selling a $20,000 contract will not share a return metric. Compare each brand with its own previous campaigns first, then with others only on cost per qualified lead or return per dollar.
Check the data before you trust the ranking
A quick sanity check saves embarrassment. Ask whether the click totals match the platform's reported link clicks within about 10%, whether any campaign has revenue but no cost, and whether any tag appears under two spellings. A short guide to this step is in how to validate campaign performance data.
Keep the campaigns, tags and results together
When campaigns, UTMs and results sit in different places, ROI work becomes a monthly chore. In Mydrop you can plan each brand's campaign in the content calendar, keep every brand in its own workspace, and read results in performance analytics, with client reports you can schedule for each brand. For a scorecard layout to start from, see the engagement-to-ROI scorecard. Agencies comparing tools can read the best campaign tracking tools for agency ROI.
Retag your three most recent campaigns today
Take the last three campaigns per brand and rewrite their UTM tags into the four-field standard. Add their costs to the shared sheet and build the first comparison table. It takes an afternoon, and next month's report will be the first one where the brands line up. If you want one place for plans and results, start free with Mydrop.
Google review
Trustpilot review
Product Hunt comment
