Campaign attribution answers whether a specific email or SMS campaign you sent produced bookings, opportunities and revenue, not just opens and clicks. Opens tell you a subject line worked. They do not tell you the campaign paid for itself. This article covers how to attribute outcomes to individual campaigns sent from the levelpartner.ai platform, how to judge them fairly with a holdout group, and where the method breaks down.
Why opens and clicks are not attribution
Email open figures are unreliable because many mail apps load images automatically. Clicks are better, but a click is interest, not an outcome. What a growing company needs to know is how many people the campaign moved to the next commercial step: a reply, a booked call, a new opportunity or a payment.
On the platform, campaign sends, replies, bookings and opportunities all sit on the same contact. That makes it possible to link the campaign to the outcome directly instead of guessing from aggregate figures.
Before you start
- Email and SMS campaigns sent from the platform's Marketing area, not from a separate tool.
- Booking calendars and the pipeline running on the platform.
- A tag naming convention for campaigns, for example the month and the offer.
- An outcome window agreed in advance: how many days after the send you will count results.
How to set it up
- Before sending, tag everyone on the send list with the campaign name.
- Split the list randomly and tag a small holdout group.
- Use tracked links that carry the campaign name in every message.
- Send the campaign to the main group only.
- After the agreed window, filter Contacts by the campaign tag.
- Count bookings, new opportunities and payments created since the send.
- Run the same count on the holdout group.
- Add the Lead Source Report to a dashboard, filtered on the campaign tag.
- Add a second copy filtered on the holdout tag, then Save Changes.
The holdout group is the whole point
Some of the people who booked after your campaign would have booked anyway. Without a holdout you credit the campaign with all of them. With a holdout, the difference between the two groups is what the campaign actually caused. It is the closest thing to an experiment most firms will ever run, and it costs almost nothing.
You give up the revenue the holdout group might have produced had they received the campaign. On a list of thousands, holding back a small share is a cheap price for knowing what works. For broader models of credit, read how multi-touch attribution shares credit.
Say the firm sends a reactivation email to 1,800 of 2,000 dormant contacts and holds back 200. Within 30 days, suppose 38 recipients book a call, about 2.1 in every hundred, while 2 of the 200 in the holdout book, 1 in every hundred. At the holdout rate, about 18 recipients would have booked anyway, so the campaign would have caused roughly 20 extra calls, not 38. The numbers are illustrative.
What it does for a firm with flat revenue
Firms that have stalled often send more campaigns in the hope that volume brings growth. Without attribution, every campaign looks like it worked because some bookings followed it. Campaign attribution with a holdout shows which campaigns create new conversations and which simply annoy contacts who would have called anyway. You send fewer, better campaigns and protect your list.
It replaces the campaign report that lists opens, clicks and unsubscribes and stops there. For the dashboard design, see building a marketing campaign dashboard.
Time it takes
Tagging, holdout and links add about fifteen minutes to each campaign. Counting outcomes after the window takes another fifteen. Over a quarter, that small cost tells you which campaign types deserve repeating.
Teams set the outcome window after seeing the results, choosing whichever window makes the campaign look best. Agree the window before you send, and use the same one for every campaign of that type.
When not to use a holdout
Do not hold back service messages, appointment reminders or anything a client needs. And for very small lists, the holdout group will be too small to show a reliable difference, so compare campaign types over several sends instead.
Troubleshooting
The campaign tag is missing from some recipients. The tag was applied after a list change. Apply it in the same step as the send, from the final list.
Bookings cannot be linked to the campaign. People booked through a link that was not tagged, such as your website's main calendar. Count bookings by the campaign tag on the contact, not only by link clicks.
The holdout group outperforms the campaign. The campaign may genuinely be putting people off, or the groups are too small. Repeat the test before drawing conclusions, and check the message for anything that reads as pressure.
Frequently asked questions
Can I attribute revenue to an SMS campaign?
Yes, with the same method. Tag recipients, use trigger links and count opportunities and payments created within your window. SMS often shows quicker results than email.
How large should the holdout group be?
Large enough that a few bookings either way do not swing the result. On small lists, run the same campaign type several times and add the results together.
Does campaign attribution replace channel attribution?
No. Channel attribution tells you where contacts first came from. Campaign attribution tells you what moved existing contacts forward. Use both, as described in marketing attribution.