You send one email to several markets. The template chooses the language, product recommendations and offer for each recipient. Your team maintains one campaign instead of a stack of copies.
Then someone asks which version brought in the most revenue. The email platform shows one campaign. Your web analytics shows visits from that campaign. Neither view necessarily tells you what each person saw.
We discussed this with Radoslav Matúška and Ján Šácha from Asteroad. Our shared takeaway was simple: decide what you want to compare before you send. The answer determines what you record in the email platform and carry through to the website.
Define the comparison first
“Which version won?” can mean several things. You might want to compare languages, markets, product categories or offers. If three blocks change independently, each recipient can receive a different combination. A result for the whole email cannot tell you which block contributed to it.
Start with a question your team could act on. For example: How did order rates differ between recipients shown free shipping and those shown a percentage discount in the Dutch market? That calls for at least a market value and an offer value. If you also want to compare the recommended product category, record that separately. This comparison describes an observed difference; finding out whether the offer caused it requires a controlled test.
| Comparison | Value to preserve at send time |
|---|---|
| Market or language | The market or language assigned to the recipient |
| Offer | The offer variant actually shown |
| Product block | The category or recommendation variant actually shown |
Record the version each recipient received
If one defined segment always gets one version, you may be able to compare results using the segment recorded for each recipient at send time. Someone’s current segment may have changed since then. Asteroad’s Radoslav pointed to another option for more flexible emails: add custom parameters to the send event, where the platform supports them.
For example, the send could carry market=nl, offer=free_shipping and hero_category=running_shoes. Your reporting can then group recipients by the content they were sent, even though the email platform treats the campaign as one send.
This needs a little template discipline. If marketers copy a campaign and change the offer but forget to update its reporting label, the data will describe the old offer. For recurring emails, define the variant names in one place and include them in the test-send checklist.
If you need help building that into a campaign, Asteroad can set up the audience rules, dynamic content blocks and variant values in the email platform as part of campaign development.
“It needs some preparation and it needs some definitions up front.” — Radoslav Matúška, Asteroad
In SEINō, custom dimensions can be used to compare results when the relevant metadata is available from the source. For example, you could compare click rate by offer within each market, even when everyone belongs to the same campaign. Our Deployteq segment-reporting guide shows one concrete example of passing values at send time. The fields and setup will differ by email platform.
Keep the distinction when someone clicks through
Recording send metadata helps with email engagement. Revenue analysis needs the distinction to survive the click to your website.
Suppose every version uses utm_campaign=november_sale and nothing else identifies the variant. Website analytics can show the campaign’s traffic and orders, but it cannot split those results by the offer each visitor saw. Give each relevant combination of market and content a stable identifier in the URL tagging, such as a defined utm_content or utm_term value. Preserve what each identifier means in your reporting, and test how it appears in your analytics setup.
Worked example: a Dutch offer
- At send time: The email platform records
market=nlandoffer=free_shippingfor recipients sent the free-shipping version. For the comparison group, it recordsmarket=nlandoffer=percentage_discount. - In the links: Both versions use
utm_campaign=november_sale. The free-shipping links addutm_content=nl_free_shipping; the discount links useutm_content=nl_percentage_discount. Keep recipient IDs and sensitive segment labels out of URLs. - In the report: Group sends and clicks by the recorded offer, then group website visits and orders by the corresponding
utm_contentvalue. You can see the two Dutch offers side by side, provided the email platform exports the send-time values and website analytics preserves the URL tag through to the order.
Now you can connect three parts of the story: what was sent, how people engaged with it and what happened after the click. Check an actual link from each important test version before launch. A correct template value is of little use if the link setup drops it.
More changing blocks create harder questions
When several blocks change independently, the combinations multiply. You may know that recipients of one combination bought more, but that does not establish which individual block caused the difference. Market, audience and offer can all affect the result.
“One shop can sell products for very different interests. Personalising the content around those interests becomes more complex when you also operate across markets and languages.” — Ján Šácha, Asteroad
Record the dimensions you will use to make decisions, then compare like with like where possible. If the question is about offers, look at offers within the same market or audience before drawing a broader conclusion. Treat a pattern in campaign reporting as a reason to investigate or test, rather than proof that one content block caused the result.
Not every campaign needs this setup. Two versions of a one-off email may be quicker to send and compare separately. The case for dynamic content becomes stronger when a recurring flow runs across many markets and maintaining separate copies would mean updating and testing the same logic repeatedly. Either way, plan the measurement alongside the content.
For a recurring journey, Asteroad can also review the existing platform setup and help implement or improve the automation, including its market conditions and personalised content.
Before you send
- Write down the comparison you want to make after the campaign.
- Name the market and content variants consistently.
- Confirm which values your email platform can attach to the send event.
- Carry the relevant variant value into website campaign tagging.
- Follow links from test emails through to your analytics and check the values there.
- Interpret combinations carefully when several blocks change at once.
Dynamic content can make a campaign easier to run. A small amount of measurement planning makes its results easier to understand. The useful question is not only what did this email do? It is what was each audience sent, and what happened next?