A number you can_verify

01

Definitions before the campaign

We start with the business goal and the number used to track it. Then we agree which events are collected, what they are called and where they come from. Every number gets a definition you can show to others.

The definitions also record the limits of tracking. Cookie consent, in-store sales and phone enquiries can fall outside web analytics, and that needs to be known before results are compared.

02

Setup and testing

We build the data collection: events, the data layer and the connections between advertising, web analytics and customer data. Tracking is tested before the first flight, because data that was never collected cannot be recovered later.

An ad platform and web analytics count things with the same name in different ways. A platform's clicks can include more than link clicks, so the report explains where the gap comes from.

03

Bringing the results together

Media results show who was reached, how often and what they did. A business metric, such as sales, bookings or contracts, shows the outcome. We also ask for figures from your own sales or booking system, because they do not depend on cookies.

Brand change does not show in web analytics. It is measured with a campaign study, described on the Research page.

04

Interpretation and the next decision

The numbers say more together than apart. If reach was delivered but interest did not rise, the options to investigate include the message, the audience and the level of frequency. We explain which conclusions the data can support.

The agreed figures appear in Business Dashboard in NØRR3 Marketing Engine, described on the Dashboards page. Estimating the effect of each channel is separate work, described on the Data Modelling page.

05

What our campaigns have shown about measurement

In our own campaigns, missing purchase or enquiry tracking has made the result impossible to assess, however many clicks came in.*

In our campaigns, the platform's purchase count has differed clearly from web analytics. A platform's clicks have included likes and profile visits, and a link click has not always loaded the page.*

The effect of a brand campaign has shown first in searches for the brand name and in share of search.* So search demand for the brand is a fast signal between research waves.

A target for leads, newsletter sign-ups or visits can be met without sales growing, and sales can grow while the lead target is missed.* So every campaign also gets a business metric.

* NØRR3's own campaign bank, campaign reports 2024–2026: each lesson recurs in campaigns across at least two industries and two advertisers.